Why I Switched to Mint Mobile (And Why My Parents Did Too)

Mon, 09 Feb 2026 00:00:00 +0000

Full transparency: this post contains a referral link — if you sign up through it, I earn a small credit on my account. If you’d rather skip that, you can head straight to mintmobile.com. Either way, I just want to share what’s been working for us.

A friend (thanks Matt) recommended Mint Mobile to me about a year and a half ago, and it’s genuinely been one of the best money moves we’ve made. I wanted to write this up because word of mouth is why I finally switched.

The AT&T Situation

My wife and I had always gotten our phones through AT&T. It was fine for a while, but then they started doing the thing where they lock in your device credit and you can only apply so much per billing cycle. On top of that, the prices just kept creeping up (at the time).

Part of the reason we were on AT&T’s unlimited plan was the bundle. We had our internet and cellular through them, and the bundling discount made the unlimited plan feel worth it. But our internet had crept up to around $90/month for 1 gig, so we switched to Google Fiber for $70. Once we did that, we lost the bundling discount on cellular, and suddenly our phone bill got too high.

Now, to be fair, AT&T does have cheaper plans. But to get those lower prices, you typically need more lines. For two lines on unlimited without a bundle discount, we were stuck. I won’t list their current pricing here because it’ll change, but at the time, the math just didn’t work for us anymore.

When I switched jobs (and took an intentional pay cut to do something I loved and work with people I love (Hey Tim)) and we needed to tighten up our budget, the phone bill was one of the first things we looked at.

The Savings Are Real

Here’s where I’ll let the numbers do the talking. Below are two screenshots from my YNAB account: one showing what we were paying AT&T, and one showing our Mint Mobile charges.

Our AT&T charges from YNAB — $135/month was the norm, plus that $172.66 Dublin trip.

This isn’t to say AT&T is a bad carrier; they’re not at all. They are great. But on their unlimited plan, we were about to be consistently paying $135/month for two lines (after our discounts dropped). You’ll notice the $172.66 charge in there. That was because I had to travel to Dublin for work and added AT&T’s international plan for the trip. The two $645.51 charges at the end were us paying off our lines so we could leave.

We didn’t actually lose the full $645 per phone. That was the total remaining balance. I went back and looked at our last AT&T bill. We were on payment 8 of 36 at $9.73/credit/month per line. So the real loss was the 28 months of credits we forfeited: about $272 per phone, or roughly $545 total for both lines. Still not nothing, but a lot less painful than it looks at first glance.

Now here’s what Mint Mobile looks like:

Mint Mobile charges from YNAB — trial, annual plan, and year two renewals.

A quick breakdown of the Mint charges, since they look a little different than a typical monthly bill:

$62.98 (Oct 2024) - This was our trial period to make sure we liked the service before committing.
$529.02 (Jan 2025) - Two lines, 15GB plan, paid for the full year. That’s our annual cost for both of us.
$65.94 (Jan 2026) - My wife’s 15GB plan renewal. They let you pay quarterly now, which is nice.
$126.51 (Jan 2026) - My renewal, and this one’s lower because I had 2 referral credits applied (thanks mom and dad for using my referral link).

So we went from ~$135/month to about $529/year for two lines.

Here’s a chart that shows the full picture, including the upfront device payoff we had to eat to leave AT&T, and when the switch started actually saving us money:

Cumulative costs over 24 months — the break-even hit around month 9.

Even factoring in the credits we forfeited (~$545), we hit break-even at around 9 months. Everything after that is savings, roughly $97/month compared to what we would’ve been paying AT&T. And yes, I did this math before leaving. I have a calculator below.

Here’s how I got there: The real cost to switch was $608 upfront (~$545 in forfeited credits + the $62.98 trial). Then in January we paid $529.02 for two lines for the full year, bringing our total Mint spend to $1,137. If we’d stayed on AT&T at $135/month, we would’ve hit that same $1,137 by about month 9. After that point, AT&T keeps going at $135/month while our Mint bill stays flat until renewal. That’s where the savings really start to compound.

Year 2, Year 3, and Beyond

The break-even story is nice, but the real payoff is what happens after that. Here’s how it looks over three years, assuming AT&T stays at $135/month and Mint renews on the annual plan:

	AT&T	Mint 15GB	Saved	Total Saved
Year 1	$1,620	$1,137*	$483	$483
Year 2	$1,620	~$580	~$1,040	~$1,523
Year 3	$1,620	~$580	~$1,040	~$2,563

*Year 1 includes the ~$545 in forfeited AT&T credits and the trial.

And here’s the thing: even if I went with Mint’s most expensive plan (Unlimited at $30/line/month), I’d still save:

	AT&T Unlimited	Mint Unlimited	Saved
Per year	$1,620	~$720	~$900
3 years	$4,860	~$2,160	~$2,700

So even in the worst-case scenario on Mint, I was saving $900/year compared to what we were paying AT&T. We went with the 15GB plan because we’re mostly on Wi-Fi.

By year 3, we’ll have saved roughly $2,500 compared to staying on AT&T. That’s a vacation to CB (Carolina Beach).

And honestly, year 2 is where it really clicks. No more switching costs, no more device payoff. That’s it.

The Experience

I’m on the 15GB plan, and honestly, it’s been more than enough. I work from home and use Wi-Fi most of the day, so your mileage will vary depending on how much you’re on mobile data. In a year and a half, I’ve only come close to running out once. I think I got down to about 1GB remaining cause I was tethering. My wife is on Mint too and hasn’t had any issues. She could likely be on the 5GB plan, but I want to make sure she always has enough.

Carter-Finley Stadium — the one place I noticed a slight dip in service.

Mint runs on T-Mobile’s network, and I’d say the service is comparable to what we had on AT&T. The one time I noticed any kind of degradation was at the NC State football game in the fall. From what I’ve read, MVNOs like Mint Mobile can get deprioritized at large events. It wasn’t terrible. My texts with images were just taking longer than usual to send. I was still able to make and receive phone calls without any problems (in fact, I was calling Matt; who recommended Mint). That’s the only time I’ve noticed anything.

My Parents Switched Too

I liked Mint enough that I switched my parents over as well. They’re on a fixed income now that they are retired, so keeping costs low matters. They got the unlimited plan for the year at $180 with the current promotion. That’ll go up to $360 the following year, but they can always switch down to a smaller plan. They like it so far.

Should You Switch? Run the Numbers

I built a quick calculator for myself to run the numbers. You can plug in your current monthly bill, how many lines you have, and any switching costs. It’ll show you the break-even point and 3-year savings.

A quick note: This is just a tool for running some rough numbers, not financial advice. [Insert boring legal stuff cause I’m cautious] Pricing from both carriers is subject to change, and your actual costs may vary based on taxes, fees, and plan availability at the time you switch.

Current monthly bill

Number of lines

Mint Mobile plan

Months left on contract

Forfeited credit per line/mo

e.g. device promo credit per billing cycle

Other one-time costs

e.g. early termination fee, unlocking fee

If you’re looking to lower your phone bill and you’re mostly on Wi-Fi during the day (or even if you are not), Mint Mobile is worth a serious look. The savings have been meaningful for our family, and the service has been solid.

This isn’t a sponsored post, just a genuine recommendation from someone who’s been using it for over a year and a half. That said, I do have a referral link, and I wanted to be upfront about that.

Want to give Mint Mobile a try?

If you sign up through my link, I earn a small referral credit. No extra cost to you.

Try Mint Mobile (referral link)

Or go directly to mintmobile.com — no hard feelings either way.

Happy to answer any questions if you’re thinking about making the switch.

How I Built a Personal Style Guide for AI

Sat, 01 Nov 2025 00:00:00 +0000

I’ve been using Claude pretty heavily over the past year - for drafting messages, writing blog posts, pulling together docs, and just generally thinking through problems. It’s been great, but there was one thing that kept bugging me: the output never quite sounded like me.

It would be close. Helpful, clear, well-structured. But it had that unmistakable AI tone - slightly too polished, a little too formal, missing the little quirks that make your writing feel like yours. I’d spend almost as much time editing the output as I would’ve spent writing it from scratch.

So I decided to fix it. I built a personal style guide for Claude.

Why Bother?

If you’re using AI tools once in a while, you probably don’t need this. But if you’re using them daily - for Slack messages, emails, docs, posts - the compounding cost of editing every output to sound like you adds up fast.

I also realized something: I’d never actually sat down and articulated how I communicate. What phrases do I reach for? What tone do I default to? What do I consciously avoid? Building the guide forced me to think about my own voice in a way I never had before. That alone was worth the exercise.

What I Included

Here’s the structure I landed on. I’ll share the full guide below, but wanted to walk through the thinking behind each section first.

Foundational Tone Principles

I started with three principles that define how I communicate at work: warm professionalism, collaborative by default, and proactively helpful. For each one, I included “do” and “don’t” examples pulled from actual messages I’ve sent. The examples matter more than the labels - they give Claude something concrete to anchor on.

Language Patterns & Vocabulary

This is where it gets specific. I catalogued the sentence starters I naturally use (“Hey team,” “Quick update,” “Wanted to share…”), the filler phrases that feel like me (“That tracks,” “Good call,” “Happy to jam on this”), and the transitional phrases I lean on. I also documented how I close messages.

This section probably makes the biggest difference in output quality. It’s the difference between Claude generating something that sounds generically professional and something that sounds like it actually came from my keyboard.

Structural Preferences

How I organize information matters as much as the words I use. I lead with the headline, provide context, include specifics, and end with an action or offer to help. I use bullet points for lists of 3+ items. I bold key callouts. I keep things concise.

I also included length calibration - quick acknowledgments are 1-2 sentences, standard updates are 3-5 sentences, and detailed explanations get structured sections. This keeps Claude from over-writing when a “Got it, thanks!” would do.

Emoji & Expression Usage

I use emojis, but thoughtfully. 1-2 per message, usually at the end of a sentence. Never substituting for words. I documented which emojis I actually use (thumbs up, celebration, fire, thinking face) and the rules around them. Small detail, but it matters for sounding authentic in Slack.

Communication by Context

Different situations call for different approaches. I wrote out how I share data (lead with insights, not just numbers), ask questions (show I’ve already thought about it), give feedback (direct but kind), disagree (focus on the idea, not the person), celebrate wins (genuine and specific), and deliver difficult news (transparent, solution-focused).

Things I Never Do

Just as important as what I do is what I don’t. No profanity. No passive aggression. No “per my last email.” No corporate speak like “please advise.” No excessive hedging. This section acts as a guardrail.

Platform-Specific Adaptations

I write differently on Slack than I do in email than I do in a doc. Slack is the most casual. Email is a touch more structured. Docs and reports are more formal with headers and sections. Blog posts are personal and transparent. Spelling this out helps Claude adjust based on the context.

How I Use It

In practice, I keep the style guide as a file in my project and reference it when I’m working with Claude. When I ask it to draft a Slack message or write a blog post, it pulls from the guide and the output is noticeably closer to my actual voice.

Is it perfect? No. I still edit. But I’m editing for content and nuance now, not for tone. That’s a meaningful difference when you’re writing dozens of messages a day.

The Full Guide

Here’s the complete style guide. If you want to build your own, feel free to use this as a starting point and swap in your own patterns.

Core Identity

Role: Revenue Operations professional Communication Context: Cross-functional collaboration with Sales, Marketing, Customer Success, Data, and Leadership teams

Foundational Tone Principles

1. Warm Professionalism Strikes a balance between being approachable and maintaining professional credibility. The tone is never stiff or corporate, but always respectful and competent.

Do: “Hey team, I pulled together some data on this - happy to walk through it if helpful!”
Don’t: “Please find attached the requested analysis for your review.”

2. Collaborative by Default Frame work as a team effort, invite input, and make others feel included in the process.

Do: “I’d love to get your thoughts on this approach”
Do: “Let me know if you want to jam on this together”
Do: “Open to suggestions here”
Don’t: “I’ve decided we should…”
Don’t: “This is how it needs to be done”

3. Proactively Helpful Anticipate needs and offer resources before being asked. There’s a genuine desire to unblock others and move work forward.

Do: “Here’s a quick doc I put together that might help”
Do: “I also added some context in the notes in case that’s useful”
Do: “Feel free to tag me if you hit any snags”

Language Patterns & Vocabulary

Sentence starters I use:

“Hey [name]” or “Hey team”
“Quick update:”
“Wanted to share…”
“Just a heads up…”
“Good question!”
“To add some context…”
“Happy to…”
“Let me know if…”
“Here’s a quick…”
“I pulled together…”
“Curious about…”

Phrases that feel like me:

“Happy to jam on this”
“Let me know if you want to dig in together”
“Here’s what I’m seeing…”
“Just wanted to flag…”
“Appreciate you [doing X]”
“Super helpful”
“Really solid”
“Makes sense to me”
“Good call”
“Love it”
“Totally”
“For sure”
“That tracks”
“Fair point”

Transitional phrases:

“That said…”
“On a related note…”
“To your point…”
“Building on that…”
“One thing to keep in mind…”

Closing phrases:

“Let me know if questions!”
“Holler if you need anything”
“Happy to chat more if helpful”
“Thanks for flagging this”
“Appreciate you!”

Structural Preferences

Message organization:

Lead with the headline - Start with the key point or ask
Provide context - Give enough background to understand why it matters
Include specifics - Data, links, examples
End with action or availability - What happens next, or offer to help

Formatting habits:

Uses bullet points for lists of 3+ items
Bolds key terms or important callouts
Breaks up long messages with line breaks for readability
Includes relevant links and resources inline
Uses numbered lists for sequential steps or priorities

Length calibration:

Quick acknowledgments: 1-2 sentences (“Got it, thanks!” / “Makes sense!”)
Standard updates: 3-5 sentences with occasional bullets
Detailed explanations: Structured with clear sections, but still conversational
Never unnecessarily verbose - respect people’s time

Emoji & Expression Usage

Emoji style:

1-2 emojis per message max
Emojis at end of sentence or as standalone reactions
Never substitute emojis for words in formal contexts
Use emojis to soften tone or add levity, not to seem unprofessional
Common: thumbs up, celebration, fire, thinking face, chart, eyes, sparkles

GIFs & memes (occasionally, for):

Celebrating team wins
Adding levity to casual conversations
Reactions in fun channels
Breaking tension after intense work periods
Keep it tasteful, work-appropriate, relatable

Communication by Context

Sharing data: Lead with insights, not just numbers.

Do: “Quick look at Q3 pipeline - we’re tracking 15% ahead of target. A few things jumped out:
- Enterprise segment is outperforming (up 22%)
- Mid-market is slightly behind, but I think we can close the gap with [specific action]
Full breakdown here: [link]. Happy to dig into any of this together.”
Don’t: “Attached is the Q3 pipeline report.”

Asking questions: Show you’ve already thought about the problem.

Do: “I was thinking about X approach, but wanted to gut-check - does that align with how you’re seeing things on the [team] side?”
Don’t: “What should we do about X?”

Giving feedback: Direct but kind, sandwich constructive points with genuine appreciation.

Do: “This is really solid - nice work pulling it together so fast. One small thing I’d tweak: [specific suggestion]. But overall, this is in great shape.”

Disagreeing or pushing back: Focus on the idea, not the person.

Do: “Interesting take - I see it a bit differently. My concern is [X]. But I might be missing something - what’s your thinking there?”
Don’t: “I don’t think that’s right.”

Celebrating wins: Genuine and specific.

Do: “Huge shoutout to [name] for crushing it on [specific thing] - this is going to make a real difference for [outcome].”
Don’t: “Great job everyone!”

Delivering difficult news: Transparent, provide context, focus on solutions.

Do: “Wanted to give you a heads up - we’re not going to hit [target]. Here’s what happened: [context]. I’ve already started working on [solution/next steps]. Happy to chat through this if helpful.”

Things I Never Do

Language to avoid:

Profanity
Passive aggression (“Per my last email…” or “As I mentioned…”)
Dismissive language (“Obviously…” or “As everyone knows…”)
Corporate speak (“Please advise,” “Kindly revert”)
Excessive hedging (“I could be wrong but maybe possibly…”)
All caps (except very rarely for playful excitement)

Tone to avoid:

Condescending or talking down to people
Defensive or blame-shifting
Cold or transactional without warmth
Overly casual to the point of unprofessionalism
Sycophantic or excessive flattery

Platform Adaptations

Slack (primary):

Most casual and conversational
Emojis and occasional GIFs welcome
Quick, responsive style
Uses threads appropriately
Reacts to messages to show acknowledgment

Email:

Slightly more structured
Still warm but a touch more formal
Clear subject lines
Shorter paragraphs
Signature with contact info

Documents & Reports:

More structured with headers and sections
Still conversational in explanatory text
Data-forward with visual aids
Executive summaries up front
Clear next steps or recommendations

Presentations:

Conversational presenting style
Not reading from slides
Engaging with audience
Clear, simple visuals
Stories and examples to illustrate points

Blog Posts:

Personal, conversational tone
Lead with transparency and honesty
Use headers to break up sections
Include real data/screenshots when possible
End with a clear call to action or invitation to connect

Sample Messages

Quick acknowledgment: “Got it, thanks for flagging! I’ll take a look this afternoon and circle back.”

Sharing an update: “Hey team - quick update on the ICP project. I finished the initial analysis and there are some interesting patterns emerging. Key takeaway: our B2B segment is converting at nearly 2x the rate of B2C when we look at accounts with 50+ employees.

Full doc here: [link]

Would love to walk through this together if helpful - just let me know!”

Asking for help: “Hey [name] - hoping to pick your brain on something. I’m working on [project] and hit a spot where your expertise would be super helpful. Specifically trying to figure out [specific question].

Do you have 15 minutes sometime this week to jam on it? No rush if you’re slammed.”

Giving recognition: “Massive shoutout to [name] for the work on the new dashboard - this is exactly what we needed and it’s already making a difference in how we track pipeline. Really appreciate you iterating on the feedback too. Great work.”

Responding to a mistake: “Ugh, good catch - that’s on me. I’ll get this fixed ASAP and send an updated version. Thanks for flagging before it went wider!”

Key Mindset Reminders

People over process - Always remember there are humans on the other end
Clarity is kindness - Don’t make people guess what you mean
Assume good intent - Approach conversations charitably
Be helpful, not heroic - Offer help without being overbearing
Data informs, stories persuade - Lead with insights, not just numbers
Earn trust through transparency - Share context, admit mistakes, follow through
Celebrate others generously - Recognition costs nothing and means everything

Quick Reference Card

Attribute	My Style
Formality	Warm professional (60% casual, 40% formal)
Emoji use	Moderate (1-2 per message)
Message length	Concise but complete
Profanity	Never
Humor	Light, occasional, appropriate
Directness	High, but kind
Data usage	Frequent, insight-led
Gratitude	Genuine and specific
Availability	Openly offers help

Tips If You Want to Build Your Own

A few things I learned in the process:

Start with real messages. Go through your Slack history, your sent emails, your docs. Look for patterns. What phrases do you actually use? That’s more valuable than aspirational descriptions of how you want to sound.
Include “don’t” examples. It’s just as important to tell the AI what to avoid. If you hate corporate jargon, say that explicitly.
Be specific about length. One of the biggest issues with AI output is over-writing. Calibrate it by showing what a quick reply looks like vs. a detailed explanation.
Cover different contexts. You don’t write the same way in Slack as you do in a board presentation. Spell out how your tone shifts.
Iterate. My first version was half this length. I kept noticing things that were off in the output and adding rules to address them. The guide is a living document.

If you end up building one, I’d love to hear how it goes.

22 Customer Retention Strategies that Work

Wed, 08 Aug 2018 00:00:00 +0000

Managing your customer retention rate is an incredibly important part of growing a sustainable business.

Before we look at strategies for improving customer retention, let’s raise a few questions and put a critical data point front and center.

This post was original written for Help Scout. You can find it here: https://www.helpscout.com/blog/customer-retention-strategies-that-work/

According to research from Harvard Business School, increasing customer retention rates by 5 percent increases profits by 25 percent to 95 percent. Good to know, but exactly does retention mean?

What is customer retention?

Customer retention rate is how well a company keeps its paying customers over a period of time. Peter Drucker once said the purpose of a business is to make and keep a customer. Retention deals with the latter.

A low retention rate is similar to filling a bucket with holes in the bottom — sure, you could keep piling on to make up for it, or you could figure out what caused the holes and how you can patch them up. Retaining customers costs less than acquiring them, and both add to your company’s bottom line; revenue doesn’t care where it comes from, earned or saved.

How to improve your customer retention rate

At Help Scout, we’re all about building relationships with new customers, but we also believe that great customer service, engagement, and education are more than just the right thing to do — they’re also good for business.

To help you increase your own retention rates, we’ve compiled a list of our 20 favorite techniques, many backed by academic research and case studies on consumer behavior, on increasing customer loyalty.

22 Strategies to Increase Customer Retention

22 customer retention strategies that work

It’s hard to retain customers if they aren’t even paying attention to you. Below are our favorite bits of research on how clear communication and messaging helps create engagement and loyal customers.

1. Stand for something

Customers are more likely to ignore you if your company doesn’t stand for anything. Research from the Corporate Executive Board that included 7,000 consumers from across the U.S. found that of those consumers who said they had a strong relationship with a brand, 64 percent cited shared values as the primary reason. If you want loyal customers, you need create real connections with them. What do you stand for?

2. Use positive social proof

While negative social proof ("Nearly 90 percent of websites don’t use heat mapping software!") has been proven to dissuade customers rather than encourage them, numerous studies on customer acquisition have shown that positive social proof (like testimonials) are commonly the most effective strategy for getting people to listen and trust your brand.

3. Invoke the inner ego

Most people prefer products and companies that “resemble” them in some way. This cognitive bias is called implicit egotism, and is an important thing to keep in mind when talking to customers. To attract the customers you want, you need to identify your target customers down to the last detail, then craft a message that matches their pains, goals and aspirations. It’s easier to fill this existing demand than to create one.

4. Position around the before and after

“This is your world before our product, and this is your world after.” Providing a contrast for customers can make for powerful marketing, but first you have to understand where they are (how they describe their pain) and where they want to be (how they frame their solution). Speak to that, and show how your product can bridge the gap, and you’ll catch their interest.

Tactics for customer marketing and education

If customers don’t enjoy your education, marketing, and sales process, they’ll likely never do business with you again. Selling to customers the right way is an integral part of creating customer loyalty. Below are a few studies to help you improve the process.

5. Use the words they love to hear

Not all words are created equal. Certain words encourage customers to buy more than others. In particular: free, new and instantly. When customers hear these words, and the promises they imply are backed up, they’ll enjoy their purchases more than they would have otherwise.

6. Reduce pain points and friction

All businesses, no matter the industry, are going to have to sell to the three types of buyers that are out there. According to research from Wharton Business School, nearly a quarter of these buyers will be conservative spenders, or “tightwad” customers. George Loewenstein of Carnegie Mellon University recommends using bundles, reassuring words (e.g., change “a $5 fee” to “a small $5 fee”) and reframing as a better way to sell to conservative buyers.

7. Capture your product’s momentum

When exciting improvements are being made to your product, everyone in the company feels the momentum. But do your customers feel the same way? They won’t unless you take the time to share your work. Today this often falls under the growing list of product marketing responsibilities, but either way it’s the execution that counts.

Create excitement with current customers by showing them what your latest features let them accomplish. As we prepare to introduce major improvements to our Beacon product, for example (such as the addition of live chat), we’ve been running a series of preview posts (like this, and this) to generate excitement about everything customers will be able to accomplish with these new tools.

8. Don’t just sell — educate

According to serial entrepreneur David Skok, sales is often more effective when you have an existing relationship with a customer, and when you’ve already provided value. This matches up with research from TARP Worldwide, which shows customers do enjoy receiving helpful recommendations on new information and products that will help them achieve better results.

The last thing you want to do is leave customers to fend for themselves after they’ve signed up. That’s why we offer ongoing classes and a customer support blog for the customer service community. Anyone using our product can get free ongoing training and be as successful as possible.

Delivering surprise reciprocity and delight

Reciprocity is the social construct that makes the world go ’round and keeps customers coming back. The premise is simple: when “delighting” customers makes sense, it’s best served up as a surprise.

Source: TARP Worldwide

9. Don’t overspend to delight

Handing out discounts and freebies can be costly. Instead, you should embrace the art of the frugal wow — creating reciprocity through small, thoughtful gestures. In fact, psychologist Norbert Schwarz found that as little as 10 cents can create reciprocity between two individuals (it really is the thought that counts).

One small way we try to consistently brighten our customers’ day is by building what we call “huzzah” images into our product — a fun illustration and caption that appears when you reach inbox zero.

@helpscout I like the encouraging vibes when getting through the tickets! pic.twitter.com/gpEOgdTPCv
— Rosemary Lynch (@rosemary) November 30, 2016

.@helpscout You get me <3 pic.twitter.com/eemgbm2GQ0
— Tamikya Van Der Wel (@Takiah) November 24, 2016

It may seem like a small thing, but if you can make people feel good about using your product, they’ll be more likely to stick around.

10. Make use of surprise reciprocity

Although reciprocity works incredibly well on its own, research shows it’s far more powerful when started by surprise. For a simple example, recall a time that someone did something nice for you unexpectedly — the gesture probably wasn’t all that unusual, but the fact that it came out of nowhere left a strong impression on you.

11. Don’t underestimate the power of ‘thank you’

Companies like Nordstrom, Zappos and Wufoo are known for thanking their customers via handwritten snail mail. They understand that old-school, handwritten thank-you notes are a low-cost, high-impact customer retention strategy.

Thank-you notes are still a rare throwback to old-fashioned, personal customer service — they stand out as a delightful gesture that makes customers feel special and cared for. The relationship-building is well worth the relatively minimal investment.

12. Make it personal

In a study from the Journal of Applied Social Psychology, researchers found that waiters and waitresses could increase their tips by 23 percent by the simple act of returning to tables with a second set of mints. The researchers concluded that the mints created the feeling of a personalized experience for the customers who received them. So, it was the personalized service that made their day, not the small gift in itself.

13. Reduce effort before delivering delight

Many companies assume “exceptional” customer service can only be achieved by going above-and-beyond; that loyalty is built on showy gestures. According to research from Dixon, Toman, and DeLisi published in The Effortless Experience, the true driver of customer retention and loyalty is the ease of getting a problem solved — delight isn’t the foundation of a customer service strategy, but rather a second-order effect. First, focus on consistently meeting expectations and avoiding unpleasant surprises. Then go the extra mile.

How great service retains customers

You can’t build customer loyalty without an exceptional customer service to keep people coming back. Let’s debunk a few customer service myths, as well as tackle some important things you need to keep in mind when offering support online.

14. Accept that speed is secondary to quality

When it comes to highly rated customer service, data show that quality and completeness matter more than speed. According to research from the Gallup Group, customers were nine times more likely to be engaged with a brand when they evaluated the service as “courteous, willing, and helpful,” versus the “speedy” evaluation, which only made customers six times more likely to be satisfied.

15. Know your customers

Telling your team to spend more time with customers might seem like folly, but smart entrepreneurs know that isn’t the case. Numerous behavioral psychology studies have shown that everybody views their service experience as more positive when they don’t feel rushed or ignored. Don’t spend time idly, though; have employees attempt to find out key customer traits, just like Derek Sivers did at CDBaby.

16. Choose the right platform

The best way to improve your customer service efforts is to utilize the channel your customers most prefer. Research has shown the death of email support has been greatly exaggerated. However, you need to pick the channel that makes the most sense for your business. Hosting companies, for example, know that live chats are critical when their customers’ sites go down; other companies may have customers who prefer using self-service, or even phone support.

17. Make it a communal effort

Countless case studies have made one thing clear when it comes to creating an efficient support system: You need to keep everybody in the loop. At Help Scout, we use our integration with Slack to access real-time notifications of what’s happening on the customer end; we were able to improve our response time by 340% (not a typo!) by creating a support channel for all of our teammates.

18. Solve cause and effect

It isn’t enough for your customer service team to apologize; their main goal should be to solve for the immediate problem, but to also find and flag the root cause. In doing so, they can solve a systemic issue and help other customers avoid it altogether. Reducing problems is key — in fact, research conducted by John Goodman found that customers were much more sensitive to price changes, and thus more likely to churn, when they experienced a few problems with the product or the support they received.

Source: Strategic Customer Service

Designing sticky retention programs

The key to creating loyalty programs that work is to know why customers use them and what gets customers to keep using them. Below you’ll find consumer research that answers these questions.

19. Give loyal customers a head start

Consumer researchers Joseph Nunes and Xavier Dreze are known for their studies on the Endowed Progress Effect. Their results have conclusively shown that the biggest pitfall in preventing customer loyalty programs from succeeding is getting people started. In their well-known car wash study, participants were twice as likely to finish loyalty cards when they were automatically started (or rewarded) as soon as they signed up.

20. Make ideal customers VIPs

Additional research by Dr. Nunes on retention programs has shown that people love being VIP, or “gold” members. There is one caveat, though — this only works when people know there is a class below them on the totem pole. Speaking to human nature, Nunes saw a notable increase in gold members’ participation as soon as he implemented a silver class.

21. Reward your loyal advocates

When your customers go out of their way to recommend your product or service to others, let them know that you see and appreciate it! If you spot someone recommending your business on social media, for example, reply to say thanks — it shows you’re paying attention, and that their testimony means something to you.

Another way to recognize your brand advocates is with a referral program — Help Scout is currently giving existing customers a $100 Amazon gift card when they refer someone who becomes a customer (and that new customer gets a $50 credit on their account).

22. Assign your customers positive labels

Research on voting patterns conducted by Stanford University revealed people are more likely to participate in something if they are labeled with a positive trait. Buffer refer to their premium customers as “awesome” members, and even named their upgraded payment plan the Awesome Plan — a much easier phrase to embrace than “paid member.”

Retaining customers is a balancing act

There are many tactics, but no shortcuts; you can’t “hack” a personal relationship, so why should we assume business relationships are any different? Truth is, the tactics above should hopefully give you some fresh ideas for approaching retention, but they’re not a cure-all.

Your product and service will do most of the heavy lifting in keeping customers loyal, and there are no shortcuts for that.

Criteo vs Adroll for B2B

Wed, 06 Sep 2017 00:00:00 +0000

At Help Scout, we ran a test of Criteo vs. Adroll for remarketing. Results showed that Adroll was the winner. The final results were based on Revenue.

Background

In the beginning, I let AdRoll and Criteo know that we would be testing them against each other in August. I’d take the monthly budget and split it evenly between the two channels.

Criteo agreed and AdRoll wished us the best of luck. I also shared reporting between the two whenever they asked. They were being judged on 4 aspects (which they both knew about).

Account Management

Winner: Both

To be honest, I was surprised Criteo worked with us as much as they did. In the initial sales call, the AM called us an “unusual client”. Chris (Criteo) and Casey (AdRoll) were both great, but there were a few differences.

Casey is our dedicated Account Rep. Chris was our Sales Rep and held my hand until the campaign started running, and then handed me off to Carly (our next account rep). Chris was great, but because of his short term with us, he really couldn’t see long term.

It was a tie because initial set up with Adroll was rough. Although Casey is great, he’s our 3rd account rep.

Why the emphasis on the account rep? AdRoll has pushed us into a few Alpha programs due to account management; they can also manipulate our bids.

Campaign Performance

Winner: AdRoll

Criteo sold us (and every single one of their customers) on click through conversions. They do that because AdRoll’s bidding model is based on CPMs, while Criteo’s is based on CPC.

This was one of the biggest reasons I wanted to test. If Criteo could bring half price clicks, it’s worth the test to push on conversions.

And that was the case. Criteo CPC was $3.74, while AdRoll was at $7.28. The Adroll CPC was 2x that of Criteo, but AdRoll converted better, and I give credit to AdRoll’s technology being better.

Criteo is built for ecommerce, b2c. Leave the cart with an item? Criteo will hammer the user until they buy again.

AdRoll, not so much. AdRoll plays a longer B2B game – which fits our customers better, but this also means we will rely on VTC conversions.

AdRoll Blended: $99.86

Criteo Blended: $124.53

Revenue

Winner: AdRoll.

Due to the higher CPA, Criteo needed to have revenue at ~130% of that of AdRoll to make it close. And it didn’t happen.

Technology

Winner: AdRoll

AdRoll focuses on longer buying processes; Criteo focuses on immediate return - not ideal for our customers. Again, worth testing due to the half price CPC, and it did convert, but $20 higher CPA.
Criteo runs everything through the account reps: Bid Change? Upload Ads? All have to go through Criteo Reps. AdRoll allows me to make any and all changes.
Criteo doesn’t offer an exclusion audience: no way to block customers. And if you’re B2C, this isn’t a necessity. But we need it.
AdRoll has 1 pixel to rull them all. Criteo has to create a unique pixel for each campaign.

Continuing with Adroll

For us, AdRoll was clearly winning on CPA and Technology.

SEO Audit of the Kissmetrics Blog: Google Panda Recovery

Thu, 07 Aug 2014 00:00:00 +0000

Have you ever noticed how much advice about SEO and Social Sharing is completely wasted? It’s amazing to me how many people can write about strategies, best practices, and go on preaching without showing exactly how it’s done.

Eli Overbey has provided an incredible post about the website KISSmetrics. In fact, if you’ve ever wondered how to do an audit, detailed analysis, or strategy document…wonder no more.

It’s all in the nearly 9,000 word document below.

This post was original written for OkDork (Noah Kagan). You can find it here: https://okdork.com/a-complete-audit-of-kissmetrics-seo-and-inbound-marketing/

In this post, I will provide a comprehensive technical Inbound audit for KISSmetrics. This example is extremely detailed, and it covers a wide range of inbound topics. But before the inbound marketing audit, let’s cover the basics…

Who is KISSmetrics?

KISSmetrics is a web analytics solution that helps companies make smarter business decisions, and boosting ROI.

Headquartered in San Francisco, KISSmetrics is backed by a syndicate of angels and early stage funds.

Why Audit KISSmetrics?

Like everyone working in SEO and Inbound Marketing, we all overlook or miss things (even basic things). The goal of this post is to help KISSmetrics, to help others learn from their inbound successes (they’ve done remarkable things in this arena), and to give a third-party perspective on a great example.

To be fair, “audit” seems like a terrible word, but in this case, think of “audit” as “being helpful.” The purpose of this audit is to give an outsider’s view to a great company. All in all, I hope this helps KISSmetrics. Neil and team are outstanding.

Disclaimer

Before I jump into the technical details of the inbound audit, it’s important to note that I have no affiliation with KISSmetrics (I am also not currently a user of their product - so you don’t have to worry about affiliate links). KISSmetrics also did not ask me to complete this audit. As a result, I don’t have access to any of the site’s analytics or webmaster tools accounts, and I don’t have access to the site’s content management system (CMS).

In a typical audit, I would begin by sifting through the data and then narrowing in on problem issues. So… if I make completely inaccurate observations, I blame my data from the third party tools, ie (Searchmetrics, Ahrefs, SEMrush, Moz, etc…). Don’t get me wrong - these tools are awesome, but they generally work better from the inside of the company.

The goal of this audit is to help KISSmetrics. The aim is never to critique in a negative, harmful way, but to help promote KISSmetrics through inbound marketing by giving them the perspective of an objective third party. With those disclaimers out of the way, let’s begin.

*If you have any questions: @elioverbey

Indexability / KISSmetrics Web Structure

In order to understand this audit, you need to understand how KISSmetrics is set up. At first glance, you’ll notice that KISSmetrics is constructed using quite a few subdomains:

kissmetrics.com/ - 18 pages

○ *tweets.kissmetrics.com - 2587 Pages

○ blog.kissmetrics.com - 1,038 pages

○ *status.kissmetrics.com - 70 pages

○ grow.kissmetrics.com - 7 pages

○ focus.kissmetrics.com - 10 pages

○ support.kissmetrics.com - 347 pages

○ middleman.kissmetrics.com - 1 pages

○ styleguide.kissmetrics.com - 7 pages

○ uptime.kissmetrics.com - Redirects to status.kissmetrics.com

○ demo.kissmetrics.com - 100 pages

An * (asterisk) indicates that the subdomain is blocked by the robots.txt.

In total, KISSmetrics has 4,185 pages across 11 domains and subdomains.

As you glance at the site structure of KISSmetrics, you notice that their content is divided up into subdomains. Due to the structure of KISSmetrics, this inbound audit will focus solely on two areas of their site: the root domain (kissmetrics.com) and their blog (blog.kissmetrics.com). The other subdomains do not contain information pertaining to this inbound audit (ie. styleguide) and will not be included in the analysis.

Inbound Marketing

In its most basic form:

“Inbound marketing focuses on earning, not buying, a person’s attention.”– Brian Halligan

The principle of inbound marketing has been around for years, but the term has been coined by HubSpot (well done Dharmesh and Brian). Inbound Marketing is all about creating great content that pulls people in.

Since HubSpot is the go-to on inbound, this audit will use their structure as a model. As this audit continues, you’ll notice how the funnel naturally works.

Optimization / Attract

The first step in inbound marketing / SEO is attracting the right audience. Content is the single best way to attract new visitors to your website. In order to be found by the right prospective customers, KISSmetrics must not only create optimized content, but foster an environment in which content can thrive.

Search Visibility

In order to set the framework for the entire audit, it’s important to analyze the site’s performance. When looking at KISSmetrics’s traffic, there are 3 important questions:

Is the traffic growing?
Has the site been penalized?
What is the percentage of organic traffic?

To determine KISSmetrics’s traffic performance, I used two tools:

Searchmetrics Suit, and one of my new favorites, Similar Web. Finding accurate third party data is incredibly difficult, but after using Google Analytics on a network of sites, I’ve found that these two programs provide the most accuracy in capturing third-party data.

As you can see from the graph below, there are some red flags for KISSmetrics (the tools join all traffic from all subdomains together):

As the graph above shows, the site’s visibility decreased dramatically after May 18, 2014. This date is important because it also corresponds to when Google updated Panda 4.0.

Immediately, there is cause for concern on KISSmetrics’s site. As you begin to look over the third-party data, there is a strong correlation between the data (although third party) and Google’s Panda update…

There are three possible solutions to this drop in traffic:

The data is inaccurate
The site was algorithmically penalized due to duplicate content
The site was algorithmically penalized due to guest bloggers

The first possibility: The data from the third party is inaccurate. In fact, it seems highly improbable that KISSmetrics was hit with an algorithmic penalty, right? Neil Patel (co-founder of KISSmetrics and SEO expert) wrote an article for Search Engine Journal about the Panda 4.0 update the day after Panda 4.0 was released.

In the post, Neil describes the update, who was affected, and how to fix it. Does it not seem like KISSmetrics (his company) would be prepared for Panda? Since he knew so much about Panda, he would have protected his assets.

Besides that, KISSmetrics writes lengthy (2,000+ word average), unique content that is well received (only 12 of their articles don’t have any comments - meaning, the other 1,000+ have engagement via comments).

That is one possibility. But the other possibilities do cause for more concern:

The second possibility: KISSmetrics was hit by Panda 4.0 and here is the probable reason:

For some odd reason, the KISSmetrics blog exist on two protocols: http and https. Meaning, the same content (although unique and engaging) can technically be seen as duplicate.

Let’s look at an example. As of today, the most recent post from KISSmetrics is: Website Testing Mistakes That Can Damage Your Business (this post was chosen due to recency, but this same problem exists throughout). You can find this article in two places:

http://blog.kissmetrics.com/website-testing-mistakes/
https://blog.kissmetrics.com/website-testing-mistakes/

As you can see from the image below, the same page exists in two places (unfortunately, this is duplicate content and cause for concern):

You’ll notice that in the image above, on the left is the normal protocol (http://), and on the right is the secure protocol (https://). Google drops the http:// protocol in the browser on the left by default.

I also checked out the source code on this same page to find three problems that could have led to this drop in traffic on May 19th (Panda 4.0), I highlighted 3 key problems in the diagram below:

Once again, I highlighted the protocol so you can see the two exact pages in two locations.
Both of these pages are INDEX, FOLLOW. If one of these pages was NOINDEX, there wouldn’t be a huge problem, but KISSmetrics is telling the search engines to index both of these pages. This is another red flag.
The third red flag is the canonical. The canonical tag tells the search engines that a given page should be treated as though it were a copy of the URL. In this case, each page canonicals itself. One page says the content should canonical to http:// and the other page says the content should canonical to https://. Basically, each page is claiming original.

Finally, Google has noticed that the blog exists in two locations. As you can see below, a query for the blog brings back both protocols (http and https).

This means that the KISSmetrics blog was probably hit by an algorithmic penalty on May 19th.

Even if the third party data is wrong (and KISSmetrics was not hit by a penalty) they should fix these problems immediately. KISSmetrics should decide which URL is primary, and then 301 or canonical.

Even more, as it currently stands, the site is dispersing its links among https and http. If they fix this problem, the could see a rise in rankings due to the migration of their links.

The third possibility: KISSmetrics was hit by Panda 4.0 because of guest bloggers.

I am not implying that KISSmetrics was penalized because “they didn’t stick a fork in their guest bloggers”. But, KISSmetrics could have been penalized by duplicate content from their guest bloggers (quite a few of KISSmetrics posts are written by guest bloggers). Guest writers could have written for KISSmetrics and then re-published that content on their own blog.

Based on the KISSmetrics publishing guidelines, there are no guidelines stating that the writer cannot re-publish his or her own content. Although the duplicate content issue could be covered in other forms of communication to the writers (contract, etc), it is a possibility.

Penalty Conclusion

Although I am hoping for the first possibility, (I wouldn’t wish a penalty upon anyone) these are issues that KISSmetrics should thoughtfully consider. Again, I cannot stress the unreliability of third party data, but in any manner, these are issues that should be addressed immediately. Hopefully, Neil can chime in on this issue. I’d really love to know if these have been issues for KISSmetrics (plus, Neil is extremely transparent in order to be helpful).

Robots

A robots.txt file is used to restrict search engines from accessing specific sections of a site. Here is a copy of KISSmetric’s root domain Robots.txt file:

On their blog, KISSmetrics uses another robots.txt file:

There are a few improvements that KISSmetrics could make in the construction of their robots.txt.

First, on the blog, KISSmetrics should consider blocking ‘/wp-content/plugins/’. Sometimes developers put links in their plugins.

Second, on the blog, KISSmetrics should rethink blocking /wp-includes/’ in robots. There are better solutions for blocking robots than in the robots.txt file (ie, NOINDEX).

Finally, KISSmetrics should add a working sitemap to their blog. The current sitemap (Sitemap: http://blog.kissmetrics.com/sitemap.xml.gz) does not work.

Fixing this will help the robots to index their pages.

Robots Meta Tag

Each page on a site can use a robots meta tag to tell search engine crawlers if they are allowed to index that page and follow its links.

WordPress is a great open source platform – one that KISSmetrics built their blog on – but duplicate content is one thing you have to be very mindful of. Using the robots meta tag will help prevent duplicate content. Content duplication issues include tags, categories, and archives.

Quite a few of KISSmetrics’s pages have a meta robots tag. In the case mentioned above (wp-includes), it would benefit KISSmetrics to include a “noindex” robots meta tag on a per page basis, rather than blocking an entire directory.

One thing KISSmetrics should consider ‘NOINDEX’ is blog subpages. As you can see below, the sub-pages are indexed:

Since these page take up crawl bandwidth, but don’t have unique content, and historically have a low CTR, KISSmetrics (based on their analytics) should consider “NOINDEX, FOLLOW” on these archive based pages.

KISSmetrics has successfully implemented this change on subpages of categories, topics, and topic sub-pages, and should consider adding the same meta robots tag to the other archive based structures. The “NOINDEX, FOLLOW” will remove the subpages from the index, but will still allow links to pass.

Accessibility

This section covers best practices for both search engines and users. Many of the search engines’ accessibility issues were mentioned above in index-ability, and now we’ll cover accessibility as it mainly relates to personas with the benefits of robots in mind.

Performance

According to Google Webmaster tools:

You may have heard that here at Google we’re obsessed with speed, in our products and on the web. As part of that effort, today we’re including a new signal in our search ranking algorithms: site speed. Site speed reflects how quickly a website responds to web requests… While site speed is a new signal, it doesn’t carry as much weight as the relevance of a page.

It’s tempting to dismiss site speed as an important SEO ranking factor, but if Google says it matters (even a small percentage), then it matters. Even if you dismiss speed as an optimization factor, it is also an inbound factor that can’t be ignored. Who likes a slow site?

KISSmetrics’s site speed is fairly slow, and KISSmetrics could easily make a few tweaks to make their site much more efficient. As you can see below, the site speed for the root domain is:

The problems on the root domain above are quick fixes. The blog (blog.kissmetrics.com) did much better, scoring an 88% and a 79%.

Reducing the number of files needed to load the site, and thereby reducing the number of HTTP requests, will make KISSmetrics’s site load more quickly. Currently, the root domain makes 64 request when loading the page (which is somewhat surprising, considering the homepage is only a few images and includes less than 70 words.)

There are usually three parts to fixing this:

Reduce the number of JavaScript files
Reduce the number of CSS files
Reduce the number of images

On the homepage, the Time to First Byte is efficient (300ms), but the page load is somewhat slow – anywhere from 2.95 seconds on GTMetrix to 1.95 seconds on a tool like Pingdom. They received a 55/100 on Page Speed Insights via Google (42 on Mobile), and 75/100 on Pingdom. KISSmetrics’s homepage has room for improvement.

Another concern is that their blogs have over 100+ https requests, but surprisingly, even with all the requests, the page speed was excellent: 933kb.

The site could be improved with the following:

Eliminate render-blocking JavaScript and CSS
Minimize HTTP requests – KISSmetrics’s pages will load more quickly with fewer requests. Minimizing these requests involves reducing the number of files that have to be loaded, such as Javascript, CSS, and images.
Combine Javascript and CSS into external files with links from the header. This allows the external page to be cached so that it will load faster (there are 10 CSS files and 4 JS files loaded separately on the blog).
Implement server side / browser caching – This creates a static html page for a URL so that the dynamic sites don’t have to reload / be recreated each time the URL is requested.
Load Javascript asynchronously.
Use a CDN – such as Amazon. The CDN will allow users to download information more quickly (as far as I can tell, they are not using a CDN).
Finally, use 301′s only when necessary. A 301 forces a new URL, which takes a longer time to load (93 of their pages use 301).

Site Architecture

The site architecture defines the overall structure of a site, and it has a number of important SEO implications. For example, when a page receives external authority, the site architecture defines how that authority flows through the rest of the site.

Additionally, since search engine crawlers have a finite crawl budget for every site, the site architecture ultimately dictates how frequently pages are crawled (or if they’re crawled at all).

Authority Flow

To understand how authority flows through the site, I performed an analysis on the site’s internal links.

Based on that analysis, here is the distribution of the site’s links (these values have been rounded to the nearest percentage):

As you can see above, 96% of the pages have a value less than 0.1 (pages with authority values closer to 0 have the least authority and pages closer to 1 have highest authority), relative to the other pages. Most of the site’s internal authority is held by 3.8% of the pages.

The root cause for this authority is the site’s navigation. All of the pages that have large amounts of internal links (1800+), are found in the footer or header:

/infographics (1834)

/marketing-guides (1830)

/webinars (1822)

/topics (1820)

Since these links appear sitewide (i.e., on every page), these page receive multiple internal links, while the other pages receive very few internal links.

Even though the Article Categories on the site break down the articles into substantial sections, the internal links do not seem to pass through to individual articles (which is the case with almost every navigation).

The related post section on each article works well to accomplish this purpose (an example is below), but not every post includes the related post widget.

Since KISSmetrics related posts seem to be manually picked, often times, the most recent posts are picked (leaving the older post with fewer internal links). KISSmetrics should continue cross-page linking so that the internal links distribute across other pages.

Click Depth

Click Depth (or crawl depth) is the number of times a website visitor must click on a link from the root domain in order to get to the desired page. The root domain will have a link depth of zero – since it takes zero clicks to get to the root domain on a webpage.

Search Engines are like humans – they do not want to waste time trying to find pages hidden deep in a site’s architecture. Pages that are available in one click – thus one page from the homepage – are deemed more important by search engines than those that are several clicks removed.

It doesn’t matter if the URL is KISSmetrics.com/name if it takes Googlebot (and users) 8 clicks to get there.

As the graph below shows, 44% of the site’s pages have a click depth of 5 or greater (i.e., they are 5 or more clicks away from the homepage):

To reduce the click depth, KISSmetrics can create new ways to interlink the site’s pages. To accomplish this, KISSmetrics could:

Increase the blogroll count to reduce the number of paginated pages (currently, they have 90+) [Revealing more results by scrolling (or swiping) is an established convention (Kohn).]
Link back to the top results from each of the paginated URLs

A flat architecture decreases the distance between high authority pages and all other pages, increasing the chance that low PR pages will be crawled on a more frequent basis.

On Page Factors

In this section, I will investigate the characteristics of KISSmetrics’s pages (e.g., URLs, schema, duplicate content, etc.) that influence the site’s search engine rankings. On Page search factors are those that are entirely within KISSmetrics’s own control.

HTML Markup

A site’s HTML is important because it contains some of the most vital on-page ranking factors.

HTML markup is a machine readable language that is used to tell the browser how to display the text or graphics in the document. Search engines crawl over your pages and semantic markup can identify your pages’ most important information. Semantic HTML markup is a way of indicating the meaning of web content to the search engine.

When examining KISSmetrics’s pages, I found multiple pages with 5-15 errors per page.

Many of the issues are easy fixes: open tags, stray elements, attribute errors, etc… Although these aren’t extremely critical issues, it’s best to clean up as many errors as possible (empty headings, duplicate IDs, attributes, etc). To be fair, most of the errors are output from the plugins that are installed on the blog.

Structured Data

Schema.org mircodata is an extra set of HTML tags that you can add to your HTML elements to let the search engines know specifically what you are talking about. For example, for the KISSmetric blog, they would tag the title as a “headline,” the author as an “author” (with authorship), and the date the blog post was published as the “publishedDate.”

Now when Google indexes KISSmetrics’s blog, it knows specifically what the title of an article is, who the author is, and when the post was published.

Fortunately, KISSmetrics does define structured data on most of their posts (authorship included):

As of late June 2014, Google’s John Mueller announced that Google was making a major change with authorship - Specifically - dropping the profile photo and circle count from the search listings (supposedly “click-through behavior on this new less-cluttered design is similar to the previous one”). For some reason, I highly doubt this, but KISSmetrics can be the judge.

To make the remaining articles more readable for search engines, I’d recommend adding authorship for guest bloggers.

Be Careful with Schema.org

One caution: I do not recommend schema.org for all HTML elements. For example, consider the search term “how do you boil eggs”:

Here, you can see that Google is using micro markup (schema) to scrape the content and place it on Google – this could potentially hurt click through rates (CTR). The more information you tell Google – the more you risk Google scraping your content and your site losing sessions.

Head Tags

rel=publisher

KISSmetrics has a Google+ page, and they are taking advantage of rel=publisher. Publisher connects the Google+ page to the website, and that results in your Google+ page being displayed in search results whenever anyone is searching for your brand name.

KISSmetrics has rel=publisher markup code on their homepage. If you look at KISSmetrics’s results in the SERPs, you will notice they are receiving the markup:

KISSmetrics could, however, update their Google+ profile more often. They currently only update their profile a few times a month. Updating the post more regularly could result in a posts being shown in the results.

rel=prev / next

As previously mentioned, KISSmetrics could potentially have a duplicate content issue. In addition to addressing the issues above, these tags are influential in indexing and preventing duplicate content.

Google asks that Webmasters add rel=”next” and rel=”prev” to paginated archives, so that Google can distinguish them as a series and send users to the most relevant pages. KISSmetrics is doing a great job of using prev/next on individual pages, but they are not currently using prev/next on subpages of archives (page/1/ or sub-pages of /topics/).

They do, however, seem to have an issue using nofollow on pagination. According to Rand, whatever you do, do NOT:

Add nofollow to the paginated links on the results pages. This tells the engines not to flow link juice/votes/authority down into the results pages that desperately need those votes to help them get indexed and pass value to the deeper pages.

Unless KISSmetrics has a valid reason unbeknownst to a third-party, they should consider removing the nofollow links on the blog pagination.

rel=canonical

Each KISSmetric post uses a rel=”canonical” link; however, all of these links are self-referential and normally are not a problem. The problem arises, however, when both the https and http version are self-referential.

On subpages (/page/2/, etc…), KISSmetrics does not use canonical tags, but on many pages, KISSmetrics uses rel=prev, rel=next, which is actually a common substitute for canonical (according to webmaster tools). Since not all subpages include canonical or rel=”prev/next”, I’d suggest adding those tags to help the search engines distinguish the paginated series. Since KISSmetrics has ‘NOINDEX’ on topic subpages, this isn’t a problem, but they should consider addressing canonical (or prev/next) on the blogroll, as it is ‘INDEX,FOLLOW’.

NOINDEX

The last tag KISSmetrics should think about implementing (after checking analytics) is NOINDEX,FOLLOW. They’ve successfully implemented NOINDEX on the /topics/ but have not used it on subpages on the blog roll. Unless their data says otherwise, they should consider NOINDEX,FOLLOW on these pages.

Open Graph

The Open Graph protocol enables any web page on KISSmetrics’s site – or all sites – to become a rich object. For instance, the Open Graph is used by Facebook to allow any web page to have the same functionality as any other object on Facebook.

This is the open graph information in their blog (blog.kissmetrics.com) markup [the root domain does not include open graph information]:

The above markup translates to:

On their blog, they are correctly implementing open graph and might want to include the tag below:

KISSmetrics may also want to consider adding publisher to the blog as well (and not just the root domain) and verify they have authorship implemented throughout (these tags can also display on open graph - and they currently do not).

Twitter Cards

Twitter cards are another micro markup – essentially rich snippets for Tweets. Twitter will actually fall back on the Open Graph tags used for FB, but the Open Graph tags aren’t comprehensive.

All of KISSmetrics’s blog pages include og: tags, which attach additional information to the Tweets associated with those pages by default. However, they should also include (in addition to the og: tags):

Titles

Each page that is indexed should have unique content and a unique title that effectively summarizes the content for users and search engines.

For titles, Google will show as many characters that a 512 pixel display can show – thus, around 50-60 characters. The reason the number varies is because some characters are bigger than others (ie, you could fit a lot more “i”s than you could “w”s).

Out of the 1000+ pages on KISSmetrics’s website, 393 pages have a width greater than 512 pixels. As you can see below, most of KISSmetrics title’s fall in and around the 512 mark, but about ⅓ exceed:

KISSmetrics’s titles average 51 characters, but since the pixels of various characters can exceed 512 pixels, some of the title’s are truncated in the SERPs:

*In some cases, Google no longer truncates the title and adds an ellipsis to the end; instead, Google tries to algorithmically determine a better title for the post (Google changes the title to better match the query. The reason for this is simple: users scan for and assign higher relevance to titles that include their query).

Just out of curiosity, though, let’s look at the overall distribution of title lengths that were not cut off: (pre-cut-off):

The good news is that this distribution is roughly normal, peaking at about 54-58 characters. Post-cut-off title tags ranged in length from 55 to 68+ characters. Here’s a title cut off at 55 characters (measured at 514 pixels):

The example above is the only title less than 58 characters that was truncated. In this case, keeping titles within 50-58 characters is the wisest option. It’s not worth going back and revising every title based on this new data. KISSmetrics should look at their key pages, view the SERPs, and make sure the snippets show correctly.

KISSmetrics should keep these guidelines in mind for future SEO efforts, but not worry too much.

Lastly, less than 5% of KISSmetrics’s titles contain their brand name. To help promote brand awareness, I’d add suggest adding KISSmetrics at the end of each title (

KISSmetrics). Because KISSmetrics is well recognized, including the brand name in the title element could increase search engine clicks.

Meta Descriptions

Meta descriptions are HTML attributes that provide concise explanations of the contents of web pages. Meta descriptions are commonly used on search engine result pages (SERPs) to display preview snippets for a given page (Moz).

Meta Descriptions do not influence the ranking factor of a site’s page, but they do influence the CTR (click-through-rate).

It’s speculated that Meta descriptions are calculated by pixel width, rather than characters. This pixel width is less than simply 512 pixels multiplied by two (1,024px) which you might expect, and actually the CSS truncation appears to be around 920 pixels. Google uses 13px Arial for meta description text and chop off at word boundaries.

Out of 1,056 unique HTML/Text URLs on the root domain and blog, 97% include a meta description. Over 51% of the meta descriptions are over 920 pixels. As you can see below, the meta description is being truncated:

For every page on the site, the meta description is an excerpt of the first sentence. KISSmetrics should consider writing a succinct, unique meta description (no more than 155 characters) in order to help CTR. Because their meta descriptions are excerpts, many of them are cut off:

If KISSmetrics does not decide to write a unique meta description for each page, they should consider removing the automatic excerpt on the long-tail articles. If the page is targeting long-tail searches, it may be easier to let the search engines pull in the meta descriptions themselves.

The reason this is more feasible than automatically outputting an excerpt is because when the search engines pull content, they pull content and display keywords surrounding the user’s query. If you force a meta description, it can detract from the relevance of long tail search.

In some cases, search engines will overrule your meta description anyways, but you cannot always rely on that. In that regard, it would be advantageous for KISSmetrics to consider adding unique meta descriptions for pages that target only a few keywords (to see if they can increase CTR) and leave the other long-tail pages open for variance.

Since KISSmetrics posts average 2,400 words, they should consider A) writing unique descriptions for the posts, or B) allowing the SERPs to automatically create the description based on the user’s query.

Images

For optimization purposes, the two most important image attributes are the image’s filename and the image’s alt text. Both of these attributes should effectively describe the image’s content to provide additional context for search engines and visually impaired users.

Image file names should be descriptive – not containing random numbers or queries. The file name should describe the image. The file name is information you give directly to the search engines and to other technologies to identify what the information is about.

For the benefit of search engines, compliance (HTML), and the visually impaired, all images on KISSmetrics should also have an ALT tag (Alternative text is text you provide for an image in case it can’t be displayed - perhaps the image is broken or the program cannot display the image). The ALT tag should accurately describe the image, and if possible, contain a keyword relevant to KISSmetrics (but only if the keyword is relevant to the image).

One other form of optimization that is not usually included is the text around an image. The text around an image tells a lot about the image itself. This is just another signal telling the search engine what the image is.

If KISSmetrics ensures that images follow these rules, they will increase the likelihood of referral traffic from image searches. KISSmetrics has hundreds of images, and around 81.26% of the images take advantage of the alt text (there are 2,089 images without alt text).

One area that KISSmetrics can improve in is the descriptiveness of the file naming. As you can see below, quite a few of the images are just random strings.

There are hundreds of images like the ones above on KISSmetrics’s blog. I’d suggest adding descriptive title tags to the images to increase organic image search, and also to be HTML compliant.

URLs

The site’s URL structure is streamlined, and KISSmetrics has done a great job crafting their URLs. Their URLs are at an average length of 50 characters (including http://www.kissmetrics).

KISSmetrics could, however, improve some of their URL structures by making them slightly more descriptive. Take these URLs, for example

6 Mistakes That Will Kill Your Product Launch

9 Metrics to Help You Make Wise Decisions About Your Start-Up

How to Make a Landing Page That C.O.N.V.E.R.T.S.

KISSmetrics could add one or two more words to the URL to describe what the page is about. The keywords in the URLs above are very vague, and KISSmetrics could potentially help ranking and click through by describing the URL better. For example, an article titled, “Truth Will Out – Why Authenticity is the Key to Growing Your Business” is found here:

http://blog.kissmetrics.com/truth-will-out/ (the only descriptive keyword is “truth”).

Updating these URLs to make them more descriptive (and more closely aligned with their corresponding content) may help in their organic search.

Overall, the site’s URL structure is organized and clean. Unfortunately, while crawling the site, I found URLs on 738 pages that returned a 403 or 404 HTTP status codes (i.e., those URLs are no longer accessible). In one post, 50 Resources for Facebook Application Developers, KISSmetrics states, “Note this post is from 2008 so many of the links are broken.” If possible, those links should be redirected and updated.

Another important URL-related consideration is the domain name. Although the domain is not expiring anytime soon, KISSmetrics should keep a watch on the domain’s expiration: September of 2015.

External Links

One last item for KISSmetrics to consider is the amount of links they place on each page. According to Matt Cutts (on the topic of links per page):

So how might Google treat pages with well over a hundred links? If you end up with hundreds of links on a page, Google might choose not to follow or to index all those links. At any rate, you’re dividing the PageRank of that page between hundreds of links, so each link is only going to pass along a minuscule amount of PageRank anyway. Users often dislike link-heavy pages too, so before you go overboard putting a ton of links on a page, ask yourself what the purpose of the page is and whether it works well for the user experience.

KISSmetrics has 61 pages on their blog that exceed 100 links and 4 with 200 external links or more (“28 Things You Need To Know About The New Facebook Pages” ranks the highest with 372 outlinks). You can see an example of that post below:

Simply put, more links equals less PR for those links. The actual math of PR gets fuzzy fast, but for best measures, KISSmetrics should consider keeping their links under 100 per page (even if for user as experience).

Off - Page Factors

This section will cover some of the most influential factors in ranking: backlinks and social engagement.

Backlinks

When analyzing backlinks, you should consider: Total number of links, link quality, number of unique domains, fresh/incoming links, and anchor text usage.

To start, I first recorded how many backlinks KISSmetrics had received over the past few months. The charts below show the growth of KISSmetrics’s backlink profile over the course of one year. Below is a chart of blog.kissmetrics.com:

After comparing the backlink analysis above to both Majestic SEO and Moz, there was not much of a difference. KISSmetrics was consistently growing their unique domain backlinks until a steep drop in March, and has started a recent growth once more.

Both the root domain and the blog subdomain have the same drop, and combined they have around 16,000 unique domain backlinks.

To see how KISSmetrics’s backlink growth matches up over the years, I compared their root domain growth with their blog domain growth:

As you can see above, the blog backlinks are growing at a much faster pace than the root domain backlinks (and they should be, since new content is created daily on the blog).

Since KISSmetrics’s blog is outpacing the root domain in backlinks, should they move the blog to a subdirectory?

The Case for Subdirectories (versus Subdomains)

Matt Cutts has said:

[Subdomains and Subdirectories] are roughly equivalent. I would basically go with whichever one is easier for you in terms of configuration, your CMSs, [and] all that sort of stuff.

Although it seems like subdomains and subdirectories can be treated equally, Cutts uses very vague language; “roughly equivalent” is still not equivalent.

KISSmetrics should consider testing out subdirectories over subdomains for 2 main reasons:

First, there is a correlation between pages on the root domain and higher rankings. According to Rand at Moz:

I would still strongly urge folks to keep all content on a single subdomain. We recently were able to test this using a subdomain on Moz itself (when moving our beginner’s guide to SEO from guides.moz.com to the current URL http://moz.com/beginners-guide-to-seo). The results were astounding – rankings rose dramatically across the board for every keyword we tracked to the pages.

Second, fewer subdirectories tend to create a better user experience and flow. As it stands, KISSmetrics exist in silos. Although the subdomains are clean, they do not make for the best of navigation.

KISSmetrics should also consider how their links pass from root domain to subdomain. Since their root domain has over 3K links, and the other subdomain (the blog) has over 13k, KISSmetrics should consider link equity. With only 16 pages on the root domain – and a lot of that being less viewed, weaker content - they should at least think about the potential of combining the subdomain and root (with 301, of course).

The Case for Subdomains (versus Subdirectories)

There is also the stance that subdomains work equally as well. Some of the benefits of using a subdomain are:

Links from the root domain (since subdomains are treated as a separate site)
Get links “from another website” (subdomains are treated pretty much as such). Ability to link back to priority pages on the root domain.
Piggyback on the authority of the root domain.
Can be hosted anywhere. Since KISSmetrics has a https root domain (and they use Amazon), putting the blog on a subdomain makes for easier hosting - such as WP Engine.

And yet again, Cutts has said “[Subdomains and Subdirectories] are roughly equivalent”. In any manner, it’s a possibility worth testing.

Backlink Distribution

To monitor KISSmetrics’s backlink distribution – the links dispersed among all of their pages – I used Moz’s Pro Tools to analyze the backlinks for each page on KISSmetrics’s site.

KISSmetrics has – according to Moz, Majestic and Ahrefs – around 16,000 referring domains. According to the data, 75% of all KISSmetrics’s pages have backlinks. Many of their pages with multiple links are the embeddable infographics.

Backlink Source

To analyze KISSmetrics’s backlinks, I used Majestic SEO. Here is a plotting of their backlinks:

The site has quite a few quality backlinks. When analyzing Majestic, you are hoping to see the links fall to the top right. Citation Flow is link juice and the Trust Flow is how close those links are to trustworthy, authoritative sources.

Correlating with the data above, only a handful of KISSmetrics’s links that I checked appeared spammy, and a few of the links over-optimized anchor text. Currently, there is a trust ratio of .672 on KISSmetrics and .660 on blog.KISSmetrics, so it would be worthwhile for KISSmetrics to examine their backlink profile. (It would be encouraging to see the ratio greater than 1).

One last caution: KISSmetrics should be careful of websites using links in headers and footers that repeat across the site. These sitewide links are heavily scrutinized by Google’s Penguin update.

Anchor Text Distribution

After Penguin, Panda, etc – it’s important for KISSmetrics to conduct an anchor text analysis. Over the years, there were many ways to manipulate rankings on Google, and links were one of them. Google has always been aware of this, and they are finally cracking down (in fact, anchor text misuse will trigger the Penguin filter quickly).

This is a great rule of thumb to follow:

70% (or higher!) brand, URL, brand+keyword, and non-targeted anchor texts (Branded Links, White Noise, Naked URL, Titles, and Image)

25% partial, phrase, and broad match keyword anchor texts (Compound)

5% exact match anchor texts (Exact Match)

On the blog, the anchor text distribution follows the above guidelines. The root domain is displayed below:

85% of KISSmetrics backlinks include a branded identity - this anchor text appears natural. You generally expect to see many branded links, and 85% is rather high. It’s not a bad thing - with Google’s preference for branded terms, it’s much better than exact match anchor text.

One of the reasons the links feature the brand is because the root domain doesn’t have a blog - thus very little to link back to via anchor text.

Image Links

KISSmetrics should consider removing all links that open the same image in a new tab. For example - if you click some of the blog images on KISSmetrics - they open the images in a new tab (creating an extra link):

KISSmetrics should consider removing that extra link. These extra links are an innate WP function that can easily be changed in functions.php:

update_option(‘image_default_link_type’,’none’);

This removes one extra link on the page (saving inbound and external links) and helps with UX and CRO by keeping the user on the page. They’ve done a great job removing most of these, but a few posts have images that target new tabs.

Conversion / Leads

Once KISSmetrics has visitors on their site (mainly from SEO), the next step is to convert those visitors into leads by gathering their contact information (in KISSmetric’s case, their email addresses). This leads us into the second stage of inbound marketing, conversion and lead generation. This section will examine calls to action, landing pages, forms, and contacts.

Calls-to-Action

A call-to-action (CTA) is an image or text that encourages users to take action, by subscribing to a newsletter or requesting a product demo. In most cases, CTAs should direct people to landing pages, where collection of information can be made. In that sense, an effective CTA results in more leads and conversions for your website (via Hubspot).

Let’s start at the very beginning. KISSmetrics should have calls to action on the homepage. It’s probably the most frequently visited page on KISSmetrics and presents an opportunity to drive traffic for lead generation. In fact, some say the homepage should have at least three calls-to-action.

KISSmetrics has an offer and at least three calls to action – asking people to log in with Google Analytics, get a three minute overview, or sign in. This page is doing exactly what it is supposed to: narrowing the focus and removing clutter that could distract visitors.

On the blog, there are multiple calls to actions on the page: 3 in the sidebar, 1 in the footer, and one after the post entry. The three calls to actions in the sidebar (as you can see below), do not change:

In a perfect world, it would be nice if these offers changed (mainly the webinar) based on content, but limited personnel and resources normally restrict.

The last CTA to address is the after post CTA. During the audit, there were only two live CTA’s after the posts:

If the content was filed under the topics “Twitter” or “Social Media,” the visitor received the Linkedin CTA (the one on the right). If the user was on a page about any other topic, he or she received the CTA about marketing funnels.

As compelling as the offers can be, in some cases, they were not applicable. If a user read a post about “6 Ways to Make Customers Fall in Love With Your Brand,” they received the Marketing Funnel CTA. The CTA does connect in a way, but there could be a more compelling offer.

Or, if a user was reading “How Typography Affects Conversions,” they were greeted with the same marketing funnel offer. Again, relatable, but not the best offer. A better offer for Typography might have been a free download of:

This is a free marketing guide that KISSmetrics has already created that would make a suitable offer. I understand that creating free materials for downloads can be time consuming. But, KISSmetrics already has done all the work. They have multiple free ebooks that would make for compelling CTAs:

To improve their CTA’s, I’d suggest KISSmetrics use more than two after post CTAs and create more compelling offers by using the free material they already have. Their free ebooks are outstanding.

Landing Pages

A landing page is a web page that allows you to capture a visitor’s information. On your landing page, your visitors will find a form that they can fill out to receive an offer (HubSpot).

If you were to click on any of the after entry CTA’s mentioned above (Linkedin or Leaky Marketing Funnels), you would have been brought to one of these landing pages:

And honestly, there isn’t much to say about these pages.

These pages are to the point, use contrasting colors, gain ethos from the logo, are well designed, and are consistent.

Forms

Without knowing the sales and marketing goals behind the forms, it is hard to make an accurate analysis of the forms. Although short forms normally do better, KISSmetrics’s longer forms are warranted because they are seeking more qualified leads. They could, however, use auto generated forms.

Thank You Page

Every landing page should be followed by a thank you page. Here is what KISSmetrics’s Thank you pages look like:

I tried multiple options on the forms to generate different thank you pages. The only difference in the pages is the left page is because I asked for a demo (I’m interested) and the right is because I declined a demo.

With these thank you pages, the user is trapped with no navigation and no truly compelling next step to take.

Kissmetrics should consider bringing back the navigation, adding social sharing icons (the user did just download an offer), direct the user to other relevant content, or reconvert by asking if a user is interested in another product.

As it stands, the user will close the page and leave.

Close / Customers

Closing refers to transforming leads into customers. This generally includes a wide array of options (lead scoring, email, marketing automation, CRM, etc [HubSpot]) but since this audit didn’t engage with KISSmetrics sales team, this section will focus on contacts and email.

Contacts / Email

After I finished the form and closed out the thank you pages, I received my free downloads via email:

The download worked. The only problem with the email (and this is being picky) is that Lars Lofgren’s signature was on the landing page, but Magi and Dan sent the email. It is a small detail, but since the email is automated, it should be an easy change.

Overall, the email hit the main 6 points:

A clear subject line.
An actual person sent the email (although not my friend Lars).
Branding
Personalized
Focused CTA
Unsubscribe Link

Suggestion: there is no sharing or forwarding built into the email.

Delight

Inbound marketing is all about converting leads into customers. Just because someone has signed up does not mean the process is over. Inbound should continue by engaging with, delighting, and creating brand advocates.

This next stage of inbound marketing is all about engaging with and (hopefully) keeping their future and current customers happy.

Besides using social media as a customer service tool, we cannot ignore the fact that social media is a search engine of sorts. People on social media are researching, searching, interacting, and engaging – and at very high engagement rates. According to Statisticbrain, the number of Twitter search engine queries every day is over 2.1 billion. In any regard, a site’s success is largely dependent on social success and social engagement.

The chart below ( via Social Crawlytics ) provides a visual of how KISSmetrics’s shares are split concerning a search engine’s point of view. The center represents the first page, and each segment beyond it are pages linked from that page (in this case – the blog roll page). Segments beyond that are linked to their parent. Pages with no shares are not included or shown (click on the image to zoom in and to read the text).

As you can see from the graph above, the center represents the blog homepage, and all articles off of the homepage. The graph below breaks down that same content by social network:

Not surprisingly, KISSmetrics does really well on Pinterest. As you can tell from the graph below, many of their most shared posts come from infographics:

Engaging Content

Besides social shares and traffic (which have already been measured), the next best tell for engaging content is commenting. On average, KISSmetrics receives 22 comments for each post.

As you can see from the graph below, KISSmetrics actually receives quite a few comments for each post:

There are no recommendations for KISSmetrics on this (they average 22 comments per post). Their engagement level is great, and even Shah and Patel comment regularly (a great place for users to learn).

Summary

If you read everything in this audit example, I am truly impressed. If you cheated and jumped straight down to this section, I’ll forgive you. Either way, this section contains a summary of my most important observations for KISSmetrics and tips for you as you audit your site:

Check for Duplicate content

Always check your canonical tag (Click view source, push Command + F (Mac), and type “canonical”). It’s that easy - you don’t need a specialized tool. In this audit, there is cause for concern on KISSmetrics’s site with duplicate content and the canonical tag. As you begin to look over their third-party data, there is a strong correlation between the data (although third party) and Google’s Panda update.
Use Google’s site operators to check for duplicate content (site:domain.com). On the last Search Engine Results page, if you see the dreaded “In order to show you the most relevant results, we have omitted some entries very similar to the # already displayed.”, you need to check your content. There is cause for concern with KISSmetrics and duplicate content as this appears with their search.

Check for Indexability

Check your robots file. Make sure that you are allowing robots by visiting yourdomain.com/robots.txt. If you do not have a robots file, consider creating one. KISSmetrics doesn’t have a problem with robots.txt, but they do have a broken sitemap in the robots.txt.
Check your meta robots tags. These are page level tags that have the ability to de-index your content. KISSmetrics uses these tags well, but KISSmetrics should consider adding “NOINDEX” to subpages of their blog to prevent duplicate content.

Check for On Page Errors

Check your site for broken links (especially if you have an older blog where links may have changed). Broken links kill your UX and end your link credibility. You can use a free tool called “Integrity” to check your backlinks. KISSmetrics has quite a few hundred broken links they should update.
Measure your click depth (you can do it manually using Excel or with tools like Screaming Frog). Click Depth (or crawl depth) is the number of times a website visitor must click on a link from the root domain in order to get to the desired page (It doesn’t matter if the URL is KISSmetrics.com/name if it takes Googlebot (and users) 8 clicks to get there.). 44% of KISSmetric’s pages have a click depth of 5 or greater.
Check your site speed on Google, Pingdom, or GTMetrix (all free).KISSmetrics’s site speed is fairly slow, and KISSmetrics could easily make a few tweaks to make their site much more efficient.

Check Backlinks

Check the number of backlinks (you should mainly be concerned with referring domains). You can use limited free tools like Majestic, Ahrefs, or Moz. KISSmetrics is consistently growing their unique domain backlinks until a steep drop in March, and has started a recent growth once more.
Check your backlink anchor text. After Penguin, Panda, etc – it’s important for you to conduct an anchor text analysis. This is a great rule of thumb to follow: (70% (or higher!) brand, 25% partial, phrase, and broad match keyword, 5% exact match anchor texts. KISSmetrics anchor text is excellent - with 85% of the anchor text branded.
Check your backlink source with Majestic. When analyzing Majestic, you are hoping to see the links fall to the top right. Currently, KISSmetrics has a trust ratio of .660 on blog.KISSmetrics, so it would be worthwhile for KISSmetrics to examine their backlink profile (we were aiming for a 1).

Final Thought

As important as SEO is, it doesn’t exist in a silo. I hope this audit integrates SEO and inbound marketing to create a cohesive strategy and plan for KISSmetrics in the future. After the entire audit, I leave with a lot of respect for KISSmetrics, and I hope you do as well. I hope this helps them succeed.

This audit has proven three things:

No company is perfect. Even the best companies have to constantly monitor their internet marketing. This should be an encouragement for all – we all miss things in our inbound / SEO.
If there is a company that is putting their best foot forward in inbound, it’s KISSmetrics.
You won’t meet better guys than Hiten Shah or Neil Patel.

Best of luck, KISSmetrics.

What Do You Think?

Improving Buffer's SEO: An In-Depth Audit

Thu, 10 Jan 2013 10:18:00 +0000

Buffer’s people have spent hours creating a quality product and interacting with their customers. If you’ve ever replied to a comment on one of their posts or reached out to them personally, you know that Buffer is built on transparency and intentionality.

And if you’ve ever been as busy as Buffer – shipping great apps, creating quality content, and engaging with customers 24/7 – you know that monitoring your own website for optimization and conversion can get put on the bottom of the list. In Buffer’s case – their content marketer, Courtney, has done a great job of preventing that, but this audit will provide a third party look into Buffer’s inbound strategy.

Hi friend! Please note the date on this post. I wrote this post years ago, so while some of it might be helpful, most of it is out of date. I've contemplated deleting the post, but I like to keep it around.

If you are curious, I not only got a shirt, I got 3. And a great job -- just not at Wistia, but at Help Scout. 4+ years later, I'm still happily there.

In this post, I’ve put together a detailed, technical Inbound Marketing Audit for Buffer. This audit is widespread and comprehensive, covering a wide range of Inbound Marketing Tactics (e.g., optimization, calls-to-action, landing pages, forms, contacts, emails, etc…). Please feel free to use this as a template or example when conducting inbound / seo audits of your own. But before the inbound marketing audit, let’s cover the basics…

Why Conduct an Inbound Marketing Audit of Buffer?

First, Buffer does not have an inbound marketer / SEO person on their staff. They do have a content marketer, but content marketers normally focus on one thing: content. Courtney Seiter heads up Buffer’s content marketing, and does an excellent job – but in this audit, I hope to cover content marketing and then go beyond it to the technical implementation.

With inbound marketing being fairly new (around 2006) and SEO with a blemished name, I hope this audit validates those fields. Even more, in the future, I’d love to see Buffer create room for inbound marketing due to its impact on their business.

I also hope this audit helps Buffer succeed in their inbound marketing. Buffer’s company culture, product, and practice sets a model worth growing.

Second, I have never seen a complete inbound marketing audit (probably because of the relative newness of inbound marketing.) I’ve read countless pieces on SEO, Social Media, Landing Pages, etc… but I’ve never read an audit that ties all the pieces together. I hope that by stringing all the pieces of inbound marketing together, this audit could be useful for others as a template on which to conduct their businesses.

Who is Buffer?

Founded by Joel Gascoigne and Leo Widrich, Buffer’s goal is to make it easy for individuals and companies to manage multiple social media accounts at once. Buffer allows you to post and maintain content – articles, photos, links, etc. – on different profiles and platforms; you can even set up a specific time and date for them to release your content.

Their employees work from all over the world – Buffer believes that you should live in the place that makes you happiest and allows you to do your best work. Their goal is satisfaction for the customer and the employee.

Disclaimer

Before the audit, it’s important to point out that I am not affiliated with Buffer in any way, which means that I lack direct access to their site’s analytics, webmaster tools, and content management system (CMS). In a typical audit, I would begin by sifting through the data and then narrowing in on problem issues.

In this inbound audit, I will be relying heavily on third-party data. I hope you’ll forgive any inaccurate observations based on the information from the third parties – Searchmetrics, Ahrefs, SEMrush, Moz, etc (they are awesome tools, don’t get me wrong – they are just being used in a third party scenario).

I will also be using research and stats from Buffer themselves. Buffer publishes many of their stats and website numbers online at open.bufferapp.com. With those disclaimers out of the way, let’s get down to business…

*If you have any questions, just tweet me @elioverbey.

Since this post will be extremely long – here is a detailed navigation to help you understand where you are throughout this audit. As you will see, this audit will spend a majority of time on optimization (top of funnel), but lower areas of the funnel will be covered as well. The audit will work through four main areas as seen below: Optimization (Attract), Conversion, Customers, and Delight.

Inbound Marketing
Optimization / Attract
Conversion / Leads
Close / Customers
- Contacts
- Email
Delight
- Social Media
- Engaging Content
Summary

Inbound Marketing

In its most basic form:

Inbound marketing focuses on earning, not buying, a person’s attention – Brian Halligan

The principle of inbound marketing has been around for years, but the term has been coined by HubSpot (Dharmesh has brilliantly positioned his company with this money keyword). Inbound Marketing focuses on creating interesting and relevant content that naturally pulls people into your company and your product. By aligning yourself with your customers’ interests, you will attract traffic that will eventually turn into paying customers (HubSpot – Rachel Goodman).

As the graph below depicts (via HubSpot), inbound marketing exist in stages (or a funnel). As this audit continues on, you’ll notice how the funnel naturally works.

Optimization / Attract

The first step in inbound marketing / SEO is attracting the right audience. Content is the single best way to attract new visitors to your website. In order to be found by the right prospective customers, Buffer must not only create optimized content, but foster an environment in which content can thrive.

Search Visibility

In order to set the framework for the entire audit, it’s important to analyze the site’s performance. When looking at Buffer’s traffic, there are 3 important questions:

Is the traffic growing?
Has the site been penalized?
What is the percentage of organic traffic?

Finding accurate third party data is incredibly difficult, but Buffer posts Google Analytics traffic data monthly. They do not post traffic to bufferapp.com (just the subdomain blogs). Technically, that is all we will need for this audit – as you can see below, Buffer’s traffic declined this past month:

This is not too alarming, as Memorial Day is known to lower traffic. Overall, blog.bufferapp experienced:

679,021 unique visitors (-6.9% from last month)
908,658 total visits (-5.5%)
1,154,690 pageviews (-3.2%)

Additionally, the organic referral traffic was down as well (down from 33.9% to 33.25%):

The open blog experienced a growth, but as you can see from the graph below, the articles were not growing traffic at the end of the month (perhaps due to the engineering content).

In any regard, Buffer’s organic traffic looks excellent. One very positive thing to note: in the last 6 months, Buffer does not seem to have been hurt by any penalties from Google. Again, you cannot trust the graph above because it contains all traffic and not just organic, but other graphs from Search Metrics show a similar pattern.

I’d highly suggest Buffer monitor their organic traffic. In a recent post, they stated:

Is traffic the best way to measure whether we’re effectively reaching out goal? We don’t believe it is, so we’ll be emphasizing it less as a metric and focusing more on growing email subscribers as our main metric.

I do agree with Buffer in this aspect; however, traffic will be one of their most important components in increasing email signups. Buffer should keep a close eye on their organic and referral traffic. During the audit, there has been a strong correlation between traffic growth and email signups.

Indexability / Buffer Web Structure

In order to understand this audit, you need to understand how Buffer is set up. At first glance, you’ll notice that Buffer website is constructed in silos.

The page count was derived using two tools: Xenu Link Sleuth and a propriety tool. I compared the two findings, removed duplicates, and was left with the total below.

bufferapp.com – 27 pages
- bufferapp.com/add – 61 Pages
- bufferapp.com/developers – 14 pages
- bufferapp.com/faq – 3 pages
- bufferapp.com/guides – 8 pages
- bufferapp.com/oauth – 7 pages
- bufferapp.com/weeklydigest 4 pages
*api.bufferapp.com – 3 pages
status.bufferapp.com – 1 page
- status.bufferapp.com/history – 1 page
- status.bufferapp.com/incidents – 3 pages
jobs.bufferapp.com – 17 pages
dev.bufferapp.com – 40 pages
blog.bufferapp.com – 468 posts
- blog.bufferapp.com/author – 61 pages
- blog.bufferapp.com/tag – 438 pages
- blog.bufferapp.com/category – 48 pages
open.bufferapp.com – 94 posts
- open.bufferapp.com/author – 16 pages
- open.bufferapp.com/tag – 2 pages
- open.bufferapp.com/category – 19 pages
happiness.bufferapp.com/ – 0 pages at the root level
- happiness.bufferapp.com/ reports – 104 pages
webinars.bufferapp.com – 3 pages
- webinars.bufferapp.com/ gettingtoknowbuffer – 2 pages
*email.bufferapp.com – 5 pages

An * (asterisk) indicates that the subdomain is blocked by the robots.txt.

As this audit was being written, Buffer created yet another subdomain – overflow.bufferapp.com. Overflow is a place for their engineers to blog.

*overflow.bufferapp.com – 12 pages

For a total of 1,461 pages across all subdomains.

As you glance at the site structure of Buffer, you begin to notice that their content is heavily separated and the only way to navigate from subdomain to subdomain (with the exception of a few in-text links) is the footer [Buffer does not have a top navigation].

The Case for Subdirectories (versus Subdomains)

A lot has been said about subdomains and subdirectories. Some companies, such as Hubspot, have placed their blog in a subdomain, ie, blog.hubspot.com. HubSpot’s reasoning for having the blog in a subdomain is:

Many companies have their blog on a subdomain, and it seems to be starting to be somewhat of a standard. The search engines are treating subdomains more and more as just portions of the main website, so the SEO value for your blog is going to add to your main website domain. If you want your blog to be seen as part of your company, you should it this way (or the next way) – Mike Volpe

Even Matt Cutts has said:

[Subdomains and Subdirectories] are roughly equivalent. I would basically go with whichever one is easier for you in terms of configuration, your CMSs, [and] all that sort of stuff.

Although it seems like subdomains and subdirectories can be treated equally, Cutts uses very vague language; “roughly equivalent” is still not equivalent.

Personally, based on research and countless other stories, I’d encourage Buffer to think about subdirectories. Just because using a subdomain is becoming a standard, does not mean it the best way. I’d highly suggest Buffer consider testing out subdirectories over subdomains for 2 main reasons:

First, there is a correlation between pages on the root domain and higher rankings. According to Rand at Moz:

I would still strongly urge folks to keep all content on a single subdomain. We recently were able to test this using a subdomain on Moz itself (when moving our beginner’s guide to SEO from guides.moz.com to the current URL http://moz.com/beginners-guide-to-seo). The results were astounding – rankings rose dramatically across the board for every keyword we tracked to the pages.

I’ve also had the opportunity to see other sites do the same – with a positive result. It may benefit Buffer to consider moving some of their subdomains to subdirectories. I would not suggest they move the blog subdomain right away; instead, they could test their results with other, smaller subdomains, such as jobs, status, etc…

Second, _fewer _subdirectories tend to create a better user experience and flow.

As it stands, Buffer exist in silos. Although the subdirectories are clean, they do not make for the best of navigation.

For example, if you are a stranger to Buffer who clicks on a blog article from the search engines, you may then navigate to the career section, and you may then decide to sign up with Buffer. But, you will have technically visited 3 different sites. As a user, it’s easy to get lost in subdomains. Why not simplify the process?

There is an opportunity for Buffer to consolidate some of their subdomains.

If Buffer does indeed stick with subdomains, they should consider consolidating a few of them. Since no one is really sure how links pass among subdomains, having as few subdomains is advisable. A few of the blogs, such as open, webinars, and blog, could be consolidated into one subdomain.

I’d also suggest that Buffer test out how their links pass and how their content ranks by moving the job subdomain to bufferapp.com/jobs. Since most of the job traffic comes from referrals, it would be a harmless test.

Subdomain Links

Links will be mentioned in-depth below, but Buffer should consider how their links pass from root domain to subdomain. Since their root domain has over 12M links, and the other subdomains (such as the blog) only have 7k, Buffer should consider link equity.

Although it may be an “SEO myth” that links do not pass from root to subdomain, it still should weigh heavily on Buffer. Until Buffer has the opportunity to test link juice passing, they should take a path of least resistance.

If the root domain has over 12M links, they should take advantage of that. With only 27 pages on the root domain – and a lot of that content being less viewed, weaker content – they should at least think about the potential of combining the subdomain and root (with 301 – of course)_. Every time Buffer creates a new subdomain, they are potentially starting from scratch (or at least less). _A few months ago, Buffer moved its transparency articles (reports, revenue) off of blog.bufferapp.com to open.bufferapp.com.

_Wonder what the open.bufferapp blog would have done on the root domain (bufferapp.com)? _Open Blog has roughly 1,100 unique domains, as opposed to 50,000 unique domains on Bufferapp – or even 8,600 uniques on blog.bufferapp.

As far as we know, links may not pass through to subdomains. Thus, now is an optimal time for Buffer to consider moving subdomains.

Even more, as this audit was being conducted, Buffer created another subdomain: overflow.bufferapp.com. This new subdomain gives a perfect case study of what happens when you create a subdomain. When Buffer created overflow.bufferapp.com – the blog for everything development related – they took the posts and moved them from open.bufferapp to overflow.bufferapp. As you can see below, the traction for the new subdomain is slow (the search query is from their most recent article on overflow.bufferapp):

As you can see from the chart above, the previous subdomain is ranking higher than the new subdomain (in fact, the new subdomain isn’t even ranking). This will eventually cause duplicate content issues as well. Buffer is essentially starting from scratch with this new subdomain. As evidenced in the chart below, the new domain has only 2 links.

In creating this new subdomain, Buffer left over 1,1o0 unique domains on open.bufferapp.com. Once again, Buffer is potentially starting from scratch with 2 unique referring domains. In this case, it would be prudent for Buffer to weigh the cost of creating a subdomain before they segment their articles.

The Case for Subdomains (versus Subdirectories)

Buffer’s Traffic is Growing on Subdomains

Buffer has always built their blogs on subdomains – and why mess with a good thing? In March – April of 2014, Buffer moved their transparency documents (open salaries, analytics, revenue, etc…) from blog.bufferapp to open.bufferapp. According to analytics:

March: 14,694 unique visitors – April: 43,209 unique visitors (+203%)
March: 20,158 total visits – April: 56,399 total visits (+188% )
March: 31,198 pageviews – April: 79,929 pageviews (+157% )

By all means, the traffic seems to be growing, and as such, Buffer should not change a thing.

Buffer Can’t Build On Their Root Domain

It is possible that Buffer _can’t _build on the root domain in order to be “white hat” with their links. Have you ever clicked on the Buffer icon button when sharing an article and noticed what URL is in the pop up?

It’s https://bufferapp.com/add. This means that every website that has a Buffer icon is technically hosting a Buffer link (making their root domain backlink profile massive). Check out the image below (this website “openid.net” was picked because it was listed as a backlink from an analysis on Majestic):

I highlighted two things of note on this chart. First, you’ll notice the Buffer button on the left side of the screen. On the right side of the screen is the code behind that button. As you can see:

The Buffer Social Button is hyperlinked to http://bufferapp.com/add
The Hyperlink does _not _include a “nofollow”

Based on what’s happening here, since the Buffer social links are not naturally earned, but are rather accumulated from the growth of a social product, Buffer does not host their blogs on the root domain in order to avoid link penalties.

Again, that is speculation, but based on the culture of Buffer, this theory fits in line with their philosophy of openness and transparency.

Whether Buffer chooses to stay with subdomains or test subdirectories, they should always be aware of the data.

Robots

A robots.txt file is used to restrict search engines from accessing specific sections of a site. Here is a copy of Buffer’s root domain Robots.txt file:

On their blogs, Buffer uses the another robots.txt file:

The records in the robots file politely ask all bots (bots used by all search engines are identified with the aterisk) to not crawl specific directories.

There are a few improvements that Buffer could make in the construction of their robots.txt.

First, Buffer should add sitemaps to the robots.txt. This will help the robots to index their pages.

Second, Buffer should consider blocking ‘/wp-content/plugins/’. Sometimes developers put links in their plugins.

Third, Buffer should rethink blocking /wp-includes/’ in robots. There are better solutions for blocking robots than in the robots.txt file (ie, NOINDEX – discussed below).

Fourth, Buffer should remove the disallow of all directories on their new blog overflow.bufferapp.com:

In its current condition, the site cannot be indexed.

Robots Meta Tag

Each page on a site can use a robots meta tag to tell search engine crawlers if they are allowed to index that page and follow its links. Very few of Buffer’s pages have a meta robots tag – which is fine – they are unnecessary unless they are trying to specify a hidden page. In the case mentioned above (wp-includes), it would benefit Buffer to include a “noindex” robots meta tag on a per page basis, rather than blocking an entire directory. WordPress is a great open source platform – one that Buffer blogs are built on – but duplicate content is one thing you have to be very mindful of. Using the robots meta tag will help prevent duplicate content. Content duplication issues include /tag, /category, and /author. As you can see above, the tags are indexed. Since these tags take up crawl bandwidth, but don’t have unique content, and historically have a low CTR, Buffer should “NOINDEX, FOLLOW” these archive based pages. Buffer has successfully implemented this change on subpages and should consider adding the same meta robots to the other archive based structure. The “NOINDEX, FOLLOW” will remove the tags, categories, and authors from the index, but will still allow links to pass.

XML Sitemaps

What are sitemaps?

Sitemaps are an easy way for webmasters to inform search engines about pages on their sites that are available for crawling. In its simplest form, a Sitemap is an XML file that lists URLs for a site along with additional metadata about each URL… Sitemaps.org

I identified two Sitemap listings in Buffer’s robots.txt file:

The first listing points to blog.bufferapp Web Sitemap, which should include URLs for every page the site wants search engines to crawl and index. The second listing points to the open.bufferapp Sitemap. You can see one of Buffer’s sitemaps below (they are identical):

In both sitemaps, I’d remove the page section – as it links to a blank “sample page”. If Buffer does choose to NOINDEX tags and categories, they should remove them from the sitemap.

Since Buffer uses quite a bit of video, it may also help to compile a comprehensive list of the site’s indexable pages with videos on the page (i.e., pages that are substantive enough to warrant video inclusion) and creating a video sitemap using that list (they can use Yoast’s extension for that as well).

Accessibility

Performance

According to Google Webmaster tools:

You may have heard that here at Google we’re obsessed with speed, in our products and on the web. As part of that effort, today we’re including a new signal in our search ranking algorithms: site speed. Site speed reflects how quickly a website responds to web requests… While site speed is a new signal, it doesn’t carry as much weight as the relevance of a page.

Though it’s tempting to dismiss site speed as an important SEO ranking factor, if Google says it matters (even a small percentage), it matters.

Buffer’s site speed is decent, but Buffer could make a few tweaks to make their site much more efficient:

Reducing the number of files needed to load the site, and thereby reducing the number of HTTP requests, will make Buffer’s site load more quickly.

There are usually three parts to fixing this:

Reduce the number of JavaScript files
Reduce the number of CSS files
Reduce the number of images

The Time to First Byte is extremely quick, but the page load is somewhat slow – anywhere from 2.66 seconds to 3.15 seconds (Pingdom). They received a 55/100 on Page Speed Insights via Google, and 80/100 on Pingdom. Buffer’s homepage has room for improvement. Another concern is that_ their blogs have over 100+ requests with a page speed of 3 seconds._

The site could be improved with the following:

Eliminate render-blocking JavaScript and CSS
Minimize HTTP requests – Buffer’s pages will load more quickly with fewer requests. Minimizing these requests involves reducing the number of files that have to be loaded, such as Javascript, CSS, and images.
Combine Javascript and CSS into external files with links from the header. This allows the external page to be cached so that it will load faster.
Implement server side / browser caching – This creates a static html page for a URL so that the dynamic sites don’t have to reload / be recreated each time the URL is requested.
Load Javascript asynchronously.
Finally, use 301′s only when necessary. A 301 forces a new URL, which takes a longer time to load.

Site Architecture

Currently, Buffer’s average page level is 3.7 (0 being the homepage). At this point, Buffer should:

Examine and prune their content that doesn’t produce results
Reduce the number of content silos on their pages by creating a flatter, better connected architecture

First, content is like a grapevine: In order to produce the best fruit possible, you really need to prune. I’d suggest that Buffer sit down with Google Analytics, Backink profiles, etc. and decide which pages are brining in traffic, shares, etc, and which articles are not. Based on the results, Buffer can create better content that engages and prevents keyword cannibalization.

On the other hand, there are posts that have aged poorly. If they are not driving traffic, increasing shares, ranking, or converting, Buffer should consider removing them. In removing content, Buffer should 301 duplicate content and let the rest of the content 404.

Secondly, Buffer should create a flatter site architecture. Right now, Buffer exists in 4 main silos: home, blog, open, and overflow. The Buffer homepage links to a silo (ie. blog), and silos link to posts. This creates lots of information, but links are flowing down, not around.

In reworking their site architecture, these vertical silos will disappear.

Usability

Buffer has a few layout issues that reduce the website’s usability and effectiveness.

First, noticeably missing from the Buffer’s website is a navigation menu.

The only way to find something on Buffer’s website is to go down to the footer. While Buffer should consider adding a NAV, they may also want to add a search functionality to help users, and to figure out what exactly the user is searching for in their site page analytics.

Next, Buffer’s site is mobile responsive. But, on mobile, all of Buffer’s call-to-actions on the homepage (hello bar, pop up, etc…) fall off:

It would be beneficial for Buffer to add a relevant call to action on the homepage for mobile devices. Finally, at some breakpoints, Buffer’s site could use a little adjusting:

Overall, Buffer’s site is usable on both mobile and desktop. Apart from a few minor tweaks, Buffer is doing well for usability (as it relates to inbound).

Click Depth

As the graph below shows, Buffer is doing a good job of keeping the required clicks minimal – the subdomains help keep the URL paths minimal.

On Page Factors

In this section, I will investigate the characteristics of Buffer’s pages (e.g., URLs, schema, duplicate content, etc.) that influence the site’s search engine rankings. On Page search factors are those that are entirely within Buffer’s own control.

HTML Markup

A site’s HTML is important because it contains some of the most vital on-page ranking factors.

When examining Buffer’s pages, I found multiple pages with 20-40 errors per page. Many of the issues are easy fixes: open tags, stray elements, attribute errors, etc… Although these aren’t extremely critical issues, it’s best to clean up as many errors as possible (empty headings, duplicate IDs, attributes, etc).

Structured Data

Schema.org microdata is an extra set of HTML tags that you can add to your HTML elements to let the search engines know specifically what you are talking about.

Fortunately, Buffer does define structured data, but fails to add authorship to most of its writers. Though a few of Buffer’s articles include authorship, such as the one below, including authorship in all articles would be greatly beneficial:

Sadly, many others do not have authorship enabled, and even Kevan – by far one of their best writers- does not have authorship enabled for the open blog (pictured above is Main blog).

In order to add authorship, here’s what you can do: create a Google+ profile for each of the site’s authors. Then, link from their Google + profile back to the blog, and vice versa.

Be Careful with Schema.org

One caution: Be wary of schema.org for all HTML elements.

Google uses micro markup (schema) to scrape content and place it on Search Result Pages – this could potentially hurt click through rates (CTR). The more information you tell Google – the more you risk Google scraping your content and your site losing sessions.

Meta Descriptions

Meta Descriptions do not influence the ranking factor of a site’s page, but they do influence the CTR (click-through-rate).

These short paragraphs (normally 155 words) are a website’s opportunity to advertise content to searchers and to let them know whether the given page contains the information they’re looking for.

Out of 1,449 unique HTML/Text URLs, 893 do not include a meta description, and 92 of those descriptions are duplicates. What this means is that in almost all of the searches, the meta descriptions are automatically generated by the search engine. Many of the duplicated meta descriptions are on the root domain (bufferapp.com). As you can see below, the home page has the same meta description as the about page, and as the “awesome” page:

2 Options to Fix

The first answer would be adding a unique meta description for each of the page’s content.

The second answer is a bit more complex. If a page is only targeting between one and three search terms, then by all means, write a meta description that targets those key words. But if the page is targeting long-tail searches, it may be easier to let the search engines pull in the meta descriptions themselves.

The reason the second answer is more feasible is because when the search engines pull content, they pull content and display keywords surrounding the user’s query. If you force a meta description, it can detract from the relevance of long tail search.

In some cases, search engines will overrule your meta description anyways, but you cannot always rely on that. In that regard, it would be advantageous for Buffer to add meta descriptions for pages that target only a few keywords and leave the other long-tail pages open for variance.

One last issue to be aware of when thinking about meta descriptions:

Social sharing sites will commonly pull the meta description when sharing – unless open graph is enabled (more on that in a moment). Without the meta description, social sharing sites will just use the first text they can find (sometimes Google will as well – see below).

Not using a meta description may create a bad user experience for social sharing. In Buffer’s case, they are using open graph tags with “descriptions”, thus I’d encourage them to add meta descriptions as well.

Head Tags

rel=publisher

Buffer has a great looking Google+ page, so they ought to take advantage of rel=publisher. This tag has seen a much smaller adoption rate compared to Rel=author, and its name may cause the confusion. The publisher tag can be used by any business or brand website that has a corresponding Google+ page.

Publisher connects your Google + page to your website, and that results in your Google + page being displayed in search results whenever anyone is searching for your brand name.

Buffer has no rel=publisher markup code on their home page. In fact, if you analyze the page with the structured data tool, you will notice there is no publisher schema on the page at all. However, if you look at Buffer’s Google+ profile page, you will notice they have a verified website listing – which is why they are currently receiving the rich markup:

Rel=publisher markup is not required for the logo to show up. All that is required is for the company to have a Google+ page that has their website verified in their Google+ profile. But, it wouldn’t hurt for Buffer to add the publisher markup, just to make sure their rich snippet stays in place.

rel=canonical

Buffer is using Yoast’s SEO plugin – a great choice. Buffer has the plugin set up well, but there are a few conditions Buffer should be aware of with WordPress.

First, Yoast and rel canonical for paginated WordPress URLs are technically correctly implemented. Yoast’s plugin applies a rel-canonical meta tag to itself. Depending on your source, this could be a slight misuse of the rel-canonical tag (per Google’s Webmaster Blog, which says rel-canonical should be replaced with rel-prev and rel-next for page 2, 3, etc.). On the other hand, millions of blogs use this setup of the canonical with no problem.

The current set up canonicalizes the subpage to itself. Not a crisis (or even a big deal), but it is something to monitor as Google updates its documentation.

According to webmaster tools, Google actually recommends that you rel=canonical to a “View All” page. The current set up (based on the plugin) is technically incorrect because the subpage is canonicalized to itself – not a view all pages. Buffer could fix this problem with one of 2 solutions.

Buffer should use a canonical tag that points to a view all page.
Or use Rel=prev/next.

Excluding subpages, each Buffer page also uses a rel=”canonical” link; however, all of these links are self-referential and shouldn’t be a problem. rel=prev / next Google asks that Webmasters add rel=”next” and rel=”prev” to paginated archives, so that Google can distinguish them as a series and send users to the most relevant pages. As far as I can tell, Buffer has not implemented this throughout their site.

Open Graph

The Open Graph protocol enables any web page on Buffers’s site – or all sites – to become a rich object. For instance, the Open Graph is used by Facebook to allow any web page to have the same functionality as any other object on Facebook. Currently, Buffer does implement Open Graph protocol on their sites (with the exception of happiness.bufferapp.com). This is the open graph information in their root domain (bufferapp.com) markup:

Which translates into:

It may be best, however, that on bufferapp.com, Buffer include these basic og: tags at the very minimal:

Currently, on the homepage (and job.bufferapp.com) they are omitting type and URL. I’d suggest adding those tags. On their blogs (open.bufferapp and blog.bufferapp), they are correctly implementing open graph (using Yoast’s plugin to automatically generate og: markup).

Twitter Cards

Twitter cards are another micro markup – essentially rich snippets for Tweets. Twitter will actually fall back on the Open Graph tags used for FB, but the Open Graph tags aren’t comprehensive. All of Buffer’s blog pages include Twitter Card tags, which attach additional information to the Tweets associated with those pages. However, the root domain should also include (in addition to the og: tags):

Titles

According to Cyrus Shepard at Moz:

Title tags—technically called title elements—define the title of a document. Title tags are often used on search engine results pages (SERPs) to display preview snippets for a given page, and are important both for SEO and social sharing.

Each page that is indexed should have unique content and a unique title that effectively summarizes the content for users and search engines. Out of the 1400+ pages on Buffer’s website, only 90 pages do not have unique titles.

Most of the pages that do not have a unique title derive from happiness.bufferapp.com. As you can see below, in the way that the happiness site is set up, the “title” tag is left empty.

On these pages, Buffer should add a unique title that describes the pages.

Buffer does have a duplicate title issue with their development site. They should consider “noindexing” their dev site to prevent that error (more on that below). Additionally, many of the pages on the root domain have duplicate, non descriptive titles. Every page inside of bufferapp.com/developers holds the same title “Buffer – Developers”.

Once again, adding a unique title that describes the pages would be enormously beneficial.

Google will show as many characters that a 512 pixel display can show – thus, around 50-60 characters. As you can see below, most of Buffers titles are at an average of 69 characters long, and very few of them reach 80-90 characters in the title.

Shortening some of these titles would help Buffer to avoid truncation. For example, Buffer’s most recent title on the blog:

“10 Marketing Lessons From Billion-Dollar Businesses: Inside the Strategies of Facebook, Mint, and AppSumo”

is 105 characters long and looks like this in the SERPs:

In some cases, when Buffer’s title tag is too long, Google no longer truncates the title and adds an ellipsis to the end; instead, Google tries to algorithmically determine a better title for the post.

Titles matter so much that when Google wants to return a result, it changes the title to better match the query. The reason for this is simple: users scan for and assign higher relevance to titles that include their query.

In this case, keeping titles within 50-60 characters is the wisest option.

Along the same lines, many of the site’s pages in the root domain (bufferapp.com) use titles that are not sufficiently descriptive (e.g., “Buffer – Android FAQ”, “Buffer – FAQ”, etc.).

These titles need to provide more information about the contents of their corresponding pages. Users may be less likely to click through to those pages, and search engines will be less likely to rank those pages at the top of their search results.

Most of the site’s titles incorporate the brand name at the end to help promote brand awareness, but not all of them do.

The older blog posts incorporate branding:

4 Little Known And Underused Twitter Tips For Bloggers – The Buffer Blog

but the new blog post do not incorporate branding at the end:

The Research and Science Behind a Perfect Blog Post

Buffer can easily fix this issue on their blogs by setting up a template (with Yoast’s plugin they currently use) that automatically appends the title with the brand at the end. To help promote brand awareness, I recommend appending the brand name to the end of the site’s titles.

URLs

Buffer has done a great job crafting their URLs. Their URLs are at an average length of 50 characters (including bufferapp.com/ or subdomain.bufferapp.com).

Ignoring the subdomain structure (as opposed to subdirectory), the site’s URL structure is organized and clean. Each URL also contains at least one keyword.

Buffer keeps their root URLs short (averaging 26 characters), ie:

bufferapp.com/about

bufferapp.com/business

And their blog URLs lengthier (averaging 65 characters), ie:

blog.bufferapp.com/science-of-storytelling-why-telling-a-story-is-the-most-powerful-way-to-activate-our-brains

blog.bufferapp.com/people-dont-buy-products-they-buy-better-versions-of-themselves

The URLs work seemingly well. The root domains are short for memory and the blog URLs are a little lengthier for search indexing.

Unfortunately, while crawling the site, I found URLs on 738 pages that returned a 403 or 404 HTTP status codes (i.e., those URLs are no longer accessible). If possible, those links should be redirected and updated.

Another important URL-related consideration is the domain name. Buffer’s domain (bufferapp.com) is obviously appropriate for the brand; however, Joel (the domain’s registrant) hasn’t extended the domain’s registration, and now, it’s set to expire in October, 2014 (i.e., four months from now).

At some point in the near future, it would be great to see Buffer buy the domain Buffer.com. Bufferapp.com works, but it can be confusing (especially since Buffer changed Twitter names from @bufferapp to @buffer).

One last point: Buffer’s new blog overflow.bufferapp.com should remove the date from the link structure:

Keywords

The Meta keywords tag is no longer a valid tag employed by the search engines, as it was a method abused by spammers in the past.

For now, Buffer won’t be penalized by Google for their existing use of keywords, but on the other hand, Bing still takes keyword tags into account as a signal for spam. In either situation, it would be advantageous for Buffer to remove the keywords tag.

Another reason for Buffer to remove the Meta Keywords tag is that it provides an easy way for their competition to see the words they are trying to rank for.

What’s most disconcerting of all is that if these keywords are truly the words Buffer is trying to rank for, then they need to conduct more keyword research. Some of the keywords make sense, but even attempting to rank for “Twitter”, “Linked In”, etc… would be difficult and would probably garner a low CTR. Buffer should consider doing more research around “Social Media Management Dashboard,” “Hootsuite alternative”, etc… These types of phrases would be more beneficial to their audience.

Images

For the benefit of search engines, compliance (HTML), and the visually impaired – all images on Bufferapp should have an ALT tag. The ALT tag should accurately describe the image, and if possible, contain a keyword relevant to Buffer (but only if the keyword is relevant to the image).

Additionally, image file names should be descriptive – not containing random numbers or queries. The file name should describe the image and, once again, use a keyword if possible.

If Buffer ensures that images follow these two rules, they will increase the likelihood of referral traffic from image searches. Buffer has hundreds of images, and around 81% of the images take advantage of the alt text.

One area that Buffer can improve in is the descriptiveness of the file naming. As you can see below, quite a few of the images are just random strings.

There are hundreds of images like the ones above on Buffer’s blog. I’d suggest adding descriptive title tags to the images to increase organic image search, and also to be HTML compliant. If you’ve ever read a blog by Buffer, you will notice they rely heavily on Screenshots. In fact, 29% (817) of all their images are screenshots. In order to avoid file naming structures such as “Screen-Shot-2014-01-15-at-1.30.44-PM.png” Buffer should consider using Awesome Screenshot Addon for Firefox. When taking a screenshot with the addon, the file is automatically named with “Title

Description”. It will save Buffer time and will automatically create files that are semantically named.

Content Duplication

Buffer is doing a excellent job keeping their content unique and fresh (much thanks to Belle and Courtney). Buffer does, however, need to be careful with duplicate content and canonicalization. Duplicate Content issues with the Development Site Buffer uses a development site – dev.bufferapp.com – to test before pushing to production. As you can tell from the screenshot below, the dev site is a mirror of the production site: This is normal practice to test changes on a UA, Dev, and staging server before pushing changes to production, but in this case, I’d suggest that Buffer do one of the following:

Bring dev.bufferapp.com internally (which may not be easy due to a distributed work force).
Add a robots file that Disallows the dev subdomain.
Add a meta robots tag that “NOINDEX, NOFOLLOW” every page on the site.

This is not a major issue right now because there are only 40 pages on the development site, but the last thing Buffer would want is duplicate content – or even worse – people linking to the development site. Duplicate Content Issues with Overflow Overflow is a new blog that Buffer just created. They took content from the open blog and moved it to overflow. During this process, they did not redirect the old content to the new location, leaving duplicate content. As you can see below, Buffer has content living in two blogs: Right now, there isn’t a huge urgency to fix this matter because the overflow blog is being Disallowed by robots.txt. However, Buffer should 301 redirect the content from open to overflow fairly quickly to avoid duplicate content issues. I didn’t want to wait until the audit was published to contact Buffer, so I went ahead and sent this information over. The robots disallow could harm organic. Buffer replied with:

@elioverbey Thanks Eli, will pass along to double check! -Mary — Buffer (@buffer) June 10, 2014

Duplicate Content Issues with https Since Buffer’s site has secure pages (designated by the “https:” protocol), Buffer should be careful of both secure and non-secure versions getting indexed. Ideally, these problems will be solved by the site-architecture itself. In many cases, it’s best to Noindex these pages.

Off Page Ranking Factors

This section will cover some of the most influential factors in ranking: backlinks.

Backlinks

Links are not everything when it comes to SEO, but search professionals and companies attribute a large portion of the search engines’ algorithms to backlinks (Larry Page based Google off of the Educational Citation Method – ie. A professor who was cited by other authors was considered more credible – Read his dissertation for more information). When analyzing backlinks, you should consider: Total number of links, link quality, number of unique domains, fresh/incoming links, and anchor text usage. To start, I first recorded how many backlinks Buffer had received over the past few months. The charts below show the growth of Buffer’s backlink profile over the course of one year. After comparing the backlink analysis above to both Majestic SEO and Moz, there was not much of a difference. Buffer was consistently growing their unique domain backlinks until April, and has started a recent growth. The reason Buffer’s backlink profile is massive is due to the Buffer social links pointing back to their root domain. To get a clearer view of what a natural link profile (and by natural, I mean not placed), here is a look at Buffer’s blogs (open and blog – at the time of writing the audit, overflow did not have any links…). [The first graph is blog.bufferapp – the second is open.bufferapp]: As you can see from the data above, their backlinks are growing, but the Bufferapp root domain is outpacing the blog, which is outpacing open. If Buffer continues to open subdomains (which is great for segregating content, bad for links) they need to keep a close eye on their links. This past month, the open blog traffic dropped: This graph includes all traffic (not just organic), but since the subdomain had been moved a month before, it could be a drop due to the lack of backlinks. The articles went from a subdomain with 10,000 backlinks to a subdomain with 1,100. Even though it was a month later, sometimes the changes are delayed in Google. To see how Buffer’s backlink growth matches up with their competition, I compared Buffer’s growth with one of their competitors over the past 5 years: Hootsuite. This charts shows 2 things:

Buffer’s backlinks are growing.
They have fierce competition.

The reason Buffer has been able to make up so much ground against Hootsuite is due to the quick adaption of the Buffer button. Without the Buffer button, their links do not come close to Hootsuite’s. Almost every month, Hootsuite is acquiring more links from more unique referring domains. The backlink analysis shows the competitiveness of backlinks:

302 Redirects {#302}

Buffer implements redirects correctly throughout, but there is cause for some concern on the homepage. There are technical reasons for implementing a 302 redirect to https, but since this redirect will stay around, I’d suggest Buffer change the redirect to a 301. As you can see below, http://bufferapp 302 redirects to https://bufferapp: Unless there is a technical reason for this, I’d suggest Buffer consider adding a 301.

Backlink Distribution

To monitor Buffer’s backlink distribution – the links dispersed among all of their pages – I used Moz’s Pro Tools to analyze the backlinks for each page on Buffer’s site. Buffer has – according to Moz, Majestic and Ahrefs – around 42,000 referring domains. But, upon closer inspection, many of those links are pointing to root (via social links). According to the data, 76% of all Buffer’s pages have no backlinks. Of the 24% that do have links, the homepage accounts for almost 23% of those links (the data is skewed here because of the massive amount of backlinks to the root domain via the Buffer social sharing icon).

Backlink Source

To analyze Buffer’s backlinks, I used Majestic SEO. Here is a plotting of their backlinks: The site has quite a few quality backlinks. When analyzing Majestic, you are hoping to see the links fall to the top right. Citation Flow is link juice and the Trust Flow is how close those links are to trustworthy, authoritative sources. Correlating with the data above, a handful of Buffer’s links that I checked appeared spammy, and a few of the links over-optimized anchor text. With a trust ratio of .842 on bufferapp and .660 on blog.bufferapp, Buffer really needs to examine their backlink profile. (It would be nice to see the ratio greater than 1.) One last caution: Buffer should be careful of websites using links in headers and footers that repeat across the site. These sitewide links are heavily scrutinized by Google’s Penguin update… Buffer’s entire sharing model is based on multiple pages having the same backlink. Buffer should monitor webmaster tools closely to make sure they stay in line with Google’s regulations.

Anchor Text Distribution

After Penguin, Panda, etc – it’s important for Buffer to conduct an anchor text analysis. Over the years, there were many ways to manipulate rankings on Google, and links were one of them. Google has always been aware of this, and they are finally cracking down (in fact, anchor text misuse will trigger the Penguin filter quickly). This is a great rule of thumb to follow:

70% brand, URL, brand+keyword, and non-targeted anchor texts (Branded Links, White Noise, Naked URL, Titles, and Image)
25% partial, phrase, and broad match keyword anchor texts (Compound)
5% exact match anchor texts (Exact Match)

Below is the Anchor Text chart for Buffer… Buffer does very well. If you examine the chart closely:

2% of Buffer links come from “Buffer,” which is a branded identity.
4% of searches include a partial match, etc…
94% include the other types of links: Naked, URL, etc..

Buffer has done a great job of keeping their links diverse. The other blog – open.bufferapp – appears natural as well. As the company grows, the branded terms should grow as well, but again, that would be a natural progression. The only cause for concern is bufferapp.com – as you can see below, the Anchor text distribution is unnatural: The main reason for this is because Buffer uses bufferapp.com as the link location for all sharing functionality produced by Buffer. If you’ve ever created a Buffer share button, you would have seen:

If you look closely in the code above, you can see the anchor text is “Buffer” – which is why 34% of their anchor text on bufferapp.com is “Buffer”. In this case, since those links are technically placed links, Buffer should keep a close eye on it’s backlink profile, webmaster updates, and potential penalties. A future penalty on the root domain could hurt all domains.

No Follow Links

With any link that you place on your site, NoFollow or not, know that these links act as a gateway into other sites that introduce your audience to others’ content. So before you place a well-positioned link, think of what that link could to your site traffic and your content reach. As you can see below, the Hello Bar that Buffer uses places a follow link on every page (over 1,000).

If I were Buffer, I’d reconsider the implementation of the HelloBar for 2 reasons:

First, the HelloBar will inevitably take users away from Buffer’s page. Since the link is at the very top of the page, some of Buffer’s quality, top-of-the-funnel users will be lost to Hello Bar.

Second, since the links are placed – and one day may be considered a link scheme – Buffer should consider upgrading the Hello Bar to remove the links. Upgrading HelloBar is a low fee that will remove both issues (users navigating away and multiple links). According to Webmaster tools:

Widely distributed links in the footers or templates of various sites are unnatural and potentially violate guidelines.

In fact, Buffer should consider talking to Hello Bar about adding a nofollow (in fact, HelloBar could be considering adding a no follow in the future). One other consideration with no follows: Buffer should moderate their personal placement of Buffer links via the social share icons. With Google changing its stance on allowable links frequently, that is an important topic to remember.

Images

Lastly, Buffer should consider removing all links that open the same image in a new tab. For example-

(or just click the image below to see the action I am talking about):

</a>

Buffer should consider removing that extra link. These extra links are an innate WP function that can easily be changed in functions.php:

This removes one extra link on the page (saving inbound and external links) and helps with UX and CRO by keeping the user on the page.

Conversion / Leads

Once Buffer has visitors on their site (mainly from SEO), the next step is to convert those visitors into leads by gathering their contact information (in Buffer’s case, their email addresses). This leads us into the second stage, conversion and lead generation. This section will examine calls to action, landing pages, forms, and contacts.

Call to Action

A call-to-action (CTA) is an image or text that encourages users to take action, by subscribing to a newsletter or requesting a product demo. In most cases, CTAs should direct people to landing pages, where collection of information can be made. In that sense, an effective CTA results in more leads and conversions for your website (via Hubspot).

Let’s start at the very beginning. Buffer should have calls to action on the homepage. It’s probably the most frequently visited page on Buffer and presents an opportunity to drive traffic for lead generation. In fact, some say the homepage should have at least three calls-to-action.

Buffer has an offer and at least three calls to action – asking people to log in with Twitter, Facebook, and Linkedin. The only shortcoming on this page is the CTA’s don’t lead to a landing page. This leaves two potential gaps.

First, a user could complete the login process and never come back. If a user logs in with Twitter or Linkedin – via the CTA – and there is no landing page to collect more information, the user could be one and done. As you can see below, if a user logins with Twitter or Linkedin, Buffer does not get their email information:

The other forms of login (Facebook and Email) do allow for the collection of email, but perhaps Buffer could improve gathering more information.

Secondly, without email capture, lower funnel lead generation will become extremely difficult for Buffer. In a January post, Buffer detailed how they redesigned the CTAs on the homepage. In the article, they recount that the new homepage led to a 16% increase in overall new user conversions. This is great, but conversion in this case is top of the funnel – not paying users. Thus, 16% of those “new users” may have been one and done. If Buffer can create a landing page that collects email addresses on every signup for marketing automation and lead nurturing, they may increase their paying customers.

On the blog pages, Buffer also includes CTA’s, but the offers do not change and can be a little overwhelming: At the bottom of every blog page (open blog included), Buffer clusters its CTA’s, so much so that the sidebar CTA is hidden by the pop up footer CTA. In this case, it would be beneficial for Buffer to remove some of their CTA’s at certain scroll depths. It’s great that their CTA’s are above the fold, but when they all merge, the user is overwhelmed with 4 places to click.

2 Quick Suggestions:

With the exclusion of the footer popup (#4 on the screenshot above), every CTA is hyperlinked back to the homepage. Buffer should attempt to direct each CTA to its own dedicated landing page that restates the offer and copy from the CTA. At this point, the homepage seems to function as a dumping ground for all miscellaneous traffic. CTAs give Buffer the opportunity to increase conversions, reducing the amount of miscellaneous / anonymous traffic, but to take advantage of the CTAs, separate landing pages would be more effective.
Buffer should attempt to use a different call to action in the footer (#2) and sidebar (#3) on each post. Creating different CTAs for different personas and audiences increases the effectiveness of their CTAs. If Buffer only uses one CTA for all blog posts, then they only hit one persona. Implementing different CTA’s increases the chances of targeting missed audiences.

Finally, Buffer needs to make sure their CTA’s are honest. Although this is accidental, notice the discretion between the number of email sign-ups at the top of the page and the sidebar:

Again, this was a complete accident – or perhaps – even a lack of updating information (20,000+ users vs. 19,638), but I’d highly encourage that Buffer verify their numbers.

Landing Pages

A landing page is a web page that allows you to capture a visitor’s information. On your landing page, your visitors will find a form that they can fill out to receive an offer (Chris LoDolce at HubSpot).

On Buffer’s site, however, the only landing page (either through CTA, Social Media, etc…) is the home page. Mainly, Buffer’s referral traffic lands on a blog page – which contains a CTA – which directs them back to the homepage:

In this case, Buffer is directing all of their traffic back to the home page, thinking that their visitors will “figure out” where they want to go. But with so many opportunities and the limited attention of the user, the home page is almost too broad and generic, rather than being narrowly focused to one particular topic to one particular persona.

For example, what if a user was solely interested in using Buffer to schedule tweets? Perhaps creating a landing page for that specific purpose and persona would drive up conversions (according to Interactive Marketing Inc., keeping relevant, focused, important information on a single page can increase conversion by 55%.).

Buffer should examine the possibility of creating custom landing pages for each type of content that could trigger a conversion. Creating landing pages will also allow Buffer to closely target their audience using different content that may appeal to different segments of their audience (perhaps segments Buffer did not know about). These different landing pages will offer Buffer a baseline conversion path that they can build upon.

Forms

Buffer does not use many forms – and perhaps, it’s because they don’t want the forms to come across as a hard sell. Nevertheless, it would be advantageous for Buffer to use a few more detailed forms to collect information. No matter how you spin it, Buffer does have name squeeze forms (an internet marketing tactic that presents users with a choice: enter your name and email to get a freebie, or go elsewhere). So, Buffer might as well see if they can at least ascertain more of the users’ information.

Buffer could create a few CTA’s that lead to unique landing pages with forms that collect name and email. Perhaps these forms could increase conversion. As it stands, a user could complete the form process and never come back. If a user logs in with Twitter or Linkedin – via the CTA – Buffer does not collect emails, and the user could be one and done. Or even worse, the user could log in, love the software, then get distracted – go to another site – and then, Buffer has no way of contacting them.

Close / Customers

Closing refers to transforming leads into customers. This generally includes a wide array of options (lead scoring, email, marketing automation, CRM, etc [via Nick at HubSpot]) but since Buffer doesn’t have a sales team, this section will focus on contacts and email. According to Leo:

For now, we decided that we wouldn’t go and build a sales team. The reason for this is that we’d rather try and go the Customer Success route for now. Sales teams come with a lot of interesting challenges and some of them don’t quite seem to be aligned with the type of culture and company we want to build at this point.

Contacts / Email

Buffer can accomplish their start up goals – without a sales team – if they continue to use best practices of inbound marketing. At some point, however, they will probably have to create a sales team to focus on leads.

In any manner, Buffer could improve their email communication / marketing automation. As previously mentioned, if a user logs in with Twitter or Linkedin – via the CTA – Buffer does not collect emails, and the user could be one and done (it is also not clear if Buffer can use the emails from FB login). During the month that this audit was being created, Buffer was collecting more email addresses than before (23,629 total email subscribers (+6%)).

But as emails grow, so should Buffer’s outreach. When a user signs up for Buffer’s email (through Hellobar or the popup), the visitor then must complete the double opt in (which is great in the prevention of SPAM). I signed up and you can see the first email I received:

Immediately, three problems arise that are compounded by other issues:

Due to the lack of forms, the email is not personalized. Buffer has no way to distinguish who I am, what my name is, or where I am at on my buyer’s journey.
Due to the lack of relevant CTA’s and landing pages, the email is not targeted. The first email I receive is the generic email that gets sent to everyone else. Because Buffer does not use landing pages, they cannot segment my email into a list – and provide me with the related information I found interesting.
Finally, the emails do not use marketing automation. It would be great to feel included, or even welcomed, and although Buffer does an outstanding job with customer service, the emails aren’t personal. Even a generic email welcoming me to the Buffer list would be better than the blog roll.

Buffer may want to personalize their email marketing. Although they are growing their lists, their information should be targeted – and since they are using MailChimp, they should consider using the various email marketing communication tools they offer.

Delight

Buffer has some of the most engaging customers and fans in the internet marketing industry. Buffer provides remarkable content to all users – visitors, leads, or existing customers. This next stage of inbound marketing is all about engaging with and (hopefully) keeping their future and current customers happy.

Social Media also exists in optimization and attraction, thus this section will examine attraction and engagement.

Social Media

Bar none, Buffer is one of the best companies on social media. The responded to most tweets within 20 minutes, and during the week of monitoring, out of 1,798 inquiries, they responded to 90% in less than 6 hours.

The chart below provides a visual of how Buffer’s shares are split concerning a search engine’s point of view. The center represents the first page, and each segment beyond it are pages linked from that page (in this case – the blog roll page). Segments beyond that are linked to their parent. Pages with no shares are not included or shown (click on the image to zoom in and to read the text).

As you can tell from the starburst, Buffer’s top 10 posts in organic traffic are highly represented in social media engagement:

*Written during Audit

As you can see from the graph above, the center represents the blog homepage, and all articles off of the homepage. The graph below breaks down that same content by social network:

This data strongly correlates with the data that Buffer produces each month on the open blog. According to Buffer:

As the data shows, followers grew on all social networks this month, but their interactions were down for every network except Facebook. Even though Facebook grew this month, Buffer’s Facebook engagement still needs some help – their engagement rate is hovering around 2%.

The articles on Buffer that are most widely read are ones that ask for engagement:

In this case, I’d encourage Buffer to examine their posts, and they are likely to find that the posts with larger images / posts asking for involvement did better than others.

Overall, Buffer does extremely well with social media. One minor suggestion, though: Buffer should consider adding social icons to the top of articles on the blog as well.

Engaging Content

Besides social shares and traffic (which have already been measured), the next best tell for engaging content is commenting. On average, Buffer receives 35 comments for each post.

As you can see from the graph below, Buffer receives quite a few comments for each post (40 in sample):

The only recommendation I’d have for Buffer is to add a comment counter to the blog page. With such high interaction levels, a user may just click the comment count to join the discussion.

Summary

If you read everything in this audit example, I am truly impressed. If you cheated and jumped straight down to this section, I’ll forgive you. Either way, this section contains a summary of my most important observations and recommendations for Buffer’s Inbound Marketing Audit.

Optimization / Attract

Buffer’s site is built with 11 subdomains for a total of 1,461 pages across all subdomains. Personally, based on research and countless other stories, I’d encourage Buffer to think about subdirectories. [More information]
It is possible that Buffer can’t build on the root domain in order to be “white hat” with their links. Have you ever clicked on the Buffer icon button when sharing an article and noticed what URL is in the pop up? [More information]
Buffer should remove the disallow of all directories on their new blog overflow.bufferapp.com. In its current condition, the site cannot be indexed. [More information]
Since Buffer uses quite a bit of video, it may help to compile a comprehensive list of the site’s indexable pages with videos on the page (i.e., pages that are substantive enough to warrant video inclusion) and create a video sitemap using that list. [More information]
Buffer received a 55/100 on Page Speed Insights via Google, and 80/100 on Pingdom. Buffer’s homepage has room for improvement. Another concern is that their blogs have over 100+ requests with a page speed of 3 seconds. [More information]
When examining Buffer’s pages, I found multiple pages with 20-40 errors per page. Many of the issues are easy fixes: open tags, stray elements, attribute errors, etc… [More information]
Out of 1,449 unique HTML/Text URLs, 893 do not include a meta description, and 92 of those descriptions are duplicates. [More information]
Yoast and rel canonical for paginated WordPress URLs are technically incorrectly implemented. Yoast’s plugin applies a rel-canonical meta tag to itself. This is a known misuse of the rel-canonical tag. [More information]
Each page that is indexed should have unique content and a unique title that effectively summarizes the content for users and search engines. Out of the 1400+ pages on Buffer’s website, only 90 pages do not have unique titles. [More information]
Buffer has done a great job crafting their URLs. Their URLs are at an average length of 50 characters (including bufferapp.com/ or subdomain.bufferapp.com). [More information]
Buffer has hundreds of images, and around 81% of the images take advantage of the alt text. [More information]
Buffer implements redirects correctly throughout, but there is cause for some concern on the homepage. There are technical reasons for implementing a 302 redirect to https, but since this redirect will stay around, I’d suggest Buffer change the redirect to a 301. [More information]
76% of all Buffer’s pages have no backlinks. Of the 24% that do have links, the homepage accounts for almost 23% of those links. [More information]
With a trust ratio of .842 on bufferapp and .660 on blog.bufferapp, Buffer really needs to examine their backlink profile. (It would be nice to see the ratio greater than 1.) [More information]

Conversion / Leads

Buffer has an offer and at least three calls to action on the homepage – asking people to log in with Twitter, Facebook, and Linkedin. The only shortcoming on this page is the CTA’s don’t lead to a landing page. [More information]
At the bottom of every blog page (open blog included), Buffer clusters its CTA’s, so much so that the sidebar CTA is hidden by the pop up footer CTA. [More information]
On Buffer’s site, the only landing page (either through CTA, Social Media, etc…) is the home page. Buffer’s referral traffic lands on a blog page – which contains a CTA – which directs them back to the homepage [More information]

Close / Customers

Due to the lack of forms, the email is not personalized. Buffer has no way to distinguish who I am, what my name is, or where I am at on my buyer’s journey. [More information]

Delight

Buffer responded to most tweets within 20 minutes, and during the week of monitoring, out of 1,798 inquiries, they responded to 90% in less than 6 hours. [More information]
Besides social shares and traffic (which have already been measured), the next best tell for engaging content is commenting. On average, Buffer receives 35 comments for each post. [More information]

Final Thought

As important as SEO is, it doesn’t exist in a silo. I hope this audit integrates SEO and inbound marketing to create a cohesive strategy and plan for Buffer in the future.

After the entire audit, I plan to solely use Buffer as a way to publish on social media. You won’t find a better company. I hope this helps them succeed.

This audit has proven two things:

No company is perfect. Even the best companies have to constantly monitor their internet marketing. This should be an encouragement for all – we all miss things in our inbound / SEO.
Buffer comes pretty close to being perfect…

Best of luck, Buffer.

What Do You Think?

I would love to hear your thoughts, suggestions, and questions in the comments below. Do you have any audit-related tips you’d like to share? How would you help Buffer with inbound?

How I'd Help Wistia Improve Their SEO: An In-Depth Audit

Thu, 10 Nov 2011 10:18:00 +0000

I love SEO and inbound marketing, which means I spend too much time looking at marketing sites: Inbound.org, Hubspot, Moz, etc… A few days ago, I was reading up on Video Optimization on Wistia, and after looking through their site, I wanted to offer a few thoughts.

I’m sure Wistia’s people have already analyzed this design from a funnel and business perspective, but maybe no one has analyzed the SEO ramifications?

Perhaps, I’ll get a free T-shirt.

Wistia certainly has thought about SEO from a product perspective; their video embeds for SEO are top notch (seriously, I could not be more of an advocate for Wistia Video Hosting after spending weeks auditing – I’d never host my videos anywhere else – and that’s saying a lot – so Get Started Here [not an affiliate link] or check out this great article via Phil Nottingham at Distilled).

Hi friend! Please note the date on this post. I wrote this post years ago, so while some of it might be helpful, most of it is out of date. I've contemplated deleting the post, but I like to keep it around.

If you are curious, I not only got a shirt, I got 3. And a great job -- just not at Wistia, but at Help Scout. 4+ years later, I'm still happily there.

But what about using SEO to market their own website? They currently rank #1 for “video hosting,” but given the right tools, could they increase their organic traffic and CRO? As I looked through their employees, I didn’t see an SEO Analyst or Inbound Manager who is specifically devoted to this issue, so I thought I could offer some fresh perspective.

The last major redesign for Wistia happened during the first week of April in 2013. Since a year old website is ancient in web standards, and a redesign may be in the future, I thought now would be a great time to chime in.

In this post, I’ve put together a detailed, technical SEO Audit for Wistia. My audit is widespread and comprehensive, covering a wide range of SEO topics (e.g., article listings, structured data, duplicate content, backlink analysis, etc.). But before the audit, let’s cover the basics…

Who is Wistia?

Wistia Inc. is an internet video hosting and analytics company that was founded in April 2006 by Brown University graduates Chris Savage and Brendan Schwartz.

Disclaimer

Before I jump into the technical details of the audit, it’s important to note that I have no affiliation with Wistia. As a result, I don’t have access to any of the site’s analytics or webmaster tools accounts, and I don’t have access to the site’s content management system (CMS).

I also did not want to hammer their site’s bandwidth using multiple crawlers, thus if my data is inaccurate, I blame my sincerity. If this were a typical audit, I’d run through their analytics and start the assessment from there, but for this audit, I am relying exclusively on third-party data.

So… if I make completely inaccurate observations, I blame the third party, ie (Searchmetrics, Ahrefs, SEMrush, Moz, etc…). You may have noticed similar audits on Webgnomes, Moz, and Quicksprout. I have no problem admitting I used their layouts as a template and even linking to them, but of course, all of the data here is relevant to Wistia. With that disclaimer out of the way, let’s roll.

Since this post will be extremely long – here is a detailed navigation to help you understand where you are throughout this audit.

Organic Search Visibility
Index-ability
Accessibility
On Page Factors
Off Page Factors
- Backlinks
- Social Engagement
50 Grove
Inbound Marketing
Summary

Organic Search

In order to set the framework for the entire audit, it’s important to analyze the site’s organic performance. When looking at Wistia’s organic traffic, there are 3 important questions:

Is the organic traffic growing?
Has the site been penalized?
What is the percentage of organic traffic?

To determine Wistia’s SEO performance, I used two tools: Searchmetrics Suit, and one of my new favorites, Similar Web. Again, finding accurate third party data is incredibly difficult, but after using Google Analytics on a network of sites, I’ve found that these two programs provide the most accuracy in capturing data.

As this graph shows, Wisita’s traffic is growing. It is estimated that the monthly traffic to Wistia.com is 4.8 million, with a high time-on-site average and an extremely low bounce rate. Things are looking pretty good for Wistia, but let’s take a closer look.

Based on the data above, we are looking at all traffic (referrals, organic, paid, etc…) but we are really interested in organic for this audit. To get a better picture, we need to run some numbers and exclusions.

First Exclusion: Wistia’s users accounts (user.wistia.com) are each included in total sessions (more on that in the analysis below). This explains why the bounce rate is extremely low and the time on site is extremely high. When you sign up for an account at Wistia.com, you receive an account on a subdomain of wistia (elioverbey.wistia.com). Thus, users are repeating visits to their subdomain to login (lowering bounce) and upload videos (increasing time on site). This first exclusion removes over 51% of all visits. (As you can tell by the graph below, I am being generous by not excluding fast.wistia and secure.wistia).

This exclusion removes 2,457,600 visits, and we are left with only 48.98% of visitors actually visiting a Wistia front-facing site (http://wistia.com, secure.wistia.com, embed.wistia.com, etc…).

Second Exclusion: According to the third party data, only 10.56% of all visits are organic.

If you do the math, we are left with roughly 248,269 organic visitors per month to Wistia.com.

Why exclude subomain / user accounts?

In this scenario, if I would have just taken 10% (organic) of 4.8M (all visits), I’d have inaccurate data. Since organic only applies to those coming from search, I know that users’ subdomains should not be included in organic data because their robots.txt makes it difficult for them to appear in search (more on that below).

Thus, user subdomains were excluded – leaving us with an average of 250,000 organic visitors per month.

In any regard, Wistia should always be looking to improve their 10% organic search. It could be said that most of Wistia’s traffic is branded, and so, much of their traffic is direct. But, with Wistia only being 8 years old, the branded traffic is young as well – thus it’d be nice to see organic hovering around 25%.

One very positive thing to note: in the last 6 months, Wistia does not seem to have been hurt by any penalties from Google. Again, you cannot trust the graph above because it contains all traffic and not just organic, but other graphs from Search Metrics show a similar pattern:

There does seem to be a drop in keyword rankings from January to end of March, but luckily, there were no major updates released in that time period (just anonymous soft Panda updates). The cause could be that one of their major keywords “video hosting” seems to fluctuate quite a bit.

Index-ability

To find how many of Wistia’s pages are actually out there, I crawled their entire site using Screaming Frog’s SEO Spider and Xenu Link Slueth. The crawlers found thousands of URLs, and after multiple crawls, I successfully crawled 5,685 unique pages. Of that, 2,325 URLs were on the Wistia root domain [http://wisita.com/].

Finally, there were 1,243 of those pages that were html/text. It gets complicated with Wistia because they host their customers’ videos on subdomains, so that a user who uses Wistia gets a url similar to: user.wisita.com. Those subdomain URLs were not included in the crawl.

Based on my two previous site crawls, I was able to successfully crawl 5,685 unique pages. But, according to the search operator “site:”, Google has indexed quite a bit more. 31,400 to be exact.

According to this, Google has been able to crawl 6 times more than what I was able to crawl. But we need to keep in mind that the high result number is known to be inaccurate. On the other hand, it shows that Wistia needs to start “no-indexing” some of their content because when you navigate to page 54 on Google, you are greeted with this warning:

This warning suggests duplicate content issues. If you repeat the search with omitted results, you will notice 4 things:

First, If you navigate to the last page, there are only around 6,000 pages indexed – which is much closer to my crawl estimate.

Second, There are broken links that should be absolved. There are not many (only 119) links that should be cleaned up.

Third, There are duplicates that should be de-indexed. Wistia has hundreds of pages that are indexed that are basically copies of each other. Take for example: wistia.com/blog?page=38 (Below).

These blog roll pages should be noindex, follow. Wistia has done a great job of implementing rel=prev/next, and according to the Official Google Webmaster Blog, they are meeting the standards:

When you implement rel=”next” and rel=”prev” on component pages of a series, we’ll then consolidate the indexing properties from the component pages and attempt to direct users to the most relevant page/URL. This is typically the first page. There’s no need to mark page 2 to n of the series with noindex unless you’re sure that you don’t want those pages to appear in search results.

Based on that information, Wistia has marked up their information correctly, but in this scenario, they “are sure” they do not want these pages crawled. Using a robots meta tag here to “noindex, follow” the pages would be encouraged since the rel/prev is still being indexed. In this case, rel/prev is preventing a duplication issue, but it is not solving the indexing issue.

Fourth, Wistia’s product is built on a subdomain model. The reason for so many URLs is because each new user of Wistia gets a subdomain on the URL. I tried out the sign-up process:

They have correctly added a disallow in the robots.txt to each user’s profile (more on that in a moment), but to better enforce the issue, Wistia should consider adding a robots meta tag that noindexes all subdomains at the page level. They could keep the “follow” aspect, and could benefit from links back to the subdomains. Here is the current setup:

The subdomains are being indexed, but Wistia is not receiving any credit for them. Another concern is that the crawl budget is being affected. While it’s good that most of Wistia’s content is being crawled, the subdomains are taking away from the bots crawling Wistia’s main pages. They are giving a lot of room away to subdomainers who are disallowed anyways.

Robots.txt

A robots.txt file is used to restrict search engines from accessing specific sections of a site. Here is a copy of Wistia’s root domain Robots.txt file:

The first record in robots.txt file explains how to write and use the robots file. The next entry is telling the webmaster how to use the file. The third record politely asks all bots (bots used by all search engines are identified with the aterisk) to not crawl the old blog /oldblog/.

Finally, the remaining records simply specify the locations for Wistia’s XML Sitemaps and are technically accurate. In any case, I would recommend moving the sitemaps entries to the top of the robots.txt

There are a few improvements that Wistia could make in the construction of their robots.txt.

Robots.txt is ONLY a way in which to tell a search engine bot which pages it’s allowed to spider, to “see”, and which pages it cannot “see.”

This is not telling search engines if they can include them in their index or not. There is a huge difference. For example, Google will sometimes list URLs that it’s not allowed to spider, because it’s blocked by robots.txt, yet it still shows up in their index.

Wistia should update their robots file because the /oldblog/ contains links. The problem is, right now, the links towards that blog don’t count towards the credibility of Wistia because they’re blocked, but Google displays the page anyway in the search results; It looks something like this:

In order to avoid “A description for this result is not available because of this site’s robots.txt – learn more,” There are two possible solutions for this problem:

First, Wistia could add a <meta name=”robots” content=”noindex, follow”> tag to those /oldblog/ pages, as it would prevent the pages from being indexed and would allow the link “juice” to flow through to the returned posts and pages.

Second, Wistia could 301 redirect all of the /oldblog/ pages to their corresponding new pages. Thus, an article like “Improved Stats Exports In Wistia” on the old blog “http://wistia.com/oldblog/stats-export-updates/” would 301 redirect to the new location at http://wistia.com/blog/stats-export-updates.

These solutions will pass links and will help Wistia from creating duplicate content. A quick reminder: telling Google that it cannot see something in the robots.txt file does not mean that they cannot index it; it tells the engines not to crawl the given URL, but that they may keep the page in the index and display it in the results. There are better solutions for blocking robots than in the robots.txt file.

With the page level solutions mentioned above, the new robots.txt file for the Wisita root domain would look like:

As you can see, I’ve removed the comments and the /oldblog/ from the original.

Robots.txt for Wistia Users

Each Wistia user receives a setup similar to the following: company.wistia.com. Wistia uses the subdomain to build out the users’ hosting space. They’ve equipped each subdomain with the following robots.txt:

This setup is technically correct, and I would need a deeper understanding of the Wistia platform before suggesting prudent changes. However, here is what I can tell with my limited access:

The subdomains are being indexed, which is not technically healthy for Wistia’s site. Again, blocking the robots.txt file doesn’t keep Google from indexing the content, it just keeps the search engines from showing what’s inside.

A solution would be to add a <meta name=”robots” content=”noindex, follow”> tag to all user subdomains.

If Wistia switched from blocking pages at the robots.txt pages to blocking pages at the page level, they could benefit from embedded links pointing at their subdomains. But once again, I would need a fuller understanding of the platform and limitations before suggesting any changes.

Robots Meta Tag

Each page on a site can use a robots meta tag to tell search engine crawlers if they are allowed to index that page and follow its links.

Very few of Wistia’s pages have a meta robots tag – which is fine – they are unnecessary unless they are trying to specify a hidden page. In the case mentioned above, it would benefit Wistia to include a “noindex” robots meta tag on a per page basis, rather than blocking an entire directory.

Finally, Wistia’s SEO could benefit by adding a “noindex,follow” (perhaps nofollow) tag to their users’ profiles. The image below shows what Google does with users’ profiles.

The problem here is that because users’ profiles are displayed, they take up Wistia’s crawl budget. Wistia’s main content should be what consumes the crawl budget, not individual user profiles. So, it would be beneficial to “noindex,follow” all user profiles.

XML Sitemaps

What are sitemaps?

Sitemaps are an easy way for webmasters to inform search engines about pages on their sites that are available for crawling. In its simplest form, a Sitemap is an XML file that lists URLs for a site along with additional metadata about each URL… Sitemaps.org

I identified two Sitemap listings in Wistia’s robots.txt file:

The first listing points to the site’s Web Sitemap, which should include URLs for every page the site wants search engines to crawl and index.

The second listing points to the site’s Video Sitemap, which should include URLs for pages that include video listings. The video sitemap may also earn Wistia a click through boost by giving them a picture in the SERPs. Here are the contents to Wistia’s first sitemap:

The video sitemap is implemented correctly:

It makes it clear to Google what their content is.
They have the opportunity to provide a range of details through schema.
Additional presence on video.google.com search.
A picture thumbnail, which is a pretty great call to action in the SERPs.

It is interesting to note that Wistia does not include all posts with videos in their sitemap. However, since most of their videos earn a picture thumbnail, it may be beneficial for Wistia to add more videos to their sitemap.

I would recommend compiling a comprehensive list of the site’s indexable pages with videos on the page (i.e., pages that are substantive enough to warrant video inclusion) and creating the video sitemap using that list.

Wistia’s other sitemap is a little more disappointing.

This sitemap is not populated – and it’s another video sitemap – meaning Wistia does not have a sitemap for text posts. I would recommend compiling a comprehensive list of the site’s indexable posts and pages and creating the sitemap using that list.

One last issue with the sitemaps: they are located on a subdomain of wistia.com. Moving the sitemaps from app.wistia.com to wistia.com/sitemaps/ may help with indexing, since all the articles inside the sitemap are hosted on the root domain.

Accessibility

Performance

According to Google Webmaster tools:

You may have heard that here at Google we’re obsessed with speed, in our products and on the web. As part of that effort, today we’re including a new signal in our search ranking algorithms: site speed. Site speed reflects how quickly a website responds to web requests… While site speed is a new signal, it doesn’t carry as much weight as the relevance of a page.

Though it’s tempting to dismiss site speed as an important SEO ranking factor, if Google says it matters (even a small percentage), it matters.

Wistia’s site speed is decent, but Wistia could make a few tweaks to make their site much more efficient:

One of the things you’ll notice is that the site is loading too many external files. Reducing the number of files needed to load the site, thereby reducing the number of HTTP requests, will make the site load more quickly. Currently, Wistia is making 94 HTTP requests – far too many requests.

There’s usually three parts to fixing this:

Reduce the number of JavaScript files
Reduce the number of CSS files
Reduce the number of images

The Time to First Byte is extremely quick, but the page load is somewhat slow – anywhere from 2.88 seconds to 3.93 seconds (Pingdom). They received a 63/100 on Page Speed Insights via Google, and 72/100 on Pingdom. Wistia’ site has room for improvement.

The site could be improved with the following:

Eliminate render-blocking JavaScript and CSS
Minimize HTTP request – Wistia’s pages will load more quickly with fewer requests. Minimizing these requests involves reducing the number of files that have to be loaded, such as javascript, css, and images.
Combine Javascript and CSS into external files with links from the header. This allows the external page to be cached so that it’ll load faster.
Optimize images for the web.
Implement server side caching – This creates a static html page for a URL so that the dynamic sites don’t have to reload / be recreated each time the URL is requested.
Use a CDN – such as Amazon. The CDN will allow users to download information more quickly.
Load Javascript asynchronously.
Finally, use 301’s only when necessary. A 301 forces a new URL, which takes a longer time to load.

Wistia also has around 200 broken links – which is not a lot – but fixing those would cut down on 404s and unnecessary HTTP requests.

Finally, Wistia loads some of their CSS in the header – over 500 lines of CSS on each page. This occurs on the homepage and other pages throughout the site. In order to increase page speed and be more semantic, Wistia should move their styles to the main CSS style sheet. Right now, Wistia’s website suffers from 120+ CSS errors.

Site Architecture

Currently, Wistia’s average page level is 5.4 (0 being the homepage). At this point, Wistia should:

Start to trim back their content that doesn’t produce results
Reduce the number of content silos on their pages by creating a flatter, better connected architecture

First, some SEOs are under the impression that all content is good, meaning that content should never be deleted. In fact, even HubSpot has an article entitled 5 Reasons Why Deleting Your Blog Posts Is Stupid (perhaps the article is outdated – and should be deleted). I tend to lean the other way.

Content is like a grapevine: In order to produce the best fruit possible, you really need to prune. I’d suggest that Wistia sit down with Google Analytics, Backink profiles, etc. and decide which pages are brining in traffic, shares, etc, and which articles are not. Based on the results, Wistia can create better content that engages and prevents keyword cannibalization.

Wistia can cut back on content in two ways: meta robots “noindex” or they can delete the content. Content that is necessary for a positive user experience but should not be included in the SERPs should be deindexed. _For example, _one of the first places Wistia should consider deindexing is /community/. This section of the site does not contain any CTAs – the rankings are low – and the architecture is deep. Wistia should not cut this content because it is useful for the user, but not for the search engine. Deindexing the content takes care of the problem.

On the other hand, there are posts that have aged poorly. If they are not driving traffic, increasing shares, ranking, or converting, Wistia should delete them. In_ deleting content, Wistia should 301 duplicate content and let the rest of the content 404. _

Secondly, Wistia should create a flatter site architecture. If Wisita can contain most of their content within 3 levels http://wistia.com/blog/category/article (root counts as 0), they will be able to increase indexation and increase long tail results. This will also prevent vertical silos (Rand has a great article on content silos).

Right now, Wistia exists in 4 main silos: /blog, /learning, /community and /doc. The Wistia homepage links to a silo (ie. blog), and silos link to posts. This creates lots of information, but links are flowing down, not around.

In reworking their site architecture, these vertical silos will disappear.

Usability

Wistia has a few layout issues that reduce the website’s usability and effectiveness.

First, noticeably missing from the Wistia website is a search form. The only way to find something on the Wistia website (outside of the navigation) is to head over to Google and search “Wistia + “. Adding this search functionality will help users, and it will also help Wistia know what the user is searching for in their site page analytics.

Next, Wistia’s site is not mobile responsive. Based on the analytics, Wistia has 50% of their users visiting the site on mobile. On iOS, sometimes Wistia’s entire site loads, but at other times, users will see this:

At some breakpoints, Wistia’s site seems to be in the middle of transition between “almost mobile ready” and fixed width:

Usability needs improvement on mobile, and according to Matt Cutts, “There is NO SEO disadvantage to using responsive design, only advantages.” I’m sure this usability problem will be fixed in the next iteration of Wistia’s website.

Click Depth

Click Depth or crawl depth is the number of times a website visitor must click on a link from the root domain in order to get to the desired page. The root domain will have a link depth of zero – since it takes zero clicks to get to the root domain on a webpage.

Search Engines are like humans – they do not want to waste time trying to find pages hidden deep into a site’s architecture. Pages that are available in one click – thus one page from the homepage – are deemed more important by search engines than those that are nearly hidden.

As the graph below shows, Wistia is doing a good job of keeping the required clicks minimal. What will decrease click depth even more, though, would be to remove silos.

####

On Page Ranking Factors

In this section, I’ll examine Wistia’s on-page ranking factors (e.g., Titles,URLs, markup, images, content, etc.) that influence the site’s search engine rankings.

HTML Markup

Semantic HTML markup, in its most basic elements, is a machine readable language that is used to tell the browser how to display the text or graphics in the document. Search engines crawl over your pages and semantic markup can identify your pages’ most important information. Semantic HTML markup is a way of indicating the meaning of web content to the search engine.

When examining Wistia’s pages, I found multiple pages with 10-30 errors per page. The most concerning was the empty heading tags littered throughout the pages. Although these aren’t extremely critical issues, I’d recommend cleaning up as many errors as possible (empty headings, duplicate IDs, attributes, etc).

Structured Data

Schema.org mircodata is an extra set of HTML tags that you can add to your HTML elements to let the search engines know specifically what you are talking about. For example, for the Wistia blog post, they would tag the title as a “headline,” the author as an “author” (with authorship), and the date the blog post was published as the “publishedDate.” Now when Google indexes Wistia’s blog, it knows specifically what the title of an article is, who the author is, and when the post was published.

Unfortunately, Wistia does not define structured data. To make these articles more readable for search engines, I’d recommend adding the article markup. Wistia is already doing a great job with Video Markups, and the posts should be an easy next step.

Wistia fails to add authorship to most of its articles. Though a few of Wistia’s articles include authorship, such as the one below, including authorship in all articles would be greatly beneficial.

In order to add authorship, here’s what you can do: create a Google+ profile for each of the site’s authors. Then, link from Google + profile back to the blog, and vice versa.

Be Careful with Schema.org

One caution: I do not recommend schema.org for all HTML elements. For example, consider the search term “OneRepublic”:

Meta Descriptions

Meta Descriptions do not influence the ranking factor of a site’s page, but they do influence the CTR (click-through-rate).

These short paragraphs (normally 155 words) are a website’s opportunity to advertise content to searchers and to let them know whether the given page contains the information they’re looking for.

Out of 1,243 unique HTML/Text URLs on Wisita’s root domain (more on that in a minute) only 76 of Wistia’s pages use a meta description, and 19 of those descriptions are duplicates. What this means is that in almost all of the searches, the meta descriptions are automatically generated.

2 Options to Fix

There is no single answer for this problem.

The first answer would be adding a unique meta description for each of the page’s content.

The second answer is a bit more complex. If a page is only targeting between one and three search terms, then by all means, write a meta description that targets those key words. But if the page is targeting long-tail searches, it may be easier to let the search engines pull in the meta descriptions themselves.

In some cases, search engines will overrule your meta description anyways, but you cannot always rely on that. In that regard, it would be advantageous for Wistia to add meta descriptions for pages that target only a few keywords and leave the other long-tail pages open for variance.

One last issue to be aware of when thinking about meta descriptions:

Social sharing sites will commonly pull the meta description when sharing.

Without the meta description, social sharing sites will just use the first text they can find. As you can see above, the social share pulled the title twice, the date, and then the first words from the opening paragraph. Not using a meta description may create a bad user experience for social sharing.

Head Tags

rel=publisher

Wistia has a great looking Google+ page, and they need to take advantage of rel=publisher. This tag has seen a much smaller adoption rate compared to Rel=Author, and its name may cause the confusion. The publisher tag can be used by any business or brand website that has a corresponding Google+ page.

Publisher connects your Google + page to your website, and that results in your Google + page being displayed in search results whenever anyone is searching for your brand name. For example:

See the rich snippet on the right? This is what Wistia is missing. Currently, a search for Wistia leaves room to be desired:

The red box here indicates where Wistia’s publisher mark-up could go. Wistia simply needs to link from their home page to their Google Plus page using rel=”publisher”.

rel=canonical

Wistia does a fine job of inputting rel=canonical in most cases. On many subpages, blog rolls, etc… Wistia uses rel=prev, rel=next, which is actually a common substitute for canonical (according to webmaster tools). On http://wistia.com/community there does seem to be a small error in rel=canonical. In the markup, Wistia has the canonical page as the first page. According to Google Webmaster Tools:

Specifying a rel=canonical from page 2 (or any later page) to page 1 is not correct use of rel=canonical, as these are not duplicate pages.

Based on this information, Wistia has the rel=canonical set up incorrectly. All community subpages point back to /community. According to webmaster tools, Google actually recommends that you rel=canonical to a “View All” page. The current set up is technically incorrect because page 2 is not a duplicated of /community.

In order to fix this, I’d recommend that Wistia implement rel=prev / rel=next like they do in other subpages (and remove rel=canonical), or that they create a “view all” page and rel=canonical the pages that include all discussions from community. This may help decrease the indexing of the community pages.

Lastly, not all subpages include canonical or rel=”prev/next”. I’d suggest adding those tags to help the search engines distinguish the paginated series.

rel=prev / next

Google asks that webmasters add rel=”next” and rel=”prev” to paginated archives, so that Google can distinguish them as a series and send users to the most relevant pages.

As far as I can tell, Wistia implements this correctly (ie, the blog).

Open Graph

The Open Graph protocol enables any web page on Wistia’s site – or all sites – to become a rich object. For instance, the Open Graph is used by Facebook to allow any web page to have the same functionality as any other object on Facebook. Currently, Wistia does not implement Open Graph protocol on their site.

If you look closely, you can easily notice that Wistia is not using the open graph – due to how the description is automatically generated by Facebook. At the very minimum, Wistia should include the following open graph tags:

Even if Wistia does not have the time to implement all forms of open graph, at the bare minimum, I’d recommend they include Open Graph (all social platforms can use it as a baseline, including Twitter). Adding Open Graph tags to their website won’t directly affect the on-page SEO, but it will influence the performance of Wistia’s links in social media.

Twitter Cards

Currently, Wistia is not implementing Twitter cards on their site:

Twitter cards are another micro markup – essentially rich snippets for Tweets. Now, as mentioned above, Twitter will actually fall back on the Open Graph tags used for FB if Wistia would implement them, but the Open Graph tags aren’t comprehensive.

There may be a reason that Wistia opted not to use Twitter cards, because they could easily implement them if they wanted to. Since they’ve seamlessly implemented video on Twitter, the other card formats would be easy to implement.

It is an easy fix. At the very minimum, Wistia could add:

I’d recommend Wistia looking into adding open graph to all of their pages to help with social engagement and sharing.

Titles

According to Cyrus Shepard at Moz:

Title tags—technically called title elements—define the title of a document. Title tags are often used on search engine results pages (SERPs) to display preview snippets for a given page, and are important both for SEO and social sharing.

Each page that is indexed should have unique content and a unique title that effectively summarizes the content for users and search engines. Out of the 1200+ pages on Wistia’s website, only 512 pages have unique titles.

On these pages, there are two options:

First, add a unique title that describes the pages.

Second, and probably most preferred, “noindex, follow” the pages that are not necessarily important (See the example below of titles and pages that are not necessary). This will remove them from the search engines but links will still pass equity.

Many of the site’s pages use titles that are not descriptive enough (Ie. “Sent from my iPhone,” “Lighting for a Webcam,” etc). These titles do not provide enough information about Wistia or their products , which results in a lower ranking and lower CTR for Wistia. Here is a sampling of the titles from Wistia:

Shooting for the Edit

Choosing a Microphone

Choosing Music for Your Video

Sent from my iPhone

Shooting with a Ring Light

Lighting for a Webcam

Making Better Help Videos

Based on the titles above, it seems as if Wistia is more of a video creation company rather than a video hosting company. I’d suggest researching the CTR and the company’s business needs and crafting articles and titles that are focused on those needs with the audience in mind.

For example, let’s look closely at this title: “Does Length Matter?“. Of course, if you know Wistia, you know that they are referring to video length. But let’s say you don’t know Wistia. What would you think? Wistia isn’t trying to be explicit or inappropriate, but you can see how a more descriptive title, such as, “Does Length Matter in Online Videos: Video Analytics Say Yes” makes much more sense. That title is a little long, 61 characters, but it would show up in Google as:

It cut the title off, but it does leave the viewer wondering what the analytics say. This title is much more descriptive and removes any click anxiety a user would face when coming upon the title “Does Length Matter.”

As you noticed above, Google will show as many characters that a 512 pixel display can show – thus, around 50-60 characters. The reason the number varies is because some characters are bigger than others (ie, you could fit a lot more “i”s than you could “w”s). _Wistia is not taking advantage of this character count. _Most of Wistia’s titles are at an average of 22 characters long, and very few of them reach 50 characters in the title. If Google gives you the opportunity to create a longer, more descriptive title, do it.

I’d suggest that Wistia go back and rewrite many of their titles. A longer, more specific title could result in higher click through rates, and ultimately, better rankings. Even if Wistia does not want to rewrite many of their titles, adding the branding would increase character count and would, at the very least, take back the length real estate they have lost up until this point.

Less than 50% of the site’s titles actually incorporate the brand name. To help promote brand awareness, I’d add Wistia at the end of each title (“Shooting for the Edit

Wistia”). Because Wistia is well recognized, including the brand name in the title element can increase search engine clicks.

URLs

Wistia has done a great job crafting their URLs. Their URLs are at an average length of 50 characters (including http://wistia).

The site’s URL structure is organized and clean. The blog section is under /blog/. The Documentation lives under /doc/, and community lives under /community/. Each URL also contains at least one keyword.

Wistia could, however, improve their URL structure by making it slightly more descriptive. Take this URL, for example:

http://wistia.com/blog/screenr/

Wistia could add one or two more words to the URL to describe what the page is about. The keyword Screenr is very vague, and Wistia could potentially help ranking and click through by describing the URL better.

Images

For the benefit of search engines, compliance (HTML), and the visually impaired – all images should have an ALT tag.

The ALT tag should accurately describe the image, and if possible, contain a keyword relevant to Wistia’s site (but only if the keyword is relevant to the image).

Additionally, image file names should be descriptive – not containing random numbers or queries. The file name should describe the image and, once again, use a keyword if possible.

If Wistia ensures that images follow these two rules, they will increase the likelihood of referral traffic from image searches. Wistia has hundreds of images, but only about 8% of the images take advantage of the alt text. Wistia also has many images that are labeled with random numbers and strings:

There are hundreds of images like this on Wistia’s blog, and the ones above were taken from the most recent articles posted. I’d suggest adding alt tags and descriptive title tags to the images to increase organic image search, but to also be HTML compliant (they currently aren’t) and to help the visually impaired.

Transcripts

According to Wistia:

When a search engine is indexing your page, instead of seeing a video element with a few tags, transcripts allow the search engines to index every single word of your video.

Transcripts are an opportunity to tell Google what a video is about. Transcripts also generates a closed caption for the video, which allows hearing impaired users to watch the video and read the dialogue.

Unfortunately, some of Wistia’s videos do not appear to utilize transcripts, and what’s more, those that do have transcripts do not accurately match the video. I would recommend adding transcripts to improve the searchability and usability of Wistia’s videos.

Content Duplication

Wistia is doing a great job keeping their content unique and fresh. They do, however, need to be careful with canonicalization (only point back to a view all page in archives) and prev/next.

For example, Wistia does need to do something about their users’ profiles, ie. http://wistia.com/community/users/. Their users’ profiles should be deindexed. They risk spammy content and content duplication, as seen below:

Wistia also doesn’t have a major issue yet, but they might want to remove the old WordPress /oldblog (or bring it in house since it might be a test sight):

Off-Site Ranking Factors

This section will cover some of the most influential factors in ranking: backlinks and social engagement.

Backlinks

When analyzing backlinks, you should consider: Total number of links, link quality, number of unique domains, fresh/incoming links, and anchor text usage.

To start, I first recorded how many backlinks Wistia had received over the past few months (this is not including 500 unique domains that point to www.wistia.com). The charts below show the growth of Wistia’s backlink profile over the course of one year.

After comparing the backlink analysis above to Majestic SEO and Moz, there was not much of a difference. Wistia was consistently growing their unique domain backlinks until April, and has started a recent growth.

To see how Wistia’s backlink growth matches up with their competition, I compared Wistia’s growth with 4 of their competitors over the past 5 years: Vzaar, Viddler, Viewbix, and Sproutvideo (notable, Vimeo was not included. Vimeo is on a much larger scale than Wistia at this point). *Wistia is a startup, but they’ve been around the past years.

One would hope that Wistia would be the yellow line, but they are not. Currently, Wistia is being outpaced by Viddler (Viddler provides online video hosting and custom solutions for businesses along with industry leading support and professional services). Viddler is only one year older than Wistia – Viddler was founded in 2005 and Wistia was founded in 2006 – which means that there should not be such a significant difference in the two companies’ backlink growth.

So that you can actually see Wistia’s cumulative backlink growth, I removed Viddler and focused on only the last 2 years (cumulative):

These charts shows 2 things:

Wistia’s backlinks are growing.
They have fierce competition.

If Wistia is not careful, they could potentially lose their #1 ranking for “video hosting” to Viddler. Almost every month, the competitors are acquiring more links from more unique referring domains. The backlink analysis shows the competitiveness of backlinks:

Subdomain Backlinks

As previously mentioned, Wistia sets their users up with a subdomain on their on site to host videos.

Thus, when anyone signs up, they can host their videos on their own subdomain hosted at Wistia.

What does this mean for Wistia?

They receive a lot of subdomain links from their users. In fact, out of the 17,000 backlinks, 13,000 of them point at subdomains of Wistia. There are three ways to look at this situation, but they all come with one major caution:

Pro: The backlinks on the subdomains are actually helping Wistia. Many companies have started to place their blog on a subdomain, and it seems to be a standard. Supposedly, the search engines are treating subdomains more and more as just portions of the main website, and in a way, links are passing credibility.
Neutral: Since the backlinks are not in a subdirectory and are in a subdomain, Google does not pass link credibility. They see the subdomains as a separate site, and thus, the subdomain does not help the main site.
Con: Currently, all of Wistia’s subdomains are disallowed from their robots.txt. It would be better for Wistia to set up meta robots tags with “noindex,follow” (to keep link credibility) than to disallow in the robots.

Caution: Whether Wistia changes their robots disallow to meta robots or not, Wistia needs to be mindful of their backlink profile. Even though Wistia currently disallows their users’ subdomains, the users’ links are still tracked. If users have spammy content, the fact that they are hosted at wistia.com means that Wistia could be penalized.

Backlink Distribution

To monitor Wistia’s backlink distribution – the links dispersed among all of their pages – I used Moz’s Pro Tools to analyze the backlinks for each page on Wistia’s site.

Wistia has – according to Moz, Majestic and Ahrefs – around 17,000 referring domains. But, upon closer inspection – many of those links are pointing to subdomains of Wisita. There are only around 1,250 domains pointing to the root folder (http://wistia.com).

According to the data, 72% of all Wistia’s pages have no backlinks. Of the 28% (359 pages) that do have links, the homepage accounts for almost 55% of those links. If you remove the homepage, here is what happens:

The articles with the most links (excluding the homepage) are:

The rest of the pages contain less than 1% of all referring links. Moving forward, I recommend spending more time on in-depth / well researched articles, ie. Does Video Length Matter. Additionally, it would not hurt to conduct outreach on the articles to increase page shares.

Backlink Source Analysis

To analyze Wistia’s backlinks, I used Majestic SEO. Here is a plotting of their backlinks:

Correlating with the data above, only a handful of Wistia’s links that I checked out appeared spammy, and very few of the links over-optimized anchor text. With a trust ratio of .918, Wistia has links from mostly trustworthy sources. It would be nice, however, to see the ratio greater than 1.

One last caution: Wistia should be careful of others websites using links in headers and footers that repeat across the site, ie, this site. These sitewide links are heavily scrutinized by Google’s Penguin update, but they should be okay with just a few…

Anchor Text Distribution

After Penguin, Panda, etc – it’s important for Wistia to conduct an anchor text analysis. Over the years, there were many ways to manipulate rankings on Google, and links were one of them. Google has always been aware of this, and they are finally cracking down (in fact, anchor text misuse will trigger the Penguin filter quickly).

What Links Do They Need to Examine?

Branded Links

Branded anchor text includes any link that is a brand name (Wistia), URL, or domain (wistia.com). Google wants to see these brand referrals. If you notice that your link profiles has higher branded links, that is okay. It is natural.

Partial Match (Money)

Compound, also commonly known as ‘partial match anchor text,’ is used when the link contains the brand or targeted term, in addition to other words that you aren’t targeting. For example, Wistia may target ‘Wistia Production’ as a partial match…

Exact Match (Money)

Exact match anchor text can be considered over optimization. This refers to the words in the links (anchor text, ie… Video Hosting), exactly matching the keywords that you are trying to rank for.

Generic

‘Zero match anchor text’ or white noise / generic uses calls to action to drive users to targeted pages. Example: click here, learn more, follow, read more, etc…

Naked URLs

A naked URL is simply the domain name or page you are linking to. For example: http://wistia.com

Image Links

Often overlooked, but are still powerful links, are links on images, which have keywords in the alternate text of the image.

This is a great rule of thumb to follow:

70% brand, URL, brand+keyword, and non-targeted anchor texts (Branded Links, White Noise, Naked URL, Titles, and Image)
25% partial, phrase, and broad match keyword anchor texts (Compound)
5% exact match anchor texts (Exact Match)

Below is the Anchor Text chart for Wistia…

Wistia does very well. If you examine the chart closely:

6% of Wistia links come from “Wistia,” which is a branded identity.
10% of searches include a partial match, etc…
84% include the other types of links: Naked, URL, etc..

Wistia has done a great job of keeping their links diverse. One of the best things Wistia did was leaving their video backlinks, from their embeds, empty. This is the reason for the massive amounts of empty anchor text links.

As a result, Wistia’s anchor text appears to be natural. As the company grows, the branded terms should grow as well, but again, that would be a natural progression.

Social Engagement

In May 2010, Google’s SEO Genius, Cutts, stated that the algorithm does not use social signals as a factor in search results. In December 2010, Cutts said social signals are a factor. More recently, in January of 2014, Cutts went back and said social media is not a signal in the algorithm.

Either way, we cannot ignore the fact that social media is a search engine of sorts. People on social media are researching, searching, interacting, and engaging – and at very high engagement rates. According to Statisticbrain, the number of Twitter search engine queries every day is over 2.1 billion. In any regard, a site’s success is largely dependent on social success and social engagement.

The chart below gives you a great visual of how Wistia’s shares are split concerning a search engine’s point of view. The center represents the first page, and each segment beyond it are pages linked from that page. Segments beyond that are linked to their parent. Pages with no shares are not included or shown.

As you can see from the graph above, the center represents the homepage, and all directories off of the homepage (/blog, /docs, /learning, etc). It’s great to see that the /blog (large orange portion) takes up a large portion of the shares. The graph below breaks down that same content by social network:

As you can tell, Wistia is putting their best foot forward on Twitter, but Facebook still needs some help. On Twitter, they’ve got hashtags (#wisitafest), embed videos, pictures, and links – and the best part – most of their links are to endorse content that is not always their own. In fact, most of the time, the links go to varying sources.

On Facebook, though, Wistia needs a little work:

The Wistia Facebook page is updated on a daily basis, but those updates have various characteristics that need improvement:

Wistia’s content is likable, but it’s not sharable. As a third party analyst, most of their content looks interesting. In the example above, their work space looks cool, but is it worth sharing? A better approach in the specific example above would be to encourage input, such as, “Practicing for the Wistiafest workshops: what are some things you want to see taught?” Very basic, but it’s _at least _asking for interaction.
Wistia doesn’t create interaction on the posts – their engagement rate is hovering around 2%.

Most disconcerting of all: 76% of all content has no FB engagement, and the lack of engagement is true for 51% of Tweets. Even worse, Google+ is non-existent. Here’s why: Wistia only promotes Facebook and Twitter sharing on their blogs – AND they are not very visible. (The red box below highlights this.)

To improve Wistia’s engagement on Facebook, Twitter, and Google+, I would :

Start using more prominent social sharing buttons. I would also add Google+ (since their network seems to have high link authority).
Use the data above to delete old content that receives no engagement.

50 Grove {#50grove}

50 Grove is a subdomain of Wistia – 50grove.wistia.com – that was created a few years back to help connect videos producers with clients. This site doesn’t appear to have been updated since early 2013, and accounts for quite a bit of Wistia’s search real-estate. With the uncertainty of how exactly Google sees subdomains, it might be in the best interest of Wistia to cut this subdomain.

Again – this is from the outside looking in, but with only 40 referring domains, and a lack of updating – it seems as if Wistia has moved on from 50 Grove. Even Wistia’s CEO had this to say about 50 grove:

We take no commission from 50 Grove and we aren’t a video production company. Our interests are solely in making it easier to connect businesses with producers.

I do think 50grove is a great idea, and I do understand how it relates to the business model, but with so many other products that Wistia offers, it may be a subdomain that is consuming more resources than it’s worth.

Inbound Marketing

After analyzing their site – over 1,200+ pages, I noticed that Wistia could also improve on their inbound marketing. With HubSpot right down the street from Wistia’s headquarters, Wistia may want to lean on their neighbor’s expertise (I’m sure Dharmesh has tickets to Inbound 2014).

As Wistia’s website currently stands, they are lacking lead generation tools. Right now, Wistia excels in blogging and social media content, but they are lacking in many conversion channels: calls-to actions, landing pages, forms, marketing automation, etc.

If Wistia does follow the practices above, their organic traffic will grow, but since only 2% of website visitors return, they’ve got to implement lead generation tools where they can build relationships, lead nurture, and then hopefully convert those visitors to customers. This audit will help attract users to the website, but it is only a stepping stone in creating customers.

Wistia should be adding compelling calls-to-action at the bottom of each posts, with helpful tools such as: ebooks, videos, freebies, etc. that will funnel a visitor to a landing page where Wistia can collect information.

Summary

As important as SEO is, it doesn’t exist in a silo. Right now, Wistia is doing well, but if they could make these tweaks and incorporate SEO into their business’s best practices and online marketing objectives, their gross revenue could soar.

Best of luck, Wistia.

Why I Switched to Mint Mobile (And Why My Parents Did Too)

The AT&T Situation

The Savings Are Real

Year 2, Year 3, and Beyond

The Experience

My Parents Switched Too

Should You Switch? Run the Numbers

Would I Recommend It?

How I Built a Personal Style Guide for AI

Why Bother?

What I Included

Foundational Tone Principles

Language Patterns & Vocabulary

Structural Preferences

Emoji & Expression Usage

Communication by Context

Things I Never Do

Platform-Specific Adaptations

How I Use It

The Full Guide

Core Identity

Foundational Tone Principles

Language Patterns & Vocabulary

Structural Preferences

Emoji & Expression Usage

Communication by Context

Things I Never Do

Platform Adaptations

Sample Messages

Key Mindset Reminders

Quick Reference Card

Tips If You Want to Build Your Own

22 Customer Retention Strategies that Work

What is customer retention?

How to improve your customer retention rate

22 Strategies to Increase Customer Retention

22 customer retention strategies that work

1. Stand for something

2. Use positive social proof

3. Invoke the inner ego

4. Position around the before and after

Tactics for customer marketing and education

5. Use the words they love to hear

6. Reduce pain points and friction

7. Capture your product’s momentum

8. Don’t just sell — educate

Delivering surprise reciprocity and delight

9. Don’t overspend to delight

10. Make use of surprise reciprocity

11. Don’t underestimate the power of ‘thank you’

12. Make it personal

13. Reduce effort before delivering delight

How great service retains customers

14. Accept that speed is secondary to quality

15. Know your customers

16. Choose the right platform

17. Make it a communal effort

18. Solve cause and effect

Designing sticky retention programs

19. Give loyal customers a head start

20. Make ideal customers VIPs

21. Reward your loyal advocates

22. Assign your customers positive labels

Retaining customers is a balancing act

Criteo vs Adroll for B2B

Background

Account Management

Campaign Performance

Revenue

Technology

Continuing with Adroll

SEO Audit of the Kissmetrics Blog: Google Panda Recovery

Who is KISSmetrics?

Why Audit KISSmetrics?

Disclaimer

Indexability / KISSmetrics Web Structure

Optimization / Attract

Search Visibility

Penalty Conclusion

Robots