Despite some finicky issues that I won’t go into here, Mixpanel is a fantastic piece of software.
Built for technology marketers, product managers, and other data-driven technology product types, Mixpanel started its life as an analytics platform. It differentiated itself from the ubiquitous (and free) Google Analytics by building its data model around events and actions instead of page views and clicks.
While the details of this data model are not necessary for this exercise, they basically mean that Mixpanel makes it easier and faster to track, measure, and analyze what people do (and don’t do) in web and mobile apps.
Once you’ve got Mixpanel installed and configured properly (you’ll need both a developer and a marketer who know what they’re doing), you can:
- Analyze funnels from your homepage to your on-boarding experience on down and measure how many people fall off at every step.
- Create fine-grained segments of users (like: people who signed up on X date, went through the on-boarding experience, but didn’t come back again).
- Send one-time or event-triggered emails or push notifications to any segment who completes (or doesn’t complete) key actions or stages of your funnel
- Track the percentage of new users you retain from any given date (both overall and for segments)
- Capture qualitative intelligence that quantitative metrics don’t offer with mobile surveys
You can also use Mixpanel’s tools to run A/B tests in your mobile apps (like a less-developed Optimizely) and send one-time, in-app broadcasts to your mobile users (like a more basic Intercom).
Mixpanel does a lot of different things. While the ways it does them have a range of quirks and limitations, the power of its tools (when you know what you’re doing) is remarkable.
The product’s utility has been good for Mixpanel’s business.
The company has been profitable since 2011, has a slew of top-notch customers, and recently raised a $33 million growth round from Andreessen-Horowitz.
You might assume that a company in Mixpanel’s position would be exceptional at telling its story. You might think that a team that builds a product that helps so many others optimize their funnels would do the same for themselves.
You would would think that after years of solid success, they’d at least have built a good website.
But if you thought any or all of those things, you’d be wrong.
Mixpanel’s site has some strengths, but overall, it sells a great product way short.
After reading the whole thing more than a few times, I’m still not sure I know what Mixpanel does or why I would need it, even though I know exactly what Mixpanel does because I’ve spent time talking to their customer success team and using the product, myself.
The whole site (top-to-bottom, left-to-right) needs an update and an overhaul, but today, I’ll focus on the homepage. It’s in the most dire need of help
Shall we begin?
My confusion starts at the first headline, then gets worse.
“Actions speak louder than page views” is clever. It’s a play on the oft-quoted wisdom that “actions speak louder than words.” Unfortunately, unless I am already familiar with the terminology of Google Analytics and Mixpanel, I have no idea what “actions speak louder than page views” means.
You could make the argument that this headline sparks curiosity, and encourages the reader to read more. Indeed, this would be a reasonable thing to argue if the rest of Mixpanel’s page fulfilled that curiosity and answered the questions it raised.
As we will see, it doesn’t. The concept of actions vs pageviews doesn’t even come back up again.
To put a fine point on it: Cleverness is risky in website copy. If you are a master of wit, you might be able to sneak in a double-entendre or a pun or play-on-words and still make your value proposition clear.
Most of us are not masters of wit, however, and should stick to the basics.
This headline needs attention, STAT.
Unfortunately, the problems with Mixpanel’s homepage are just beginning.
The subhead (“The most advanced analytics platform ever for mobile and web”) at least tells me that Mixpanel is an analytics tool. But that’s about the only thing going for it. Everything else is basically a copywriting mistake.
To be specific, the subhead:
- Focuses on the “what” instead of helping me understand the “why.” Ok, Mixpanel is an analytics platform. What does it do FOR ME?
- Speaks in broad generalizations. There are a lot of different kinds of analytics: Revenue analytics, marketing analytics, product analytics, customer support analytics, even IT service analytics. Telling me that you are an “analytics platform…for web and mobile” isn’t telling me much at all.
- Makes bold claims without any context to back them up. Mixpanel may or may not be the most advanced analytics platform ever, but big, bragging claims about your awesomeness are not convincing. They are even less convincing when made in a vacuum, without ample social proof to back them up.
- Sells the product short. Mixpanel is much more than an analytics platform. It is also a marketing automation and user on-boarding tool. It lets you survey your mobile users and A/B test entire features of your mobile apps. Boiling all of this functionality down to “analytics platform,” is not only selling Mixpanel short, it’s likely to create confusion for potential customers down the line.
The diagram looks like a silhouette of a preppy dude-bro with a bunch of crazy things coming out of his head. These crazy things (an envelope? an analog TV? some kind of growth graph?) also have things coming out of them.
This diagram may or may not have something to do with data, but I’m now confused.
If you’re going to lead with a big diagram (vs. a screenshot) make sure it illuminates the concepts, instead of piling on the confusion.
But the biggest problem with the top of this page is the navigation.
Specifically, there’s a slider at the top that I almost completely missed.
I’ve been to Mixpanel’s website a whole bunch of times, and I’ve never noticed that you can click an arrow button to see more.
Maybe it’s new. Maybe it’s just that hard to find. I only noticed it this time because I was marking up the screenshot you see above.
Other people have written eloquently about the conversion-killing effects of homepage sliders, so I won’t recap it here. The thing to know is: don’t use sliders!
And if you must use them, don’t make them almost impossible to notice.
But seriously, don’t use them.
People are far more likely to scroll to learn more then to click an arrow a bunch of times.
Ok, so after barely noticing the slider button and going back to click it, what did I find?
More copywriting in need of help.
Once you click the barely-noticeable slider arrow, Mixpanel’s product becomes a tiny bit clearer.
The best thing I can say for this part of the page is that the combination of the words “funnel” in the headline, “increase conversions” in the subhead, and the bar graph in the image below makes Mixpanel’s basic functions a bit more clear: this is something that I use to analyze my marketing and product funnels.
Everything else here needs a serious upgrade:
- The headline (“See where you lose customers with funnels”) is clunky. Sure, I want to understand where and why my funnel is leaking, but there are a lot more elegant, benefit-oriented ways to say this.
- The subhead (“use funnel analysis to increase conversions by showing where visitors get lost in any process”) contains a benefit, but that’s about all it’s got going for it. The rest uses jargon (“use funnel analysis”) and has a drawn-out sentence structure that feels like a run-on.
How about “Plug the gaps in your leaky funnel” for a headline? And for the subhead: “With Mixpanel’s ‘on-the-fly’ funnel builder, you’ll know where you’re losing customers—so you can make the changes necessary to stop the leaks.”
While this is only a quick first pass, my version is better because it focuses on the customer, not the product. Instead of centering on what the feature is (funnels), it talks about the core benefit you get (a less-leaky funnel).
My version would be meaningfully better with a credible metric attached to it, but it’s hard for me to know what is reasonable without talking to Mixpanel’s customers.
Let’s move on to the next slide.
The next feature section leads with a benefit. That’s good.
“Improve retention.” Finally, a headline that offers me a benefit. Unfortunately, it’s a bit generic.
If I am a marketer or product manager with a web or mobile app, I know I want to improve retention. It’s like telling anyone in business to “Make more money!” Yes, I want that. And yes, you can be much more specific and compelling.
I might change this to: “Build a product people want.”
While this may be over-promising (no analytics tool, no matter how awesome, will solve your product/market fit issues) it’s getting closer than “Improve retention.” If you learn that, two months after launching your shiny new feature, your retention rate hasn’t moved at all, you can start asking yourself what you’re doing wrong.
If my suggestion is going a bit too far, it could be shifted slightly to “Ensure you’re building the right product,” or something more polished to that effect.
The subhead (“See if changes to your product or marketing improves customer retention”) is a weak sentence. Along with the flagrant grammatical error (changes…improves), it just falls flat.
How about: “With Mixpanel, you’ll always understand the impact of changes you make to your product or marketing.”
Again: this is something I, product manager or marketer, care deeply about: measuring (and getting credit for) the impact I have on the product’s success.
As always: speak in clear language to people’s innate needs and desires, and your copy will win.
The next feature slide features more weak copy
“Get to know your customers.” Again, this is “what” that lacks a compelling “why.”
OK, sure, I might want to get to know my customers. But what benefit does that deliver to me, the target reader of this page?
Even though it’s going a bit far, I’d test something like “Make lasting connections with every customer.”
The subhead (“tie data to your customers so you can see who they are and what they have done in your app”) is confusing: How does one “tie data” to a human being?
Since I’m already coming to this page with a solid grasp of what Mixpanel does, I understand that this means you can attribute events and actions people take in your product to the people taking them. But without that pre-existing context, the idea of tying data to customers doesn’t make any sense.
How about “Use Mixpanel’s People Analytics to get to know your users and measure their activity inside your app.”
It’s not great, but it’s a lot less awkwardly phrased.
The next headline leads with a “what instead of “why.
The headline for Mixpanel’s Mobile Survey feature (“Answer tough questions with mobile surveys”) at least hints at a benefit: I can use surveys to get answers to qualitative questions that quantitative data can’t offer.
But this should be spelled out more clearly: “Get answers to critical questions your numerical data just can’t answer.”
Like the subhead in the previous slide, the subhead here (“Mobile surveys let you instantly ask you customers anything so you can finally answer the questions that have been difficult to measure”) is a cluster bomb of a sentence. It goes on and on and on.
How about something much simpler and more digestible: “With Mixpanel’s Mobile Surveys, you’ll stop wondering who your customers are and why they do the things they do. Just create a beautiful survey and press send.”
Still needs work, but you can see the improvement.
The next slide makes a bold promise. Too bold, methinks.
Look, I get it: people who work for companies that build good products tend to think the products they work on are the best things in the world.
If I worked on a product as solid as Mixpanel, I’d be proud of it, too.
But when all a feature can do is broadcast a basic image and some text to mobile users, it is not “The most advanced way to communicate with your customers.”
My issues with this headline don’t stop there. Not only does it write a big check that Mixpanel’s current functionality can’t cash, it also focuses on the feature, itself.
Titles like “The most AMAZING/ADVANCED/POWERFUL way to do X” tend to be overstatements, but they are easy to write This is probably why you see so many of them.
Whenever you start to write a headline like this, pause, breathe deeply, and recite this mantra in your head: “Focus on the benefits. Focus on the benefits. Focus on the benefits.”
The subhead (“Broadcast important messages to your customers. Just upload an image, craft your message, and hit send. It’s that simple”) reinforces my earlier point: this functionality sounds nice, but quite basic. It is not particularly “advanced,” and it is certainly not “the most advanced.”
Slow down, Mixpanel.
The mobile a/b testing slide could use an a/b test of its own.
Let’s start with the headline: “Experiment & surprise yourself with mobile a/b testing.” Again, the copy is focusing on the “what” of this feature, and not the “why.”
While you might argue that “surprise yourself” is actually a “why,” I’m not sure why, in particular, I’d want to surprise myself. Sure, it migbht be nice to surprise myself once and a while. But it’s not something I urgently feel the need to do right now.
How about: “Optimize everything about your app with mobile a/b testing.” This is not amazing. Ideally, it would have concrete results like “boost your engagement and retention by 25%,” but over-promising is just as bad as being bland.
The subhead (“A/B testing gives you the power to test your idea like a science experiment. Change the color of a button. Edit the marketing copy. If you’re really ambitious, change the velocity in your game”) is another round of bad copy.
Let’s break it down.
“Test your idea like a science experiment” is straight-up clunky. You don’t test science experiments. You create them for the purpose of testing. Meanwhile, “change the velocity of your game,” is not something I would understand or want to do unless my app is a game.
Ok, maybe a large chunk of Mixpanel’s most profitable customers make mobile games. But there are plenty of people who would benefit from Mixpanel who don’t. These people have no idea what it means to “change velocity” and they certainly would feel left out by “your game.”
Even some game makers won’t understand what it means to change their game’s velocity, or why they should.
That’s all for the feature slides.
Let’s look at the rest of the page.
The social proof section is decent.
Spotify. Uber. TomTom. These are some impressive brands! When I see that brands like these use Mixpanel, I know the product has some chops. This social proof is the strongest part of the page.
But it could still be better, starting with the headline: “You’re in good company. 2,906, to be specific” is another attempt at cleverness. While I like the numerical social proof (2,906 customers? That’s solid!), being clever is almost never as good as being direct about results.
How about: “Every day, 2,906 companies of all shapes and sizes use Mixpanel to understand, engage, and retain hundreds of millions of users.”
This way, you get the strong numerical social proof and the results in one, easy package.
It’s the most advanced way to write a social proof headline!
The case study section needs help, too
First the positive: they have faces of real people here! As much conversion research has demonstrated, faces of real people on your website help you sell.
But everything else, from the headline down to the blurbs below the faces, needs to be kicked up a few notches.
First, the headline (“Lessons learned from industry experts”) tells me neither what sorts of lessons we’re dealing with our what industries these experts are in.
The subhead sounds a bit pleading. It’s long, so I won’t write it out here, but the most unfortunate part is “We hope you find them as enlightening to watch as we do.”
There is good content in these videos, but I need to be sold on watching them.
The faces, themselves, seem random until you hover your mouse cursor over the images and the boxes slides down to reveal who the people are and what companies they work at.
That information should be easy to find on a quick scan, not buried like they are right now.
The text under the videos don’t tell me enough or sell me enough to convince me to click.
Ok, nearing the home stretch.
The customer testimonials section badly needs a refresh.
Using screenshots of positive tweets as customer testimonials is a risky game.
First, you have no control over the message. You have to rely on the happy things your customers say spontaneously, instead of in response to a specific question you asked.
Second, unless you constantly refresh them, they get dated. As I write this today, it’s April, 2015. The most recent dates in Mixpanel’s testimonial section (March, 2013) was over two years ago.
This could make the casual reader conclude that 2013 was the last time anyone said something good about Mixpanel on Twitter—not an idea you want anyone to have.
Meanwhile, potential customers care about tangible, measurable results, and these are nowhere to be seen.
At least, if I read these, I now know that Mixpanel is in the same product realm as Google Analytics, and that it has something to do with “customer engagement and retention.”
This is the first time these concepts come up. I could have easily missed them.
Conclusion: Even companies with great products need help telling their stories.
All of this is not to belittle Mixpanel.
As I wrote in the intro, the team has built an excellent product that can do remarkably useful things for your product and marketing if you know what you’re doing.
The problem is the way Mixpanel tells its story. Specifically, there’s almost no story here at all. It’s just a bunch of feature-oriented copy. Even this stuff hides behind a homepage slider that most people will never notice, let alone click.
Unless I’m mistaken (always possible) the core design and copy on this page haven’t been updated in years. Except for the addition of the feature slider, Mixpanel’s homepage is almost completely the same page it was in 2012.
But Mixpanel’s product has come a very long way in three years. It is now much more than an “analytics platform.” It is getting close to becoming a real solution for web and mobile app companies that want to do a better job understanding, engaging, and retaining their users.
That is a good story.
Unfortunately, you wouldn’t know it from looking at the website.