Geddes helps customers get the most out of Mixpanel, one of the most powerful analytics tools that startups have at their disposal. He has a deep understanding of tracking metrics and acting upon metrics.
TOPIC GEDDES COVERS
- Mixpanel is a analytics tool
- Provides libraries for various platforms
- Easy to collect data on user behavior
- Data gets sent to cloud servers and stored
- Provides visualizations and analysis of user behavior
- Metrics matter for tracking user behavior, understanding product usage and making data-driven decisions
- Metrics matter for validating assumptions about technology and user experience
- Metrics can be used to validate a product’s success early on
- Engagement metrics can tell if a product is worth growing
- And a whole lot more
LINKS & RESOURCES
WATCH THE INTERVIEW
READ THE TRANSCRIPTION
Bronson: Welcome to another episode of Growth Hacker TV. I’m Bronson Taylor and today I have Dennis Munson with us. Good us. Thank you so much for coming on the program.
Geddes: Oh, my pleasure, Bronson. Thanks for inviting me to come.
Bronson: Absolutely. Now, Geddes, you are the solutions architect for Mixpanel, which will be exciting for a lot of our audience, because whether you know it or not, Mixpanel comes up on about every other interview on this program. Well, that’s just one of those names, you know, it’s one of those tools that everyone in the know knows about it. So we get to talk to somebody on the inside that actually helps users get the most out of it. So this is going to be really cool. But first, tell us, what is Mixpanel? Just in case people don’t know.
Bronson: Yeah. Now make the case for metrics in general. You know, your whole entire product is built around the idea that metrics matter. So tell us, why do metrics matter?
Geddes: Metrics amount of matter for a lot of reasons. The first is really to validate a lot of the assumptions that we make about of about technol, about websites, about user experience. You know, we make a lot of guesses about, you know, this line of copy better than that line of copy or this user flow will be more effective than that user flow. And before, you know, we really had a data driven culture here. And, you know, as product managers and developers in technology, a lot of times that was based on gut. And if there were disagreements between different people on the team, usually it ends up being, you know, you might have heard this phrase before, a kind of a hippo decision where it was the decision was made by the highest paid person’s opinion because what the organization is and with metrics, you don’t actually have to rely on that. You can literally validate your assumption or prove that one user experience is definitively better than another. In terms of measuring your goals. And so that’s really important. It’s also really important to validate your product, right, especially early on. So, you know, you might not have a lot of users, but if you can use metrics to show that you have the few users you do have are incredibly engaged that tells you you have a winning product. And that’s actually for someone that’s, you know, probably the kind of audience it’s watching growth hackers where you’re actually looking to grow and start your small right now and you want to grow to even know if it’s worth going after growth, you need to know, do people enjoy using my products and metrics? Engagement metrics can give you an unbiased, really good news or bad news, but it’ll give you the news definitively about whether it is or not.
Bronson: Yeah. Now, this may seem like an odd question, but in your opinion, do you think it’s ever possible to be too data driven on this program? We talk about data all the time. We love data. We love metrics. But can we go too far with it? What do you think?
Geddes: You know it. Sure. You can go too far with anything. I think that it’s still an uphill battle in a lot of organizations to make data part of your decision making process. And so I wouldn’t I wouldn’t take that as a problem on anyone’s radar right now, because. So not enough people are being are being data driven. The I mean, I guess, you know, one way to put it is, you know, if you don’t have a good analytics tool, for example, or your analytics to exist, but no one’s using it, it’s kind of an uphill battle to get people to take that and use it in there and integrate into it into their decision making process. And so, yeah, so I would I mean, yeah, there’s, there’s times when, you know, maybe you don’t need to be data driven, right? Like if you’re just coming up with your concepts, right? You know, maybe, maybe then. But in general it’s always good to have.
Bronson: Yeah. So the problem isn’t too much. The problem is too little in most cases right now, maybe in ten years the discussion will be different. But for now, that’s the problem.
Geddes: You know, our customer, you know, the customers want the Mixpanel. They’re not they’re not coming in saying, you know, we are you know, we’re being too data driven right now. Everyone that comes to us is saying, you know, they’ve one of two problems. Either they say, you know, we don’t have any data and we want it. Another problem they have is we have too much data and we’re examining it and we don’t know what to do with it. But that’s not really being too data driven. That’s just not having a good data technology. That data stacks that up.
Bronson: Yeah, I see what you’re saying. So you mentioned the companies that come to you and the problems they have. What are the kinds of companies that they get the most out of Mixpanel? Who is Mixpanel built for?
Geddes: Good question. I mean, we’re built for any digital experience, right? So whether it’s, you know, a web app or a mobile app, any app. That really is their goal is engaging the consumer, engaging the customer. That’s who makes panels really good for. So if you want to get a sense of exactly how users are interacting with your app and whether they’re, you know, and whether they’re doing the things that you want them to do, that’s. And as for and so for companies, a lot of stages, it works, right. If you’re early on the growth stage makes panels really good for that. And we have a lot of customers that are even on our free tier that are small but getting a lot of value out of next. We have huge customers as well, well-established mobile applications with large user bases where they’re at that point going for point level optimizations. Right. If we can increase our conversion rate by 1% to purchasers. Right, that represent that, that means millions of dollars for us. Right. And so it works on both e-books, both customers in a very large scale and customers on a very small scale.
Bronson: Yeah. And you know, your pricing structure compared to some of the competitors. It seems like you really do allow startups a place to begin the experience you have, you know, the trial. You have some lower price points. And so you have, I think, captured a lot of that market because of the way you’ve structured the pricing in a lot of ways. What kind of growth has Mixpanel had? I don’t know what you guys have disclosed publicly. I’m not sure you need to say anything that you know you’re not allowed to. But if you can, you know, tell us about the number of users or revenue or how many events you’re tracking. Just anything you can do. Give us a sense of scale right now.
Geddes: Yeah, no, I’m happy to. It’s been I’ve been at Mixpanel for 18 months now. I came as the seventh employee here and when I came, we had 100 paying customers. We just did the count yesterday and we’re at 1008. So it’s been a remarkable for me personally, remarkable 18 months of growth. And we think paying customers is a really good metric for us because these are people who actually every month, you know, they’re writing the check, you know, and that they’re getting value out of the tool. And so, yeah, so it’s been this remarkable period of growth and it’s been really fun to be a part of this company while it’s happened. And the great thing about our growth is we grow when our customers grow. So in addition to our number of customers increasing the number of data points that we actually process is increasing. When I came, we were just at about a billion events per month across all of our customer base. Now we’re at 12 billion events per month. So this represents both growth in new customers, but also growth within our own customer base. So we have plenty of examples of companies that have come and started on our free plan and now are sending hundreds of millions of events to us every month. And that’s just awesome to see because when they succeed, we succeed and we really love being a tool that helps them grow.
Bronson: Yeah, that’s great. It’s great when you win because they win, it means you’re, you know, your your priorities can be aligned, you know. What have been the primary channels for that growth? I mean, where do the users come from? Is it word of mouth? Is it paid acquisition? Is it, you know, some other kind of marketing? How do they actually find out about Mixpanel?
Geddes: Great question. So in my experience, it’s been entirely word of mouth. You know, we have you know, we hired our first marketing person four or five months ago, and until then, it was entirely word of mouth. Now, we did have a little bit of a paid like search campaign going on, which brings in some business, but it’s certainly not our engine for growth. Our engine for growth has been just having a great reputation and having, you know, startup founders tell their startup founders about the tool. And that’s been that’s been really good for us.
Bronson: Yeah, that’s that’s a great way to grow. Now, as the solutions architect, part of your job, if I’m not mistaken, is to make sure that the users get the most out of the product. Is that accurate?
Bronson: Yeah. So tell me, what advice do you generally give to a new user who really doesn’t know where to begin? They decided that metrics matter, but they don’t know much else. They don’t know what to do next. What’s some of the basic advice you give them to where to start?
Geddes: Yeah, it’s a great question. So what I usually tell someone who’s just starting, who’s sort of in this category of, you know, I know I want to be data driven, but I don’t know how right to do that for those folks. I actually have a pretty good way of like a paradigm almost for constructing on how they should think about their data. And it works really well, especially for companies that are just starting out. And I really I tell them, you’ve got to think about your analytics in three categories. The first category of analytics are user acquisition metrics, which is essentially a single event that you fire. When a user does something to expressed interest rate for a website, it might be this user comes on one of your landing pages, right? For a mobile app. It’s that first open of the app, that first that first app open. But it’s some metric to just say this is the very, very beginning. Now, a lot of times the acquisition metrics are where people start rate. You may have seen you see tons of press releases of apps saying we got 100,000 downloads done right and that’s an acquisition metric. But that really doesn’t tell you too much, right? Because the thing about user acquisition is you can always juice those numbers with money, right? You can always buy ads. You can always you know, acquisition is a problem that has been. You can always get you can always get downloads. Just it just takes money. And so we tell them, okay, acquisition is your starting point. But you have to look further. You also have to look at user activation and user activation. Are the users who have had that core experience of your app, right? So let’s say let’s say let’s say it was a social network, something like Facebook, right? Maybe an activated user is a user who has made a status update, right? If it’s a photo sharing app, an activated user is a user who’s uploaded that photo. Right? It’s whatever that thing is, that is because of experience. And, you know, active metrics are important because that’s when a user’s really given you a chance, right? They’ve done everything you can ask them to do based on faith and marketing, right. You made promises on your landing page. They were intrigued. They said, I’ll try it. I’ll go do the core app experience. So one thing you look at is how many of my acquired users then become activated, and if I even become activated, and if that’s a low percentage, then that means, okay, you need to work on your marketing materials or your sign up flow is confusing, or the way to get from action to activated is hard. So that’s a really important metric to look at. But that’s not even the whole story, right? You might have a great flow. It’s really easy to use and users get activated. They’ve given you the chance and then they just say, this sucks. This wasn’t useful to me. Right? And the third thing I tell staff to look at is retention. So out of your activated users that have given you a chance that had put some faith in your product, how many of them come back and use your product again and use your product again and use it again? And how often do they come back? And that’s retention. Yeah. And you know, any growth hacker is going to be looking at all three metrics to understand their businesses and their strengths. It was interesting, about a year ago we raised a venture round from Injuries and Horowitz and the reason that they got in touch with us was because all these startups were marching through their doors and were putting screenshots of the Mixpanel retention reports in their decks. Yeah, you know, so what is this Mixpanel everyone’s using? And the reason stories are putting it in their decks was because the retention report, that metric is so important for validating the product. Like I said, if you have a small number of users, but they’re really engaged and they’re coming back every day or every week, whatever, you know, the flow of your app is, then, you know, winner and you know, with money then you can solve the acquisition problem. But that retention number is really, really important.
Bronson: Yeah, that’s great insight into not just a beginner but intermediate. I mean, that’s what we need to be thinking about now. What mistakes do you see companies make when they first start using Mixpanel? Is there any common things where they all want to do this and that’s not the right thing to do?
Geddes: So the biggest mistake that people make when they start using Mixpanel is not thinking clearly through their their events. So Mixpanel is an event based analytics tool. Our data model is you, you know, you fire an event whenever a user do something interesting, right? So like I was talking about a video upload or a picture upload or, you know, sign up steps in the sign up flow. Those are all events that you fire off the next panel and then you need to send properties at those event as well so that you can send it, right? So a classic property to send would be, you know, what acquisition channel that this user come from. Did they come from a paid search campaign? Did they come from a viral channel? So you can segment that way and see which of these channels are most effective. And it really takes a lot of upfront thinking to do this. Well, a lot of people that want to get started really quickly, they might just throw up an American called Dude Page and that’s it. Right? And then at that time you’ll be using Google Analytics because all you’re doing is a list and aggregated list of what pages people were viewed. But you really think about these three categories acquisition, activation and retention. Think about what events you would put in each of those categories and add code to your site or your app to fire those events. Then you’ll get a lot more usage out of Mixpanel. So the biggest problem is actually not thinking through ahead of time what you want to track.
Bronson: Yeah. Is it a good strategy to place an event on everything and then go back after the fact and sort it out? Or is that just messy and doesn’t really work?
Geddes: You know, I tend to to tell folks not to do that. It is a temptation. Right. So that’s another thing people do. Right. You see, you know, especially for people from an engineering background who are developers, they might, you know, at an event and for every function in their app. Right. And it just becomes a lot. It becomes a lot. And so another thing is, you know, you need to be smart about what data you’re tracking because those are real resources you’re using. Right? Every every piece of data you track is going to is going to either whether use in Mixpanel or some other tool or even your own system. That’s going to have to store that data. It’s going to cost too much. And so it’s really important to think upfront about, okay, is this thing I’m tracking actually actionable? Is it a decision off of. So whenever you add that call, whatever tool you’re using, you should be asking yourself, is this something I can I’ll be able to use to make a decision? If not, then I probably shouldn’t be tracking it.
Bronson: Yeah. Yeah. What are some of maybe the most powerful tools of Mixpanel that you think get under, utilized or ignored by people? Is there any, like, you know, golden gems that are hidden that people just don’t talk about enough, in your opinion?
Geddes: Well, I’ve talked about retention a few times, going to talk about it again. It is most powerful tool that we have. It’s also the one that looks the most intimidating upon first. And so we actually have a lot of customers and we have data on this ourselves. Right. Who like, of course, they’ve got tab and they never go back because they see some numbers and it’s confusing to them. And so one of the one of our job solutions architect is to make sure to explain what is this report actually telling you and why is it important? So retention is huge. The other thing that’s really huge on the Mixpanel side is our revenue analytics and something that Israel is relatively new for us and so not a lot of people know about it, but we actually make it really easy in addition to tracking generic events when users do interesting things. We have a specialized line of code you can add whenever a user makes you money, right? If they buy a virtual good or they make a purchase or they upgrade to the pro version or whatever it is, you can you can we call it track? You track those and then we let you calculate things very easily that from a perspective you want to know, you’re going to want to know your average revenue per user, your life and value, how long it takes a user to reach their maximum lifetime value. And actually working on your paid growth campaigns, you’ll know how much you can afford because you know how much money you’ll be getting back.
Bronson: Yeah. Do you see any themes or trends with the companies that really get the most out of Mixpanel? Is there any common makeup or any common things that you see with the teams that just really utilize it to the max so that we can imitate, you know, the people watching this interview?
Geddes: Yeah. Um, so the mean I would that’s an interesting question for we definitely see a lot of success with companies that set up empowered growth teams and growth teams are really interesting phenomenon. Usually it involves at a minimum, an engineer and a product manager. And then pretty important thing about a growth team. And actually, at minimum, you can just be an engineer, right? But the important thing on a growth team is that the growth team is empowered to make changes quickly. They don’t have to go through, you know, a regular code review process. They don’t have to go through a regular release cycle, but they can do changes on the fly. That’s why I didn’t have an engineer there. And these people are going into Mixpanel, which is a real time tool and making changes a few times a day and tweaking and trying new versions of copy, putting buttons in different places and seeing how that impacts things and they see real results there. So having that like that kind of structure that allows to be really nimble and make decisions quickly based on the data as you get it is something that’s been really successful and we’ve had great events at Mixpanel with done on growth teams, right? We have a great we did a great event with Branch out of that a growth team, Airbnb has a growth team. These kinds of companies with growth teams really can use, tweak and use analytics to make to make lots of small tweaks or sometimes even big tweaks and have huge impacts.
Bronson: Yeah, that’s great. You know, a lot of people watching this are jealous. They want to grow team, so they aspire to have one. Right now. They have a growth guy or a guy that does a couple of different things, but I think that is the goal. Get to the place where you have a team of people really focus on that set of problems because that’s a big deal. What go ahead.
Geddes: Is empowering them. Right. You know, the where you where we don’t see a lot of success is if you have an analytics guy who isn’t empowered to actually make product decisions or doesn’t have a lot of influence because this person can do all the reports and make all the points they want, but it’s going to be much slower for them to actually close the loop and bring this back to the products.
Bronson: Yeah, for sure. What are some of the biggest wins that you’ve seen companies make after they you’ve used Mixpanel, they come, they see their data, they make actionable decisions based off the data, and then this happens. What are some of the case studies that you point to for things like that?
Geddes: Yeah, absolutely. So there’s one I’d like to talk about a lot, because one is that they let us see. We’ve done a case. It is Fitbit. So Fitbit is one of the emitters that’s out there, tracks number of steps that you take every day. And the way Fitbit works is you have to sync it via Bluetooth to either your phone or your computer so you can upload that data. And in a very important part of the initial user experience for the Fitbit is making that sync work, because if that sync doesn’t work, the product is useless. And Fitbit used Mixpanel to see that on there. You know, people were using the computer version that are thinking with their computer had about a 90% success rate for doing that sync. That first time, but they were also collecting data for their mobile app and they were able to see as soon as they installed Mixpanel, they saw, oh my goodness, our conversion rate for people syncing with the mobile version is only about 50%. It’s about half of what it is on the desktop. And so there’s some problem there. And the great thing was they immediately saw in the data there is an issue, if you think about it now, this is someone who just paid whatever it is, 50 bucks, a hundred bucks for a Fitbit. And so they are pretty motivated to get it to work. They’ve already spent the money and if they’re not thinking successfully, there’s something terribly wrong. And so what the Fitbit guys did was they took a look at their flow and they saw exactly the steps where they were losing people. And it was the step where is actually the you know, you hold your Fitbit. In fact, I have one right here I can show you probably.
Geddes: So this is a fit, right? And you know, that part of the step is you take your Fitbit and you hold next to your phone, right? And you’re like, come on, you sit. Right. And this. And the phone says, I’m searching for my Fitbit. And the Fitbit is just like, you know, doing nothing. Right. And so I can see these things look like and this is the step where it was falling apart and what they realized. And they made a guess. And the guess was that the Fitbit wasn’t the Fitbit wasn’t actually awake. And you have to actually tap at the end of the Fitbit to wake it up, and then it’ll successfully sync. And on the computer version, there was a big line of text saying Tap the Fitbit to wake it up. But they hadn’t put that in their mobile version. They had limited screen real estate, and so they made two changes, two simple changes you’d think this would be, you know, when they saw this, they’re like, Is there a hardware problem? We have to get our engineers involved. Do we have to ship new devices? Are they correctly? No, it wasn’t. I think that all they needed to do was push out to their mobile app to change as one advertising copy saying, Hey, tap your Fitbit to wake it up. The second change they had to make was they changed the diagram and they made a diagram of a finger tapping the Fitbit. Right. And that’s what they did. And and then their rates went back up to what they had been on the computer and they were able to solve that problem right away. So that’s an example I like to tell about how that initial user experience can even be. It can be such small things that users and by testing and releasing new versions and then seeing the results, you can really fix things.
Bronson: Yeah, that’s such a great example because you know, it shows us that the problem isn’t how hard the solutions are. The problem is you don’t know there’s a problem, right? And so I love that example. And Mixpanel solves that problem. It shows us here’s what the flow looks like, here’s where it’s broken. And sometimes the solution is actually the easy part. Now in Mixpanel, you guys have continuously innovated. There’s always some new feature, a new piece of the product that’s kind of coming out every few months, it seems like. And you’ve got an awesome many things I want to ask you about. Some of these things have been added over the last year or so. Talk to us a little bit about Mixpanel streams. What are they and how do they work?
Geddes: Yeah, okay. I have like, you know, streams actually is one of the oldest products we have at Mixpanel. It isn’t one that we’ve added recently. It predates me at Mixpanel. And you know, the thing about streams is that just it’s just trigger. So you know what it does? It takes a slice of your users who are on your site or on your app right now and pops up their names and pops up the events that they’re that they’re firing like as just like this flow that keeps scrolling down. And you can see exactly what’s going on. Now, the thing about streams is that it’s real eye candy, right? Like you can see it’s very exciting for a developer to fire up. Seems to be like people are using my app. This is exciting, right? The problem with streams is because it’s a bunch of data about its individual users. Right? And you can and it just keeps flowing and scrolling. It’s actually hard to make decisions with it because it is it’s just too it’s it’s so individualized that you can’t you get lost. You look at the tree, but you can’t see the for it straight, whatever. That’s I forget the phrase. There’s some nice phrase there, but you think you know what I mean. And and so we actually streams used to be our default view because people love seeing seeing it. They kind of got mesmerized by it. But we now have sort of buried it like three or four links down on our navigation because it actually isn’t an easy tool for making data driven decisions. You want to be looking at aggregations to do that and streams and let you do that easily. So streams is a lot of fun and it’s also like, you know, when you’re giving a demo or someone. To someone like Lou. You know, you always get that ooh moment. But the problem a lot with a lot of times with like reports that are very flashy is they’re not as actionable. And that’s the issue with streams.
Bronson: Yeah. So then tell us about engage a little bit because it’s not quite as flashy and just for eye candy.
Geddes: Yeah, absolutely. So engage is the you know, let me let me turn to screen share and show you this product, please. Definitely. Definitely can do that. So what engage is the old name for what we now call notifications. And so what notifications are it’s part of our people analytics products. And the cool thing about people analytics is it lets you build profiles about all of your users. So here, for example, I’m showing a list of all my users who created their account within the last week, and I can click on any of them and I can see all sorts of information about them. I can see their accounts. You know, this is all customizable properties about a user. I can see their account status. I can see which version of the landing page experiment they saw when they came on board. I know that they’ve left 87 reviews. The example here would be like a site that sells shoes. So you can see that they’ve got a to shoes. I can see all this information about a user and what notifications lets you do is actually have triggered automated notifications that get sent to users once they meet certain criteria. And it’s useful for a lot of reasons. One reason you might want to use it is if you wanted to send a message to a user after they may have signed up for an account, but they haven’t bought anything yet. You could notifications do that, say all right if a user has signed up for an account, two weeks have gone by and they haven’t bought anything yet. Let me send them an email that has a coupon code, right, or something like that, or let me send them a push notification that tells them about a sale that we’re having this week. That’s the kind of thing you can do with it. You know, I think because we have Skype, minor connections, too slow to do the share. So I’m going to turn back on video mode. Okay. That’s that’s the general idea.
Bronson: No, that’s great. Thank you for showing us that. Now, tell us about the revenue analytics a little bit. You mentioned it earlier. Why is that so powerful? Why is understanding revenue such a meaningful part of analytics?
Geddes: Yeah, well, if you think about revenue, it is the lifeblood of any company, right. If you’re not if you’re not making any money, then why are we here right now? Obviously, there’s early stage companies where you’re not worried about revenue right now, you’re worried about growth. And that’s why we have other Mixpanel reports. But for companies that have decided revenue is a goal of theirs right now, it’s once it becomes a goal, it becomes the goal. Right. And it’s really important. So what revenue analytics lets you do is automatically send the data to Mixpanel every time someone spends money and allows you to look at that in a very detailed way. So the obvious calculation is the average revenue per user and that tells you, okay, for every user that signs up, how much money do I make out of them? And so let’s say you had let’s say you had a game, right? And you had, you know, for every ten users that signed up, eight of them never spent any money, but two of them would. Right. And let’s say each of them might buy $10 worth of virtual goods. Right? So for ten users, you’ve made $20. And so your average revenue per user then becomes $2 per user. And why do you want to know that number is so that you know how much you can spend on acquisition, right? If you’re making $2 per user and that’s almost all profit, you could spend up to a dollar 99 and on per acquisition for mobile apps. Now that is the very generic case. Now where that breaks down is, well, let’s say you do this calculation for your organic users, you find you have $2 average revenue per user. Now that’s not actually a guarantee that if you purchase traffic that those guys will also be worth $2 per user because, you know, they might be less motivated. They have might have been invited by friends or whatever. You know, they might be doing it as some sort of incentive campaign for their game. And so it’s really important to be able to segment your revenue out by things you know about the users. So one thing you might know about the users or what acquisition channel that they come from, if they came from an AdWords campaign, am I still able to get $2 per user? Maybe, yes, maybe you get more from them, but maybe no. And so you might say, yeah, you know, if they come from AdWords, I only get $0.50 a user, so I’m only going to bid $0.49 for those for that acquisition. The the other thing is you can see how achieving certain milestones impacts revenue. So you could say, all right, my average revenue per user is $2, but what did those two users do? Spent $10 each, right? And say, oh, those users all made it to level five. So you can segment out by like most recent level achieved and say, all right users who’ve made level five, they my average revenue for them is ten users. So I’m going to do everything I can to like push users into level five to make that happen. And you can make both product learnings and marketing learnings and see how it impacts your bottom line.
Bronson: Yeah. Now when you had Mixpanel open a moment ago. On screen share on the left side of the page, there’s a number of kind of tabs, a number of different sections of the site. Do you think the goal is for a company to really master every single aspect of Mixpanel to get the most out of it? Or are there use cases where you don’t need these three things but you need these four? How do you view the interaction with the user and the fullness of the product?
Geddes: Yeah, good question. I mean, I think that both sides of the mix. So Mixpanel to understand those tabs, there’s two sides to Mixpanel. There’s the event side, the what we call engagement, and there’s the people side, which is where the notifications is and the revenues that come on that. And then I was talking about that’s where that falls under. And so both are really important, right? Engagement is all about collecting the data. People, analytics makes it easy to take action on the data, right? With people analytics, you can take action without having to make, you know, program changes to your site. You can actually send notifications without doing that. And so that’s one way to think about it is, you know, the engagement reports are very important for understanding your users. The people reports are very important for taking action right away. And that’s what we wanted to make easier. So yeah, I think that all the reports that external has are valuable to most companies now. There’s plenty of companies that don’t use their use, just engagement. They don’t use people. We’re working hard to get them on people and they see that that product is growing very quickly. But yeah, so that answers your question.
Bronson: I know. Absolutely. And the reason I ask is because a lot of people, they start with Google Analytics, which is just a flow of information, most of which has no bearing whatsoever on the stuff they need to get access to in terms of data. And this panel is a different world. It’s not a flow of information. It’s a lot of valuable information for people that are running companies. And so I wanted to make that distinction that this probably is a product that you want to master as fully as you can. That was my guess at the answer anyway.
Geddes: It is. You know, last time I logged into Google Analytics, I counted and there were 47 links on the left hand side. Right. And if you log into a, you know, and you’re not trained or you don’t you’re not an expert, you have no idea what those links do. Right? So what most companies end up doing is, you know, they hire a consultant to maybe set up three or four reports that they go into. And then Google Analytics really becomes a reporting tool, whereas with Mixpanel, we want to make it a tool that is inviting, right? And so there’s not 47 links, there’s a few links. They all have specific purposes and functions. And when you click on them, you immediately understand what the state is telling you and no matter how technical or trained you are, and then we have little tempting buttons, you can click, right? There’s a little arrow that you can expand to find out more information and to dig deeper. And so our goal there was to make a product that’s both easy to use, but incredibly powerful, and to allow people feel comfortable not just pulling reports, but feel comfortable interrogating the data. Right. Feel comfortable asking questions, asking for questions, and not feeling like they have to go through an analyst to do it.
Bronson: Yeah, that’s great. Now you also have a site that actually educates users about site analytics kind of in general. Where can people find your education platform right now?
Geddes: So it’s Mixpanel dot com slash education. And what we do is we do monthly events here at our headquarters in San Francisco and we invite great minds in analytics, whether some Mixpanel customers, some aren’t Mixpanel customers to talk about things that they’ve learned. Actually, tomorrow we’re going to be doing one with Airbnb and they’re going to be talking about two interesting things. They’re going to talk about how they decreased their conversion rate, which why do you want to decrease your conversion rate? Well, it was the Help Center. They wanted to decrease conversion rate of tickets. And so actually a low conversion rate is what they were going for. They also talk about how they increase the conversion rate on the listing side for people trying to list their apartment. And these are great stories because what we have, we invite these customers of ours and also people just using analytics in general to come in and tell us their stories and share the details. Right. And that’s the important part. It’s like not just generic. I was able to increase my conversion rate. It’s like, here is the problem we had. Here’s a screenshot of our audio, here’s a screenshot of our new view here. This changing this copy here at this impact and their specific stories. And that’s really the best way to learn is to learn from these specific stories because yeah, it’s an abstract concept, but having real people bring it down to earth we found to be a very valuable educational tool.
Bronson: Yeah. And is education a part of mix panels growth strategy in your mind? The more people know about analytics, the more meaningful Mixpanel is in their life?
Geddes: Oh yeah, absolutely. I mean, the biggest problem in analytics is education, right? Some people have this vague idea I should be using data, but they don’t know how. They don’t know how to go about it. And on the Solutions team, that’s our mission, is to educate people. You know, when we’re hiring core solutions architects, you know, one of our requirements is are they a good teacher? Because you often have to find yourself teaching about not only Mixpanel on the product, but about concepts and analytics. And so our education page is. One of the one of the three things we’re doing by making this public is giving anyone a resource and learning, well, what are the experts in the U.S.? What are the very best people in analytics doing and how can I emulate them?
Bronson: Yeah, that’s great. So what’s the best talk you guys have had so far? You said you bring in these people. They talk about specifics. Has there been one talk that really stood out as like, wow, that was great information.
Geddes: So, you know, I don’t want to you know, we have a lot of great people and they’re friends of mine. They’ve all given talks. And so, no, I don’t necessarily want to say there have been you know, one is better than the other. But I will say there are some are particularly appropriate for people on the growth hackers channel. So the one that we have by Vox here is very, very good for someone who’s just starting because Boxer and Reeves, they don’t know about Boxer. They are a walkie talkie social network. So it’s voice communication. You have a Boxer app for iPhone on Android and you essentially, instead of sending text messages, you send voice messages. And it’s pretty cool experience. And what Igal does, who’s the product manager there as he talks about this paradigm that I talked about, he talks about activate acquisition, activation and retention, and it’s a really good concrete example with examples from how they defined those for their app and how they looked at those numbers and how those numbers grew. So I’d say for folks who are in the stage where they are just starting out, the boxer case study, the boxer on education pay boxer we call I’m trying to in for the boxer educational video is very, very good.
Bronson: Yeah, that’s great. Well, get us this has been an awesome interview. I got one last question for you. Just kind of a broad question. You can take it any direction you want. What’s the best advice that you can give to anyone that’s trying to grow their startup right now?
Geddes: Yeah, good question. The best advice I can give to anyone trying to grow their startup is. Just to iterate very quickly. I mean, that’s why we do that and expand on that. When our customers do that, we see lots and lots of success. So no matter what. Obviously in a data driven way, but no matter what tool you end up using, whether you’re doing your own stuff, whether using Mixpanel or another tool, the most important thing you can do is focus on retention and iterate quickly to see how you can improve that. And those are the those are two big takeaways that I would leave.
Bronson: That’s great advice. Again, get us. Thank you so much for coming on the program today.
Geddes: This was a lot of fun.
Bronson: It was fun. Well, to do it again sometime.
Geddes: Great. Sounds good. Thanks, Bronson. Thank you.