John was a growth hacker at Shopkick, and is now an engineer at Pinterest, and he teaches us how to increase acquisition, engagement, and retention on mobile apps across iPhone and Android.
→ How to increase acquisition, engagement, and retention on mobile apps across iPhone and Andriod
→ What are some of the specific things that he did move the needle on engagement
→ What does he do as a part of that redesign that upped engagement so that our audience can learn from
→ How the different segments do he used offering or how the engagements worked
→ How does he decide what ideas to implement
→ How fast did he implement ideas like
→ How long does he spin kind of doing due diligence
→ They increase daily active users by 80% and increased revenue per user by 20%
→ And a whole lot more
Bronson: Welcome to another episode of Growth Hacker TV. I’m Bronson Taylor and today I have John Egan with us. John, thanks for coming on the program.
John: Thanks for having me.
Bronson: Yeah, absolutely. Now, John, you were a technical lead at Shop Kick. Where you headed? Up the growth hacking team. And you are currently a software engineer at Pinterest where you focus on growth and in both companies. You’ve had some notable successes that I think we can learn a lot from. So let’s just jump right in and let’s talk about shop cake for a little bit and then we’ll talk about Pinterest a little bit. And then at the end, we’ll talk about a really clever hack that you know of that I think our audience is going to like a lot. That’s my teaser, The Way People Stay the end of the interview. I see that. So let’s talk shop kick at shop kick. You manage a team of six engineers that were responsible for a complete redesign of the mobile app on Android and iPhone. And according to what I saw online, you guys increased some key engagement metrics by two X all the way up to ten X in some cases. So tell me first, what are some of the specific things that you all did that really move the needle on engagement? What is it that you did as a part of that redesign that really upped engagement that our audience can learn from?
John: Mm hmm. So just so your audience is a little bit about Shortcake, what shortcake is, is it’s a mobile shopping app where you walk into retail stores that were partnered with, like, Macy’s, Best Buy, Target, and you collect these rewards points just for walk in the door and use those rewards points to get gift cards afterwards. So in our initial version of the app, it was very utilitarian. So the main use case was you just open up the app when you’re actually at the store. There’s not much reason to open it up any other time when you’re at home. And most people generally don’t go shopping every day, at least I do. So one of the problems we encountered was people just wouldn’t open up the app that often. And when your app is in top of mind, it’s a lot more difficult to get people to sort of make it a habit. Mm hmm. So with that redesign, the goal was to make for daily use case. It’s what we called sort of the couch mode, which is when you’re just on a bus or you’re on the train or you’re just sitting at home and you have a couple minutes to kill. So what we did to redesign the app and sort of try and solve for that couch mode use case was we brought in a lot of content from our retail partners so you could almost go window shopping using our app. You could see what deals were going on, what were the latest products. And we also just made content a lot more for front and center in the app, whereas previously we had like maybe a couple of deals per store, but there’s you could go through all the content in a store in like 10 seconds. So yeah, we just added a ton of more content to sort of create a daily use case and it helped increase things like session length and like number of times people use the app.
Bronson: Yeah. So kind of takeaway maybe for somebody watching this episode right now is if you have an app that doesn’t necessarily require daily use, your goal is to get people closer to that. And one of the ways to do that is to think, okay, it may not be the core value I’m providing, but what can they do every day or what can they do some days, you know, more than what they are now and try to incorporate that. In your case, it was content from the retailers. In someone else’s case, it could be anything else, right?
John: Mm hmm. Exactly.
Bronson: Yeah. That’s awesome. Is there anything else you guys did as a part of that redesign, or is that really kind of the centerpiece that changed all the metrics for you?
John: All that was really the centerpiece. Just bringing like that content for front and center. Of course, we did a lot of stuff after that project around growth, but I think we’ll get into that in a little bit.
Bronson: Yeah, absolutely. Now, did you guys could you did an Android and an iPhone app, which is hard to do kind of development on two platforms. Did you guys kind of notice any differences about how the different, you know, segments used your offering or how the engagements worked? Or were they really just the same people on a different phone? What did you guys learn there? Anything.
John: So once someone becomes a real user and they start using your app on a regular basis, there wasn’t a ton of difference between users, but getting them to that point, definitely by platform running, right. I found it generally was harder to activate a user and get them to become a regular user. And also, there’s just fewer Android users coming in compared to iPhone users. Yeah, but I think it can sometimes depend on your app because I know there’s other companies out there like, for instance, couponing apps that really target sort of like lower income people that actually see better in. Engagement on Android.
Bronson: Okay. Probably because Android’s come with plans for really cheap and they come with more kinds of plans. Where iPhone. It’s a higher price point. It’s a different demographic. So really, if your demographic is an Android demographic, you’re probably going to see numbers slanted that direction, right?
John: Yeah, exactly.
Bronson: Now, now it’s shocking. You worked with a team of people where, you know, we kind of alluded this a second ago where you brainstorm features to really work on three things to increase user acquisition, engagement and retention. Those kind of three big ones that everyone needs to focus on. Get people in, get them using it, keep them around. Those are those are the big three. First, I want to ask you about how did this team actually function? Because, you know, like you said, there was a number of people on this team. What it look like day to day, were the brainstorming sessions look like? How did you decide what ideas to implement? How fast did you implement ideas like. Walk me through kind of the makeup of this group because there’s other people watching this and they may have a team of people and they don’t know the best way to structure their time to get the most out of them. Yeah.
John: Mm hmm. Yeah. So our team, at least at the beginning, consisted entirely of engineers. And what we would do is we’d usually pick one area to focus on. So we had, say, for the next few weeks, we’re going to focus on this acquisition or we’re going to focus on retention. And then what we would do is we would just sort of get in a room, get everyone in a room together, and just take a few hours to just start brainstorming ideas around that area. So what can we do to, like bring in more users? What about existing mechanisms, such as invitations? Can we improve and help get us more users? And then after we did brainstorm all these different ideas, then what we would do is sort of just take a vote and everyone could vote on which ones they liked and which ones they didn’t. And one of the things we found was actually that this voting process didn’t work very well. Everyone felt like they had their say, and that was great. But we ended up spending time on some projects that actually didn’t move the needle. So one of the things we learned was to really focus on doing some due diligence on projects before we actually started executing. So we started after we come up with a list of projects. And so we’ve come up with a list of what we thought might be interesting to work on. Then we would actually go look at the data, see if making running that experiment or adding that feature would actually get us enough of a gain to make it worthwhile. And so we sort of moved to more of a r y model. So we started brainstorms, but then we would pick which projects to work on based on sort of what we guessed I would be.
Bronson: Yeah. So how long would you guys spin kind of doing due diligence? Is that going to be a couple of days and then you have your session and talk about it? Or would you spend a week on due diligence? Like, how much time did that usually take up?
John: Yeah. I would sort of depend on the project, but typically it would be no, no more than like maybe a day or two of due diligence. We went in like do due diligence on every single project. Some of the projects are just we can kind of tell we’re low and many of them we’re like digging into. But all of the ones that were sort of candidates to work on for like the next few weeks to a month, we’d just go through, run through the data and then take too long to do that. But definitely helped us be more impactful as a growth team.
Bronson: Yeah. So you kind of went away from democracy and more toward data as you as you evolved as a team. Yeah. Now, with the numbers that you were trying to increase, you know, the, you know, acquisition, engagement and retention, you actually increased invites by 400%. You increase daily active users by 80% and increased revenue per user by 20%. There’s a keeper right there. So that’s the one you want to increase. Tell us about some of your favorite ideas that you guys implemented. They actually made those numbers possible. What did you do to actually get those things up and to the right? Give us some specifics here.
John: Sure. So I’ll talk about, I guess, each one time. So to get invites up invites were in user acquisition channel because we were a mobile only app and we were only at select retailers doing stuff like SEO or some other type of use. Acquisition and models didn’t really work for us. So we really focused on invitations. And there there is, I think, three key things that helped us get those big gains. One was the incentive. So because we were a Rewards Points app, we use those rewards points as an incentive to get people to invite their friends. And we started off with this sort of mediocre incentive is like you get 50 points and your friend gets 50 points and 50 points. It amounted to something like ten or $0.20, if I remember correctly. So it wasn’t too great. And then we moved to this model where it was, you get 2000 points. Inviting three friends. And so it was a little bit more expensive. But the increase in spend, actually, we got a bigger increase in terms of like number of new uses we got that was greater than the amount of increase in spend and ended up being worth it for us.
John: And then a couple of other things we did around the invitations was reducing the friction in our invite process. One of the big problems we had being a mobile only app is a good number of people who get invited will click on a link, but they’ll be on a desktop experience, so you need to get them onto the mobile app. And we did a lot there to just try and reduce the friction in the flow. So originally we had people doing, going through and creating a whole account on web and then getting them to try and log into their account when they actually downloaded the app. But then we came up with we added something like tracking that allowed us to basically close the loop on the invite as long as we got them to visit a web page on their mobile browser first. Mm hmm. And using that, all we had to do is if you’re on desktop, just give us your phone number, and then we just send you a link to the store. And if you’re already on mobile, we just send you straight to the App Store.
Bronson: Are you taking it to them or emailing it to their phone or what?
John: We’re texting it to them to make sure that it gets on their mobile phone. And then if you’re on mobile already, we just send you straight to the App Store. And so just reducing friction helped us a whole ton there. And what’s really good is that it’s at the very top of your invitation funnel. So when you get a ten or 20% increase there, it’s just massive gain.
Bronson: So on somebody’s got to refer. Now if they’re on a desktop, if you just say, hey, what’s your phone number? We’re going to text you this app instead of saying, Hey, let’s set up a profile right now. You’ve downloaded that. If they send the profile on their mobile device after downloading, it was just better numbers for you guys.
John: Yeah, what we found was just reducing the number of stops. Having them put in as little information as possible helped increase throughput. And so for mobile, we got it to where they didn’t have to give us anything. We were still able to track them and like close the loop.
Bronson: And so they don’t have to create an account. You’re just using like their unique kind of phone identifier as like that’s their account and if they lose their phone, they lose their point. But that’s the only problem.
John: Yeah. We use like some tracking methods. Like there’s companies out there like Yoyo, which does tracking and closing the loop using stuff like IP and time and stuff like that. So we use methods like that to just sort of close the loop on the invite.
Bronson: Gotcha. So they really didn’t have a log in. They didn’t have a username, they didn’t have a password. They did an app on their phone and you said, name a couple of these yoyo. That might be something other people might want to look into because if you can, you know, reduce that friction, then you can get more people using it. And with mobile phones, like it’s my phone, it’s in my hand. So the technology’s there to identify me with that phone. And so maybe using that instead of log ins, like, is the way of the future.
John: Mm hmm. Yeah, exactly. That’s one of the final things we did with invitations is we added a recommended friends section of people. When you go to invite, it’s a list of people you think you should invite. And we found that helped increase invitations quite a bit because normally you come to this list where it’s just a alphabetical list of all your contacts and you might think of like, Oh, maybe Joao Mario like it, but by having a recommended section, you can help spark ideas of other people they might want to invite.
Bronson: Yeah, because most people are not going to select all. It’s just too many random people in their list and they’re not going to scroll through to find their real friends. So you can curated on the front end, you’re going to get way more invites.
John: Yeah, exactly. Yeah. And then for increasing our daily active users, one of the biggest things for us there was just using push notifications and using badging, which is that little red circle with the number you see on all the apps. Yeah. And we found that using both push notifications and badges just helped drive a huge increase in active users, especially for badging. What we tried to do was we matched you whenever we had new content in the app, so we had like new offers. And then when you opened up the app, we tried to get you as quickly as possible to consuming that content so you would have like a useful session. And badging is sort of at the very top of the funnel in terms of activity, and it just helped drive a huge number of user activity.
Bronson: Well, I know how I feel and there’s a badge on something like my number one goal in life is to get rid of it. So I’ll log into the app, I’ll look at whatever they want me to look at just to get rid of the badge. And if I can’t figure out what to look at to get rid of it, I’ll go into my settings and turn off notifications for that app because I cannot take a red dot on the corner of anything. Yeah. So is that what people do? That is have to get rid of it, which means they’re going to log in and. Look at whatever.
John: Mm hmm. Yeah. So we tried to get them to consume content as quickly as possible. So when you open up the app, not just dump even the front screen, but get you into, like, the content that we’re actually trying to get you to consume. And we also try and personalize the content. So because we’re a shopping app, there’s definitely a heavy female element. But if you’re a male user, we would try and remove as much as that female content from the notifications that you get. So hopefully every time you get notified, it’s something that you find relevant or at least interesting.
Bronson: Yeah, that sounds great. Anything else you guys did that come to mind? The ideas you implement or those are the big things.
John: Yeah, the one last thing which helped us increase sort of the LTV of users and also increase like active users and activation rates was using Geofence notifications. So because we were a location based app, that aha moment for our app was often days or weeks after you downloaded the app, which made it really difficult if you didn’t have a mechanism to alert the user when the app was actually contextually relevant. Because if you download the app on a Tuesday, chances are you’re not going to remember to use it when you actually finally go out shopping on Saturday. Yeah, so geofence notifications were a huge, huge help in overcoming that problem for us.
Bronson: Yeah. And was it that the retail stores already had geofencing set up and you just had to enable it in your app? Or was it more intense than that?
John: It was a lot more intense. Like we spent months and months like doing like geofencing, even stuff like the basic stuff like where the store is. Like we would ask the store for a list of like the outlines of all their stores and we’d just find like their data was like wildly inaccurate and off.
Bronson: So we actually using latitude and longitude is the fence. It wasn’t like a physical device they set up around their perimeter or something. It wasn’t like beacons in their store or telling it when you’re in range. It was something more from a map level.
John: Mm. Yeah, this was before like Berkeley had really become and started to become adopted and I’d be, can I come out? So there wasn’t any like hardware technology we could really use at the time. And also even nowadays, even with I began using geofencing, it’s just far cheaper to do. You don’t have to ship out these devices to these different stores and like get them to install them and all that stuff.
Bronson: Yeah. So was geofencing, you know, through the software like we’re talking about in the map and the latitude longitude, is that kind of the de facto way that people are doing it now when they’re implementing it today? Is that still the right path to go down for most people?
John: Yeah, I think for most people, unless you already have a reason to have a device that you’re installing in the store using geofencing, just doing it in software is the right way to go.
Bronson: Yeah. Now this team that you’re coming up with, all these ideas with the we’ve been talking about how many of the ideas kind of on average, if you had to guess we’re just complete failures like you would try them and you got, you know, either didn’t move the needle or moved it backwards and only ask not to point a finger at you just because it helps us realize what’s realistic for our own kind of cycles. Our own efforts, you know, is everything is success are half the things are success. What was your experience there with that team?
John: I’d say like five, 10% of things did really well and they just killed it. 10% of things met expectations, and then the rest was just a split between disappointment and utter disappointment.
Bronson: Okay, so you got a top 20% echelon that really matters. And then everything else is just, you know, that.
John: Well, the stuff that was disappointing still had some gains and those gains do accumulate and compound over time. So it’s certainly not as gratifying as getting like a big home run that increases metrics by like 30 or 40%. But if you get like five, 5% gains, those start to add up pretty quickly.
Bronson: Yeah, no, absolutely. And how long was that team working to get all those numbers? You know, 4% of daily users, 80%. You know, was it now 80% daily users, 20% revenue? Like, was that over the course of a year or two years? What kind of timeline was that?
John: That was over the course of a year. Okay. And so when we first started, no one really had much experience at growth. We were also like neophytes. But as you just like work on it, you start to get better. You get a feeling for like what works and what doesn’t. And so our first year of projects was definitely a lot more failures. But as we got better, we started getting a lot more successes.
Bronson: Yeah. So now you’re Pinterest. Let’s kind of shift gears a little bit and talk about what you’re doing there. It seems like you’re focusing primarily on their push notification strategy and you already mentioned how important push notifications were for Shop Kick. Is it the same push notification strategy just with a different company, or is it a whole new ballgame at Pinterest? How they compare?
John: Mm hmm. So. There’s definitely some elements that of what I learn from subject matter being over. Pinterest is also very different. Company like one of our core values is putting Pinners first, which means that when you’re thinking about growth, everything you do for growth to really be adding value for the user as well and not just getting it helping the company. So one of the things that we try to do here, especially around push notifications, is making sure that every time we notify the user that notification is adding value for them and is driving a useful session. And it’s not just something to help try and like boost any numbers up.
Bronson: I gotcha. Anything else that’s different or similar about the two companies?
John: They’re actually surprisingly similar because they’re both like heavily like female demographics. Also, Pinterest has been seeing a ton of mobile usage. It’s sort of like became really popular on the web at first, but now mobile accounts, I think over 75% of our traffic. So it’s sort of an interesting point because Pinterest is starting to move toward being a mobile first company and really think of mobile first and forefront. Yeah, but because that change is sort of still ongoing, stuff like growth hasn’t really caught up entirely yet. So I’ve been able to sort of like take the best learnings from working on a mobile only app for the past three years and start to apply them here.
Bronson: Yeah. And now like with geofencing, it’s, it’s an obvious play for shop kit because they literally use it in the store. What about Pinterest? Is there a connection? Is there a way to use geofencing at Pinterest or does it just not really fit in the strategy there?
John: Mm hmm. I think there’s certainly ways that geofencing could play in. So, for instance, we have this feature called Place Pins, which is basically pins associated with the latitude and longitude. So it might be restaurants. It might be like places to go when you’re traveling. And so that’s certainly an area we could use geofencing if we could. For instance, users might want to subscribe to geofencing, notifications for a particular board if it’s like a board of like the best places to visit in Spain and then just get notified when they’re near a place in Spain.
Bronson: That’s cool.
John: Yeah, because one of the keys of what we try and do here at Pinterest is we try and get users to do things in the real world. So we want to try and inspire them and get them to then have take that inspiration and do stuff in the real world. So I think geofencing can be a huge help in meeting that goal.
Bronson: Yeah, I know. It makes sense. All right. So let’s talk about this hack that I read about on your blog. That’s really interesting. Basically, it’s a way to increase referrals based on census data. Hmm. So talk to us about that. How would you increase referrals from your mobile device based on census data?
John: Yeah. So like I mentioned earlier, one of the gains we saw in invitations was having recommended friends. And how you create those recommendations is sort of a black, black art. If you if the user Facebook connects, you can at least use some Facebook data and try and do name matching, but usually no more than half of your users Facebook Connect. And so you need to find a solution for the other half. And on iOS, you’re really limited in what information you get about contacts. You basically just get their name, phone number and like maybe like email address that they have it. Yeah. So because we were so constrained on iOS and we had seen like good gains at recommended brands on Android, we tried to come up with a way of creating these recommendations out of like the small amount of data we had. So we use things like, did they have the same area? Did they have the same last name as the person? The number of pieces of contact data you had for them, like phone number, email address, so on. But one of the more interesting things we did was Shop Cake was targeted at a very specific demographic is emails between the ages of 20 to 40. Mm hmm. So to create these recommendations, we knew we wanted to target people in that demographic. And we came up with an idea that was sort of based on something Freakonomics talked about, which is popularity of name over years. And we just downloaded the census data like number of births by year with like all the popular names on that year and like number of people born with that name. And so we sort of created this database that could tell us, given a particular first name, what is the likelihood that their female and also even give us the likelihood that they’re between the ages of 20 to 40. And so we used that thing to sort of score all these names and. And recommend to the females that we thought were most likely between 20 to 40.
Bronson: Yeah. And if you’re using it, you’re probably female between 20 and 40. And those probably are your closest associates that you would want to refer and would actually get value of it out of it. Also, I mean, that is the quintessential big data kind of way to look at the world is that there’s a census data which has nothing to do with shop cake whatsoever. It’s census data. And yet with somebody thinking like a growth hacker, they can take the census data and all of a sudden spin it into better referrals to increase your metrics. And I think that’s a great example because it gets to the heart of what data can do if you get just a little bit creative with it and don’t just think, Oh, all I have to work with is the data that my device is creating, that my app is creating, but that there’s data all over the place for free laying around that can be used to do some really interesting things.
John: Mhm. Exactly.
Bronson: Yeah that’s cool. Well John this has been awesome. I have one last question for you. Here is the way I end all the interviews and basically it’s what’s the best advice that you have for any startup that’s trying to grow just kind of a way for you to, you know, instill some knowledge into the people watching this episode.
John: Mm hmm. So I think the two things I would say, one is remove as many barriers as you can to growth. So at Shortcake, we were mobile only. We were us only. You had a shop at one of the few retailers we were partnered with, and that sort of just made it an uphill battle for growth. So I think removing as many barriers as you can just makes the playing field a lot easier for you. And the other thing I would say is just when you’re approaching growth projects, just think about the ROI on those projects and just ruthlessly prioritize the things that you think will move the needle and actually do the due diligence to verify that your assumptions and what you think will work actually is true.
Bronson: Yeah, I love that second one because, you know, due diligence is cheap, engineering is expensive, so you can spend time in due diligence. You don’t spend time in engineering, so you spend the cheap time, not the expense of time.
Bronson: That’s awesome. Well, John, thank you so much for coming on growth out of TV. I love we learned about just mobile and push notifications and geofencing and the interesting hacks you’ve done. There’s definitely gonna be some inspiring ideas for people that have mobile apps especially. So thank you for coming on.
John: Thank you for having me.
Get the strategies, motivation, and in-depth interview with all the details every week!