Ankur is the largest individual developer in the history of the Facebook platform with over 10,000 applications and a reach of upwards of 200M users. Learn 5 of his favorite growth tactics and how he teaches others to hack mobile app distribution.
TOPIC ANKUR COVERS
- Five Facebook hacks that Ankur used to hack growth
- How did his Facebook Invite rates to go up 400%
- The most efficient way of growth hacking
- How to use artificial scarcity to fuel growth
- What it means to have a multiple-point entry
- The “PAC MAN” story
- The most viral concept he worked on
- 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 Encore Naipaul with us. Encore. Thanks for coming on the program.
Ankur: No worries. Happy to be here.
Bronson: Yeah, I think we’re going to have a really exciting talk, basically, because you’re the largest individual developer in the history of the Facebook platform with over 10,000 applications. That’s not a typo. 10,000 applications and a reach of upwards of 200 million users. And you’re also an online educator that teaches people how to hack App Store growth. And I know our audience is going to love that because they’ve specifically asked me for more mobile stuff. So I think we have a really good talk.
Ankur: Yeah. I’m excited.
Bronson: Yeah. Let’s jump in here. First, I want to talk about the Facebook stuff and not necessarily because I think our audience is trying to hack Facebook, but because the things we’re going to talk about right here, the psychology behind them is going to be applicable to so many things, whether it’s an app in the app store, whether it’s a SaaS application on the Web, whatever it is. The psychology here goes deep. So I want to kind of try to connect the dots for people as we talk through that, and then we’ll get to more of the mobile stuff. And we’ll have to kind of change gears in the middle to talk about that. So first, let’s talk about. Absolutely. Yeah. So let’s talk about five Facebook hacks that have kind of served you well. The first one is Friends with Benefits. So tell me what that is. What is that hack?
Ankur: Okay. So I think this started I want to say back when I was working on the Facebook platform, we there were a lot of other people doing similar stuff. I mean, virality was was the rage at the time. You know, literally entire companies were built and died in a matter of months. Know, I think when you when you think about growth, higher level, I think that’s kind of that’s kind of a very important era when it came to online growth, because the first time businesses were truly able to reach tens of millions of people. And I so I was one among the many, you know, working building initially useful applications on Facebook. But I soon realized that that didn’t really work, like a pain in the ass to grow. So, I mean, I was in college at the time. I was like, okay, let me make money on the side. So we started dabbling in viral, viral games, viral applications. And one of the first applications I built was a gift app creator, where you send your friends gifts. And gifts were essentially just it was just essentially an application whose sole purpose was to send invites to the application. Very, very stupid.
Bronson: But a gift is like an image.
Ankur: That’s just an image. It’s an image that.
Ankur: Or something. Yeah, exactly. And that wasn’t working initially for me. And I saw it working for some of my competitors. And on the surface I was like, I can’t get what I’m doing wrong. And that’s when it hit me that we were using the default Facebook friend in beta at the time, which I mean, that’s as rookie as it gets when it comes to the Facebook platform, when you have all this social data. And that kind of began a long process of experimenting with how you can take a user’s, you know, 500 to 1000 friends and choose the best friends they have. And I mean that did absurdly well like you could to to four X acceptance rates just by that one simple step.
Bronson: Like you break this down to make sure it’s really clear what we’re talking about here. So the Facebook kind of default invite or is when you tell Facebook, I wanna invite my friends. It pops up a little modal window and it just has all your friends in it in alphabetical order. It starts with a is goes to the Z and there’s all your friends you can invite and that’s broken. Why? Why is that default loader not the right one?
Ankur: So a quick caveat in the sense that, I mean, I don’t know if they’ve changed in the last three months. I think the last time I saw it, they had a slightly modified version where you have friends that use the application and friends that don’t. So that’s one level of categorization, but it’s still used to be alphabetic. I’m not sure what it was now, but at the time, yeah, it was very broken. I mean, as someone whose name was started with a I can tell you I got a lot more invites and other people.
Bronson: Got and.
Ankur: It did. Yeah. I mean and you know, the kind of user psychology for a lot of people is, okay, I want to, you know, unlock whatever it is I am. And talking about setting invites, let me just select the first 20 people. Boom, boom, boom, boom, boom. Done, sent always was the same 20, most of whom don’t care, especially when you’re dealing with the stuff that we did where the social connection mattered. Like we had a gift, we had a gifting app, right? If you’re sending a hug to someone that you’re sending a hug to your mom, my mom’s going to click any invite I send her. It doesn’t matter what I sent or if you’d like. Oh, my God, my son. Send me something. Let me see what it is. So with the world of difference, I mean, we even had a friend quiz where you answered questions about your friends. So the way you chose the friends that we asked questions about was the reason we were successful and everyone that came after us. And trust me, at least 20 companies that tried to clone us. Only one other company actually got success in the US market. And that was Badoo. Yeah, but everyone else, they just did not do what we were doing on the back end. Choosing friends smartly.
Bronson: Gotcha. So unusual in a pop up box of friends to choose from. It’s their real friends based on common algorithm. You guys have them and then they’re inviting people. They actually care to get the invite. So let me ask it is, did the number of invites sent go down or up? When you changed how they selected their friends?
Ankur: It actually went down a little like not substantially. But the which which was interesting intuitively, I would have imagined they would have sent more invites because they recognize the people that were sending it to the number of invites went down in a couple of cases they went up, but there was not the overall was not too much conclusive data. It definitely did not lead to a big increase, but the acceptance rates went up 200 to 400%.
Bronson: So that’s really where it happened, is that they sent out less invites overall, but there were higher quality invites, which means more acceptance, and then you have a better chance of actually getting over the viral threshold. Because of the way you wrote this algorithm and you based it on who the real friends are. Which I think you said before, it was based on photos they were tagged in with their friends because they are tagged in photos of your friends. They’re your real friends.
Ankur: If that’s that’s that’s easily the biggest indicator. I mean, we use secondary indicators like who post on your wall, who likes your stuff. Mutual friends. We tried it didn’t work super well, but yeah, photo tags work absurdly well. In fact, we even devised a separate standalone app once that was called Best Friend Predictor. Mm hmm. And all it did was it read the photo tags like, I predict these are your ten best friends. And people were like, Holy shit, that’s not freaking out. And that was an example of an application that grew to 500,000 people in 6 hours. Mm hmm. Just based off a single feed story, it was not that aggressive. It just gave them the option to post it to their feed. So just each user was posting one feed story at most, on average point six feed stories and close to half a million people in a half day.
Bronson: Well, it’s incredible. And like I said, I the really to go into the psychology to connect the dots for people because it’s not just, oh, I’m not building on Facebook. So this is irrelevant. The point is we’re all trying to imitate the viral loops that we see. We all want to be like Dropbox. We all want to be like fill in the blank. But the point is, make sure the invites are really high quality. So however your systems build whatever thing you’re building, make sure the person they’re typing in to invite however you Fraser or word it or presented to them gets you the highest quality invite. So even if they send less of them, more of them get accepted because they actually know each other. So use whatever signals you have at your disposal, whether it’s APIs, whether it’s stuff in, you know, data in your own application, use whatever signals you can to send out high quality invites, and then you got a better chance of getting the K close to one.
Ankur: Abstracting out one level further, I think the high level advice I give people based on this is see what you can do on the receive side. Instead of sending out more communication in general, you always get better returns to scale by optimizing acceptance. It’s like it is possible to, you know, get people to accept two times as many requests, four times as many emails. But it’s really hard to get someone to send four times more emails. Right. It’s such an inefficient way of growth. Like the most efficient way of growth is first fix the receiving side. Try and try and maximize the percentage of invites accepted. Might it be a text message, an email, whatever it is, before focusing on sending more stuff?
Bronson: You know, it reminds me of kind of the retention versus acquisition discussion. And, you know, we go back and forth. What’s more important, getting people in top of the funnel or retaining and making sure the conversion rates are good. It’s almost the exact same thing but narrowed down to a viral loop.
Ankur: Yeah, we would then act as.
Bronson: High and then worry about how many invites they send out.
Bronson: That’s great. And then the next thing, I love this one, partly because I’m reading the book right now called Scarcity. So it’s not it’s teaching me all kinds of stuff about scarcity. But the hack you have is called artificial scarcity. What is artificial scarcity? How do you use that artificial scarcity?
Ankur: I mean, kind of goes back to what I remember in college economics where, you know, in economics you buy a commodity, then make it scarce and drive up prices. The same thing applied to psychology. Essentially, as human beings, we value anything that we perceive to be scarce and scarcity does not have to be authentic. So again, you sit on Facebook with invites. The way the way I successfully got people to send more invites was putting a hard cap on the number of invites they actually could send.
Bronson: You told them you can’t send over a certain amount.
Ankur: You can order.
Bronson: Fairly even though it would help you if they did.
Ankur: Yep, yep. Absolutely. And and that’s kind of the ethical part of it. You don’t have to enforce it. I mean, I know people that have done that and not enforced it. The one thing that I think is pretty important, which which adds to your credibility is just displaying visual feedback, like always really critical, like saying you can send 20 hugs today for it. You know, we’re kind of well, but what really made it work well is every time they sent a hug, they saw that number change even more when we had the.
Bronson: Option of it.
Ankur: Exactly. You know, just like giving them a visual indicator, it’s. Yeah, I don’t know why people tend to believe a visual indicator. Like a few chances of it being true are far greater just because it’s depicted with nice graphics.
Bronson: Yeah. Now that’s great. And so let’s kind of connect the dots on this one, how people can use it when they’re not in Facebook. Some of the ideas I had is just, you know, if you have a time limit on something, you can make scarcity of time. You have two days to do this. You have 4 hours to do that. Scarcity of invites. You can only invite five people. I think about Gmail when they first came out, you know, Oh, you have this many invites. It’s in their interest to give you a million invites. Yeah. But instead they’re like, you get 20 and now suddenly they’re meaningful. These invites added, everybody wants one. So whatever your product is, think about all the different aspects of it and think, how can I make it scarce? You know, maybe you have a SaaS product, you say, okay, only three people can sign up a day. Maybe you’re only getting two times a day, but you say only three. You can sign up a day and all of a sudden everybody wants them and they see that maybe. Oh.
Ankur: Absolutely. And I mean, there’s so many companies that have recently been founded on the premise of artifice. Of scarcity of some kind, right. Groupon, that was a big driving factor in their growth. I mean, they’re essentially selling scarcity. I can’t remember a single Robinhood stock trading platform, another waitlist invite. Only every like almost every startup right now that’s successful is using scarcity in some form or the other. Yeah. And it’s kind of a very basic principle, but something that I think you should absolutely be thinking about how you can best employ it.
Bronson: Yeah, I think there should be some scarcity somewhere in your product almost no matter what you’re doing. It’s just you got to be creative. But yeah, you can put it in there, it’ll work. And the next thing is having a multiple.
Ankur: One, one, one. Go ahead. One quick last note on scarcity is and this comes more as a user rather than a growth hacker is too many people have abused it. So there’s always a fine balance. I mean, you know, it’s you don’t want to call you don’t want to call it scarcity every time to the point that people stop taking it seriously. I mean, it definitely it definitely works to a degree, but I’m not taking specific names. But there’s an online education company that sends out 75% discount two or three times a month. Right. And at a certain point, I mean, that cannibalizes your, you know, your existing market pretty badly when people tend to just have a lower brand value based on frequent scarcity, that’s entirely artificial. And, you know, people tell how artificial it is.
Bronson: Absolutely. Yeah. You have to introduce it in quantities that make sense for you. Yeah. But definitely play with it, you know, tested. So you see how it works. And like, as I say, the next thing is having multiple points of entry and I like this one. What does it mean to have multiple points of entry?
Ankur: So multiple points of entry is again something that I discovered almost accidentally. It was not something by design, but it was so powerful, first on Facebook and then later on you see how other companies are using it. Now, essentially see if you can break down what you have. You know, you can break down your your product or your application into smaller subunits that people are more likely to add. So on Facebook, we we built there was a company that we competed with that was called Brand Fall Personality Quizzes, and they had more content than us, maybe even better content than us that were first to market. Yet we beat them just by breaking down a fraction of their content into individual applications.
Bronson: So branded applications. So instead of just being personality quizzes, quizzes in general, you would have a quiz on what?
Ankur: Anything. I mean, for instance, you’re much more likely to you’re much more likely to add an application called Dr. Phil’s personality test than ad brain for personality quizzes. Right. Gotcha. You just have if you have so many applications, it’s very possible that one specific idea will catch the interest of someone the way a generic title never will. Yeah. You know, it applied. It applied to to gifting, for instance, you know, like there’s a popular application called Free Gifts, but send your send your friends a Christmas tree very frequently did better than free gifts around Christmas time because someone would not add free gifts that I’d send your friends a Christmas tree.
Bronson: The context is right. It’s meaningful to them, has the keyword they want in it for that time in their life, for whatever reason. And so they’re apt to do something with it where the generic brand may not be meaningful. Right?
Ankur: Precisely. And obviously, again, this seems very Facebook specific, but a lot of this can be applied to other companies. For instance, for instance, whenever you give users invites, it seems to many companies, you know, kind of using the same generic invite. But it’s so easy to customize the invite landing page. And I mean, conversion always goes up, you know, I mean, it can start small with a picture of the person that invited you. But ideally, if you are depending on the context of your of your of your product, maybe you guys are sharing a group or whatever you invite them to join. That group like Dropbox invites you to join a folder, not just Dropbox itself.
Bronson: Yeah, yeah, yeah. A lightbulb just went off on my head when you started saying that stuff that I hadn’t thought about, because when somebody signs up, we pull their Twitter handle, which means we have their image, which means we could create a landing page for our referral system with their friends image on it, saying This person invited you in the picture and the thought to do that. That’s great.
Ankur: So powerful. I mean, for about two months, maybe 2 to 3 months, the entire world of Facebook advertising exploded. I mean, this was later proven to be illegal. But what they did is they injected images of friends in the banners and that CPMs went up 4 to 5 over a night. Like it was unreal. Like everyone started making so much money. There were online advertising companies built and, you know, that made 30 to $50 million in three months and then just went under after a change it. But that’s just the power of that, you know that kind of social yeah that social proof can do.
Bronson: Well I think about like Basecamp, they put out a blog post or a study where they talked about the different landing pages they used for Basecamp and the ones with just people on them did the best. And the regular people, regular users, big huge photos as the background. And so how much of a more powerful if you know the person personally that you’re being pulled into it that’s that’s a great idea and one of the ways you kind of put this is like you said, you know, we’re not all building a Facebook app that has a brand that we need to break into smaller apps. That’s that was your play. But a lot of us don’t have that play, but we can use user generated content to almost do the same thing, right? Absolutely. Use user generated content to kind of imitate this idea.
Ankur: I’ll give you I’ll give you a great example and that is meetup.com, for instance. Right. I was recently giving a talk in New York that was hosted by my friends Growth Hacking Agency, Grow Hack. And I asked the crowd, how many of you how many of you are frequent users of Meetup? Maybe half of them raise their hands and then ask them, How many of you just joined Meetup to attend to this meetup? And the other half went up. So Meetup.com is using user generated content in the form of new meetups to drive people to their product. It’s the same thing with marketplaces like Udemy. You can, of course, creators create content and that content service gets people into the product. Same with Kickstarter crowd tools, you know, basically just leveraging. I mean, those are honestly like great companies to build because in a lot of ways once you get critical mass like growth is not the biggest issue you have. I mean, you obviously want to optimize and retain users, but the acquisition piece you have, a lot of people are very vested in acquiring users on your behalf.
Bronson: Yeah. So for the people, you know, watching and listening, you know, really think through what kind of content can you get users to generate on your behalf that would serve as yet another entry point into your product? Maybe it’s SEO friendly content, maybe it’s content that’s just valuable and gets shared. I mean, for whatever reason, they get to that content and then all of a sudden you have a marketing machine that’s being built for you, not by you.
Ankur: Absolutely. And I mean, I think the easiest way is if you have a product where users can in some way, shape or form make money or have a business advantage to the content that build on it, then.
Bronson: They’re going to actually bring in people. It’s going to be in their interest to even maybe buy ads and actually spend money and do whatever.
Bronson: Yeah, that’s great. Now, this last one I love, it’s some you call the Patman story. This was the PAC Man story.
Ankur: Well, I’ll tell you the story first and then kind of go into the higher level thesis. But we were running a what I flatteringly would call the first iteration of social gaming, but really was just flash games with like Facebook leaderboards. And one of them was a Pacman based game. Sorry. And like, definitely don’t have rights to that. But I think we call it something similar. But I mean, the game engine was very, very similar. And sometimes you get so caught up kind of trying to test micro optimizations and small things that during one of our brainstorming sessions, we’re just like, you know what? Let’s just try something ridiculous. And I, I thought about how cool it would be to, like, have an actual Pac-Man machine sent to someone. So I investigated feasibility. I was like, two grand. I mean, that’s, you know, you spend that much on ads some sometimes in a day, right? And it’s. Yeah, I was like, how about we actually run a contest for referrals and send a Pac-Man machine to the person that that has the most referrals. And that did something that kind of definitely hit a nerve that an iPad would not have. I mean, these are people that clearly like the game and it’s so ridiculous having a company send you a giant ass arcade machine to your door. Right. That’s awesome. And it definitely. Yeah, definitely struck it. Definitely. It definitely hit as soon as it hit. We knew it started working really well. The other innovative thing we were doing then is Facebook banned incentivizing all their communication channels. It was a very hard ban. You could not incentivize it. So what we did is we generated referral links.
Bronson: Which is like, cool, all right.
Ankur: Like old school referral links, right? Yeah. Like way back with like a referral ID and stuff. And that way we’re not liable for people sharing it everywhere.
Bronson: Because it’s not one of their channels. It’s not one of their channel.
Ankur: On their wall.
Bronson: Exactly. Gotcha.
Ankur: And that created an army of people going to like associated groups, pages, because they didn’t post it on their wall like the guys that wanted to win. They went to groups, they went to pages and they wanted the machine. And again, we got fortunate in the sense that two people got really close to each other and they ended up in this crazy war against themselves, culminated against them buying traffic for their referral link for buying.
Bronson: Traffic for the referral link to win a machine in there. They’re buying traffic for you.
Ankur: Precisely. That worked out absurdly well. Like replicating it is probably not going to work out that well. But yeah, it was just higher level. It was just trying like ridiculous shit. Like, just like too much of growth I think gets caught up in the numbers and optimizing like if they be testing landing pages, all that good stuff. I mean, I’m asked quantitative is anyone, but sometimes you got to like just say fuck it and Trish.
Bronson: You know, no, I mean, really, you know, I think the stuff that will probably bring you the biggest wins cannot be ab tested. Like I really believe that. Absolutely. We just built our community on growth out of TV and in the community. A discussion was started about AB testing and how much we’ve AB tested growth out of TV. And I said honestly, not much because I’m too busy building an entire community, which I knew was going to be a huge win. So I could maybe test headlines or I could just build awesome features. And when I run into awesome features, then I’ll start AB testing the little stuff. But I’m not there yet. I don’t need to be there yet and the stuff I’m doing is winning, you know. So there’s a big point there. Micro optimization, there’s a time and a place, but macro optimization is just trying awesome stuff and seeing what happens.
Ankur: And it’s more fun to do. And you know, when you think about when you think about why you do what you do like, it’s so much more fun just to try stuff and then, you know, stress out too much about increasing your click through rate by 1% or whatever.
Bronson: Yeah, the big stuff is fun. Now, looking back at your 10,000 apps, what’s the most viral concept you ever worked on? I’m just kind of interested personally because you’ve seen a lot like that. The thing that just you said back like, what happened? That’s crazy.
Ankur: Yep, I think I think we soon realized we definitely hit a sweet spot with teenagers worldwide, which probably informs and it’s part of this. And so at the time our bread and butter were friend quizzes. You answered questions about your friends and you. In order to unlock your answer, you had to either pay a little bit of money taken off or ask more questions. But I was doing really well. But it reached a point where everyone and I mean everyone tried adding the same engine like dating apps like Zoosk can do. We’re doing it like social interview their entire lives start ups like friendly in it’s the other one I mean it was not budget like everyone started doing this and it reached a point where it’s still the best, but we’re losing some of our magic. So then. You know, we thought about the kind of demographic we had and I was like, okay, let’s try a version of this application where you choose the best profile picture people have out of their friends profile picture album. So it was like vanity to the max. You get an application, say Bronson shows what profile picture he thinks you should have.
Bronson: Of course I’m going to click on it, you know.
Ankur: Yeah, I did that. Oh, that was bizarre. Like, we started having applications shut that. They’re just like, nothing can grow this fast. And it reached a point where we’re like, okay, we need to artificially slow this down.
Bronson: So they didn’t shut down because you were hitting limits that people didn’t think you could hit in their algorithm?
Ankur: Yeah, exactly. And we were artificially shutting it down. And it reached a point where our first step to artificially shut it down was every 20 questions show a big ad, interstitial ad and have people, you know. So the idea of being like people click on that ad would make money as well as artificially slowed down. But that was not enough. We tried it at 16 questions. We tried it at 12 questions. Finally, we realized we could grow sustainably by showing a full page ad every four questions.
Bronson: That’s crazy. So 25% of the experience was crap on purpose and yeah. Money both ways.
Ankur: Yeah, absolutely.
Bronson: That’s great. I mean, I guess the insight there is the human ego knows no bounds, right?
Ankur: Yeah. And again, I mean, the higher level is just trying different stuff. I mean, imitation has its place, but just try things.
Bronson: Yeah. No, that’s awesome. And like I said, you know, the five things we talked about, you know, a lot of people are going apply this to Facebook because I think in a lot of ways that ship has sailed. But they can apply it to so many things they’re working on. And a lot of the great hacks, they have a psychological base that’s more powerful than the actual thing that was implemented. And that psychological base can be carried over from now for the next hundred years. It doesn’t matter.
Ankur: Yep. Highly recommend influence to anyone that hasn’t read it already. Yeah.
Bronson: That’s. That definitely was the author on that one. Do you remember?
Ankur: It’s a professor from ASU. I can’t remember his name.
Bronson: Yeah, if you Google it, they’ll find it and.
Ankur: Call her Doctor Rogue or something. I don’t know. Yeah. Don’t worry about it. You can even put up here.
Bronson: Yeah, absolutely. Now, what do you think the next big viral channel is? Do you have any insight on maybe what’s around the corner that we should be looking out for?
Ankur: I can’t tell you how many times I heard this question. Like, this is literally the one question.
Bronson: I got a question.
Ankur: Yeah, it’s the million dollar question. And frankly, I think. The first thing to think about is there might never be anything like the Facebook platform again. Like that was actually just a lot of good things happening at the right time. I mean, no company is going to open up access like that anytime in the near future. But with that said, I think the important thing to do is start thinking how about how anything can be a viral channel? I mean, the way I look at it is depending on where critical you’re in, think about what other products exist in your vertical and how you can somehow leverage their audience, right? People always talk about Airbnb and Craigslist, but I think that’s what it comes down to. I mean, if you find literally any company, you know, whenever you keep up with news, you’re like, okay, Coursera now has 4 million users. Okay, if I’m building an online education company, what can I do to get some of those users as friends? For instance, there’s a company called Study Room. What they started doing is they created referral links that they encourage people to post in the course. There are classes to build a community around specific course. There are classes, right? So at that point, you maybe even build a feature in your product around a large company or a large website that already has millions of users that you could somehow borrow for.
Bronson: Well, in essence, that’s what you’re doing on Facebook just to such a high scale. It didn’t feel like it, but you were leveraging Facebook’s audience. They have captured a bunch of people, a billion people, and then you were leveraging that. But we can do it on smaller scales all the time. It’s an open LP, leverage other people’s audiences and you just make that a mantra you’ll grow. I mean, you know, somebody will meet up. Earlier, when we were first trying to get growth out of TV off the ground, I went to me ABC.com found all the organizers, organizers, cinema counts and on promo codes incentivize them to share it with their people. It’s like they already have these groups. I’m going to tap into them somehow. And I’ve been thinking that way from day one and now I’m not doing Meetup anymore, but I’m doing every day. Like, who has an audience? How can I leverage it? Who has an audience? How can I leverage absolutely.
Ankur: The other the other kind of growth hack or viral channel that I think is massively underrated. And it’s almost counterintuitive because we’re dealing with high technology being really sophisticated, but something that I think helps a lot in growth, particularly when it comes to like enterprise and things that are hard to growth, is literally brute force man, like just having assistants, interns or whatever. Just, you know, when it comes to business development and stuff, brute force might be might be the best and most sustainable growth factor. It’s, you.
Bronson: Know, as you know, consumer facing products, they can get a viral coefficient, actually make it truly viral. If you’re charging money upfront, it’s probably not going to go viral, but it’s still a business is still legit. I mean, that’s what we are. And so sometimes brute force can do some great things in that kind of world. I mean, we have these right now, you know, hammered away, doing their thing, just brute force and some hacks for us because it’s not. SAS works in a lot of ways.
Bronson: No, that’s awesome. Now let’s switch gears a little bit. Let’s talk about mobile app distribution. It’s very different than Facebook, isn’t it? I mean, I don’t even know if there’s much carryover from what we talked about. The psychology is there. But, you know, in terms of all.
Ankur: It’s a full it’s a whole new game. It’s a whole new game.
Bronson: It’s different.
Ankur: I realize. I mean, I sucked on mobile for the first few months and, you know, it’s still it’s still such a different platform that there’s there’s almost nothing in common. I think the big high level difference and this is something to consider when you design a mobile application is most growth on mobile still happens in the real world. It still happens with ideas and concepts that are cool. It’s still people telling their friends to download this. It’s still kind of what made things grow 50 years ago. It’s not specific viral channels, and I think that’s kind of a fundamental difference.
Bronson: You know, I thought about that a lot because with the with the computer, you’re at a desk, you’re in your home, you’re you’re somewhere, maybe not in a social setting, but your phone is everywhere. Your phone’s at the party, your phone is at dinner. Your phone is, you know, in bed with you at night, like, so you can share an app because it is literally with you 24 hours a day. So it’s not necessarily a knock against it. That word of mouth is so powerful. It’s just the nature of the now. It’s actually a part of you and you have your own voice so you can tell people about it. Wired With it.
Ankur: That is one factor. I think the other factor is there’s no killer communication channel for getting people to download new applications. Yeah, if there are communication so hard. Exactly. There’s no killer communication channel to enable that.
Bronson: Yeah. Well, right now the only channel really have outside of word of mouth is the app store, right? I mean, that’s quite a bit.
Ankur: Yep. The App Store is all you have and which is why most of what I teach is just, you know, milking the app store for what it’s worth. Yeah. With that said, I mean, it’s different from Facebook. I mean, you can milk the App Store for what it’s worth, but that doesn’t mean you can use my strategies and get a top five rank. Like there’s a lot more with the mobile application, including kind of just the basic stuff like how many people see this and go like, Wow, that’s a cool idea.
Bronson: Yeah, no, absolutely. I mean, there’s no there’s no replacement for just having a product that matters. I mean, you can’t hack, you know, stuff that is not meaningful. But let’s talk about some of the. Things. You know, if the app stores are so important, it’s going to be important to know the ins and outs of the app stores. So what are some of the differences between, let’s say, the iOS App Store and the Google Play Store? Are they the exact same or they are. Things are different from different plays.
Ankur: Completely, completely different platforms. And I think that’s the first that’s app store optimization one on one. Okay. As you know, Apple is actually surprisingly unsophisticated and I don’t know why it is. And again, this is something I’ve confirmed with and it’s something that evolves all the time. So it could have changed in the last three months when I last tested this, but three months ago they definitely were relatively unsophisticated, where the biggest indicator of rankings and success was straight up. Your download velocity like how fast you can get download.
Bronson: So not necessarily total download velocity which is different.
Ankur: Yep. Total downloads is. Well, that’s a total downloads isn’t that important on either platform. But Google has a bunch of other algorithmic indicators. I mean, clearly the research company and like most.
Bronson: Things they do.
Ankur: High level, high level. I think as a company they tend to overthink so much stuff. And that’s clearly evident in Android because even though you can tell they’ve put in so much more effort into filtering, sorting, blah, blah, blah, as a user, I don’t find a massive difference between, you know, the discovery engine on either. Yeah, but Google has like really sophisticated. I guess the other reason is Google also does not really have an approval process that they have to have further measures in place. Yeah, but on Google, obviously velocity matters, but know something as simple as having a large percentage of people uninstall your application. Can kill it. Can kill it. Your velocity could be rendered entirely useless.
Ankur: Google also cares about things like how how old the application is, which is interesting because it’s counterintuitive. I’ve not heard anyone bring it up, but I’ve tested enough to know that it’s definitely a thing like the age of your application matters a lot when you’re promoting it with older applications performing better.
Bronson: Yeah, I was going to say it’s probably a carryover from SEO optimization that our old websites have, you know.
Ankur: Yeah, I mean, I don’t think about it. That makes complete logical sense. Mm hmm. That might be. That might be one of the reasons your old, old keys work a lot better reviews. Better reviews matter with Google. So Google is basically just trying to, you know, put together a quality score based on how good they think their application is.
Bronson: Is and everything. The reason velocity, they’re using total downloads a little bit. They’re using on installs, they’re using the keywords in your description. All those things.
Ankur: Are they using I forgot. I even forgot the others. They’re using keyword density. Well, it’s.
Bronson: Kind of so obvious if people don’t know that, don’t even watch TV. So you have to put words in the descriptions that are, you know, going to, you know, key them into you, be you need to some degree, etc., etc.. On iOS, do they not use those signals? I mean, is that kind of what you found out? Almost.
Ankur: I really disregarded I mean, Apple uses the keywords you specify for application because you can specify keywords as well as obviously having it in your application title, which is a huge, huge indicator. I mean, for most for most searches, it’s, you know, whether it’s in your title or not. Mm hmm. And then. Yeah, then it’s velocity that primarily dictates what goes in the ranking charts.
Ankur: So, I mean, I guess there’s two. There’s two elements to it, right? One is one is how they choose applications to show up in the ranking charts, the top 25 top whatever. And the other is how they filter it for search.
Bronson: Optimize for both or you have to really pick one and just run after it.
Ankur: You can definitely optimize for both. There’s kind of search is kind of a constant optimization that keeps giving you smaller returns. But ultimately, I mean, if you’re playing in the big leagues and you have something that you want to make a large company, you have to start thinking about top 25 overall.
Bronson: I gotcha. That’s really the way to break in, right?
Ankur: It’s it’s I mean, it’s not going to be cheap, but it’s probably the cheapest way to know if you have a viable business or not.
Bronson: Yeah, that’s great. You know, and one idea that I would, you know, encourage people to try is based on our kind of insight that maybe Google is using all their knowledge of search engine and SEO and that kind of thing to inform their store, their play store, you know, think through what you know as SEO and then just imagine how would Google apply it to the store and let me see how I can engineer something to rise in the ranks. And you might have, you know, something on your hands there.
Ankur: Absolutely. The one thing to add is whatever you do on an Android application, just ensure that not too many people are uninstalling it. I’ve seen too many applications that just die like it kills your campaign. It’s such a waste. And this becomes very important when you think about media buying, because on the iOS app store, it makes sense to buy lots of low quality traffic. It doesn’t hurt you that much, but you buy lots of low quality traffic on Google. It’s a net negative. Yeah.
Bronson: No, I didn’t say, well, that’s great. What about reviews? Is that a part of the algorithms on either store? The number of reviews. The rank of the reviews? No.
Ankur: And I mean everything like Apple puts out and PR and stuff would lead you to believe that it matters. But in my data, it just hasn’t really mattered beyond like you looking at it and being like, wow, this has one star. Like, Screw this. But on Google it definitely Apple acquired Trump. And I mean, I know they plan to start using reviews better, but on Google it definitely matters. And it’s one of those things that I think it’s easy enough to hack that. I mean, if you have a decent application, you should at least have four stars. Like there’s enough ways of going about it and being smart about it. Mm hmm. That you should. You should be able to at least have a four star application.
Bronson: Yeah. Which is really important for Android and, you know, might be important eventually for iOS as well.
Ankur: Yep. And it’s one of those things that, you know, when you think about it, it’s really not hard to do. One of the strategies we’ve used that has been very successful in getting us lots of positive reviews is, again, this comes back to influence where Doctor or whatever talks about how you how people generally like to stay congruent to what they say. And what we do is we pop up a message that says, do you do you love this application? And if they say no, we say, oh, we’re sorry. You know, let us know what we could do better. But if they say yes, it’ll be like, wow, great. We really appreciate it. You could give us a review and then send them a link to the review page. So that works really well because it’s a good get your users out and the people that like your application and people that don’t and it only takes reviews from people that do like your application. So you’re like, that’s a great legal way of always getting really high, highly reviewed applications.
Bronson: Yeah, I think I just remember his name, Cardini Influence.
Ankur: That’s my boy.
Bronson: That’s him right there. And he actually has six pieces of psychology. He says the focus on congruency is one of them. Scarcity, one that we already mentioned. That’s the process they recommend people Google right now, cardini six psychological tips or something like that. All six of them are incredible and should be baked into everything you do at some level.
Ankur: Absolutely. Yeah.
Bronson: I mean, I was actually just reading them this morning. I think I was just looking over them again, just a refresher, making sure that I’m, you know, got my chops down.
Ankur: Yeah, it’s a surprisingly readable book, too, so it was a fun read.
Bronson: Yeah, that’s awesome. Now, one of the things you’ve talked about before is something called horizontal distribution in the app stores. What do you mean by that? What’s the way to think about horizontal distribution?
Bronson: Yeah. Now that’s great. It just comes back to multiple points of. Three. User generated content. You know, not a big brand, but many brands that can be found in a lot of different ways. It’s yeah what we said it worked on Facebook and it’s not going to work the same way, but the idea is the same. Just allow people to find it the way they’re going to find it. If you end up.
Ankur: With growth, like take advantage of the fact that search is so powerful and see how you can get the most number of keywords. Yeah. And I mean, I accidently discovered this in what was a really fun experiment that got shut down eventually, but we built an entirely algorithmic way of launching new applications. So we once we scraped wiki quote which were legally allowed to do and build 100 and something smart applications based on TV shows where it would show you a quote by a random character and have to guess who said it. So it’s kind of like a stupid randomly generated application. We launched 100 and something of them algorithmically. And that was absurd. I mean, that was like Facebook levels of growth on mobile. So definitely got shut down soon after Google claimed that it was in violation of TV shows copyright, which, I mean, legally isn’t an issue because it comes under the Fair Use Act. Yeah, but either.
Bronson: It was a system.
Ankur: Yeah. They didn’t like the automated system. And it was funny because three days after that they changed their policies about automated systems and they sent out an email to it and just pretended it was completely coincidental.
Bronson: Yeah, right. Yeah. This has been great. Oncor, I have one last question for you. What’s the best advice that you have for any startup is trying to grow? What would you tell them?
Ankur: So the one thing I see a lot is so many companies die because of a lack of growth. Like I would argue that’s probably the biggest reason company companies die. Either that or not getting along with your co-founders. And at the same time, I see so many people hesitant of being aggressive. And if the one kind of higher level maxim that I’ve always stuck to in the context of growth is it’s always easier to ask for forgiveness and permission. You know, like sometimes if you’re a startup founder, chances are you’re kind of have a little bit of a gambling instinct. Go with it. You know, you are in a very low probability game and it’s all about taking calculated risks. And considering the biggest reason your company will die is the fact that it doesn’t grow when it comes to grow. Just try stuff like take risks, push, push the boundaries. I mean, obviously I’m not saying get into trouble, get banned from the App Store like there’s a balance, but be aggressive.
Bronson: Yeah, no, I totally agree. I mean, the way I see it, there’s ethical lines you don’t want to cross. Yeah, there’s legal lines you don’t want to cross. And there’s a lot of other just social constructs. Cross them. If it’s not absolute and it’s not illegal. Cross. See what happens is even grow your startup or die.
Ankur: Absolutely right. Yeah. I think the ethical thing is important. If it’s something that you wouldn’t tell your parents, you probably don’t want to do that. But if it’s something your parents will understand and they’re like, Oh, yeah, I guess time to do it. Go for it.
Bronson: Now, this has been great, uncle. I mean, there’s been so many, you know, things we’ve covered. I think people are going to a lot of this interview and just thank you so much for taking the time to come on growth TV.
Ankur: Yeah. No, it’s been fun, man.