Lenny's Podcast: Product | Career | Growth
00:25:04 · Nikita Bier
[00:25:04] S0: After building all these apps, we had these kind of, like, lingering users that stuck around and would share feedback with us, on our next app. And so there were a couple, like, there's this senior in high school that I would send screenshots of our products. And, he told me about this trend called TBH that kids were playing on Snapchat where they would post an image of a bunch of emojis, and it would say, like, I like you. You're smart. Your style is great. And you would just reply to the story with the emoji of what you felt. And I was like, this is kinda weird. You post this on your story and then people send you feedback. And I'm like, so teens are looking for this, this, like, vang... Like, this vehicle for disclosure, essentially. And I'm like, that... That's kinda cool. I wonder if you can make that into an app. We, like, had sketched some things out. And, as we were kind of sketching things out, I looked on the App Store, and the number one app in The United States was an app called Sarahah. And it was for sending anonymous messages, by adding a link to your Snapchat story. But the thing that was most interesting was the entire app was in Arabic. The number one app in The United States was in Arabic. And that was one of the most, like, the strongest signal that you could ever have that people want something. And so when I meet with founders, I often tell them, like, the way you should be searching for product ideas is this concept of latent demand, where people are trying to obtain a particular value and going through a very distortive process to obtain that value. And if you can actually crystallize what their motivation is and build a product around and and clear up what they're trying to actually do, you can have this kind of, intense adoption. And, when we saw what people were doing with Sarahah, I I also looked at some of the tweets and comments on it. A lot of people were receiving negative messages. And so I... What I saw with the game that kids were playing on Snapchat, TBH, and then Sarahah, I realized just people want to know good things about themselves, and they they don't want, like, these bullying messages that they're getting on, these anonymous apps.
00:27:31 · Nikita Bier
And I was like, well, what if instead of actually typing what you wanted to say about somebody, you, just answered polls and reauthored those polls so that we ensured everything would be... Always be positive. And, I mean, in the back of my head, I always knew anonymous apps go viral, but they always lead to, like like, these awful news stories of kids committing suicide, you know, self harm, and all that. And so I was like, I... I'll never build anything like that. But, when we came up with this new mechanic where you can only say positive things through polls, you know, who has the best smile, who's most likely to be president, and then you receive it, and it's it's anonymous, but your name is selected. What we discovered a couple of things is it made users feel a lot better. It actually solved what they were trying to do, and they also sent a much higher volume of messages. And so it was it was literally explosive adoption. Like, one school I was looking at, they sent 450,000 messages in the first seven days of adopting it. And when you look at day one, like, volume of messages sent on a messaging app, You're lucky if people send, like, three or four or something, but we were sending, like, 60. And we we couldn't even handle it, so we had to, like, we had to geofence the app because it... We... You need to scale our servers, which is actually a pretty controversial decision inside of our company because, like, why would you turn off something that's working? But at my core, I knew, like, if it's working at this many, you know, in individual schools, we could just relaunch it anytime, and it'll just... It'll it'll go viral. So, let's let's let's regroup and figure out what's happening here and then... And relaunch.
00:29:27 · Lenny Rachitsky
[00:29:27] S1: So you keep talking about how it went viral, went crazy, grew like crazy. I know that there's, like, a little trick that you came up with to help it spread. Can you just briefly talk about what you did there and to help it spread so quickly within a school?
00:29:37 · Nikita Bier
[00:29:37] S0: I I think you're referring to, there's, like, a BuzzFeed memo that... Or a memo that was leaked to, BuzzFeed while I was at at Facebook. And the the main thing we found was, like, to to be convinced to download an app, you need to see it. You need to see, like, the marketing message, like, three times or so. So you basically need to saturate an area with every kind of marketing you can. You know? So we ran ads targeted at this, a particular school to to to when we were seeding and testing these apps. And we also followed people creating a dedicated Instagram account that went to that school. Because we... I... We learned that, high schoolers identify their school in their bio. So it says RHS on their bio. And so that was how we tried to get, the entire school to adopt synchronously. We would... We'd follow them and then accept the follow backs. A big misunderstanding though, and I I get this DM a lot and people are like, I'm trying to replicate your strategy. We've just done it at 15 schools, and it's it's not working anymore. This is not the way we grew the app. This is how we tested apps. And that... There's... It's real... It's it's a little bit nuanced there. That's an important nuance because you need to get, enough intensity of adoption and density for a social network to start... To get the flywheel spinning, but the app should grow by itself after that. And people think we would just went, like, from school to school following every kid on it. Like, you can't... That... That's totally unrealistic. But for, like, the first 100 users, yes, that's how we got them. And that allowed us to know whether the product was working or not. Like, we we could get enough people on it, and then we could, with conviction, say that whether the app had legs and we wouldn't have this kind of uncertainty like, oh, did they did they add enough friends? Did we get enough people on it? Did they reach the moment because you need friends to get on... So we we wanted to eliminate that confounding variable. And so we we figured out a way to just get a bunch of people to adopt at once. And that's one thing I encourage a lot of founders to do is figure out a way to eliminate all those potentially confounding variables, so you can know immediately whether something's working or not.
00:32:02 · Nikita Bier, Lenny Rachitsky
You never wanna walk away from an experiment or test and say, well, maybe the the execution was bad. Because it takes a lot of energy to mobilize a team to test something, and you really wanna make sure your tests actually are... Provide signal. [00:32:18] S1: So your advice here is when you're testing something, test the best possible version of what that could be, whether it takes manual work or something that is never gonna scale, like test the ideal. Because that'll tell you even if this could be the best possible version, do people actually care?
00:32:33 · Nikita Bier
[00:32:33] S0: Yeah. We would try to get the... Like, an entire school to adopt just to know, like, if if everyone had 10 friends, would would would they actually derive value from this app? We also did other things like, you know, I recommend all companies do this is put live chat customer support in your app, like like, twenty four hours a day. And it sounds insane. It's like... The whole point of tech is you you don't need to do that. That's the whole point of a of a software. But then users get this white glove experience, and that eliminates another confounding variable. Like, did they think they were... Their problems are solved or they're they're treated well? But most of all, one of the reasons I actually recommend people put live chat in their app is it's the best, vehicle for getting feedback and do you... Doing user research because users will literally tell you what the problem they're having. So we we had, our person that was running this, is is the same as Michael Gutierrez. He's he's done it for all my companies, actually. He's the the the, community and customer support rep. He would, paste any interesting feedback into Slack, and then we would be like, oh, this this this user has a great idea. We we should consider turning that into a feature. So you really want your finger on the pulse as you roll these things out and so you can get a sense for what's working, what isn't, and also make users feel great and make sure at the end, they they promote your app positively to their peers.