Josh Elman, partner Greylock & product master (ex-Twitter, LinkedIn, FB) shares secrets for success

Today's episode of This Week in Startups is brought to you by SnapTerms, online legal protection made simple. Visit snapterms.com and enter the code TWIST to receive 10% off. And by InVision. Find out why so many hot startups are using InVision to prototype, present, and collaborate on design in real time. Sign up for a 90-day free trial today at invisionapp.com/twist. And by AWS Activate, the Amazon Web Services startup program. It's easy to start and scale your business with AWS. Visit aws.com/activate. Hey, everybody. Today on This Week in Startups, we continue with our incubator series. Yes, 7 companies are spending 12 weeks with us. They're all going to launch at the Launch Festival on March 2nd, 3rd, and 4th. And in tonight's episode, it's week 2 of the incubator, and you're going to meet Josh Elman. Josh Elman is a venture capitalist at Greylock Partners. Greylock has Reid Hoffman from LinkedIn, and a bunch of other really smart cats over there. But before he did venture capital, Josh did a lot of the growth growth, and a lot of the workflows, and the signup flows at a little site called Twitter, which had a lot of unique things they needed to solve for. And he gives a very detailed talk that you will learn a lot from. Stick with us.

That's what it's all about, man. They said money is the root of all evil. What?

Funny how it feeds my people. Yeah. We ain't gonna live like equals until we get the money, spend the money, and defeat you. Yeah.

Money is the root of all evil. Funny how it feeds my people.

We ain't gonna live like equals until we get the money, spend the money, and defeat you.

You know, the first thing is, you know, I'm just gonna talk about growth and just sort of how I think about growth. And, you know, I've really enjoyed hearing all of your talks, but you're probably sitting there wondering a little bit of, you know, who is this guy? And, you know, I started just as an engineer. I went to Stanford and started as an engineer and worked on this thing called RealPlayer. I list my kind of career as a is my network, because I think all the stuff that I've worked on, and kind of a lot of the stuff, even we talked about what gets really big and interesting, are networks. Networks of people, networks of marketplaces that bring lots of people together. And those become the companies that really become kind of the lasting legacy, really foundational companies of our time. And that's kind of when I think about growth, and when I think about everything of what we're trying to do to get something from a very small and beautiful early idea to something very big. It's about compounding your network of people, of users, of activity, of a team that you're building, and kind of just building this compounding network. And yet, that's what we do all of our lives. So I show my life a little bit as kind of my network, and it's interesting how it pools into lots of different pockets. But, you know, I went to Stanford and then started as a Windows engineer on RealPlayer. Do people remember RealPlayer?

Oh, yeah.

So like, look, it used to be the thing that you watched, like all the audio and video on the internet came through it. And when I joined them, it wasn't a public company, and it had the vision of, the internet should be full of audio and video, which in 1997 was a more novel idea than it is today because it's now obvious. But it was great to get to work on that, and it really succeeded. But Real didn't for a variety of reasons, most of which had to do, I think, at the time with customer empathy. It's very hard to be the product that gets to massive scale when you have a lot of people who go, I had to click through 17 install screens, and it was popping things up on my computer. They don't fall in love with your product. They kind of fall in hate, but they have to use it. And those things, can win for a period of time, but don't win for the long term. And I'd go to meetings where I would say, oh, we should do this. And they'd say, oh, well, the Excel spreadsheet shows we'll make more money if we do this. And I'd be like, well, how do I put zeros on the end of your spreadsheet to show you that your idea will go to zero? And I was a dumb engineer and didn't know how to use Excel to prove that in their models. So at that point, I moved down to the Valley. I actually started business school, but I dropped out when I heard of a couple of companies called Friendster and LinkedIn that said, hey, if we connect all the people in the world and we get them all connected, what could we build on top of those connections, whether it's dating or reinventing jobs or everything else? And I'd done the same program at Stanford as Reid Hoffman, who's the founder of LinkedIn. So I sent emails to jobs@linkedin, jobs@friendster, and was lucky to get a response back from LinkedIn, and ended up joining them instead of dropping out of business school. There was about 15 people in the company, and we had this theory of, how do we get people in this network? And we called it growth usage revenue, GUR, because we had to figure out how to get people just in the network, even if they didn't do that much, so that the network would expand, then— We'll figure out the products that will really start to engage people with more usage, and then over time we'll be able to make revenue. And I'll have a little bit of a LinkedIn anecdote or two kind of through this talk, but of just stuff that I learned. And this was 2003, or I joined very beginning of 2004. It was in '04 and '05 when like virality on the internet was not like virality is today. I went to a company called Zazzle, which is on-demand custom products. And again, I was like captivated by this vision. Now, not just connecting people, but if we connect people through a marketplace where you can design, design anything, and somebody else can buy it, and it gets made that day and shipped out. We're connecting people through actual commerce, which I thought was, like, an incredible idea. And I think Zazzle isn't as big as I had hoped it would be. It's still doing, you know, really well down in San Jose. They'll be one of— probably the number one shipper out of the San Jose airport this holiday season.

Right.

They have a big factory where they're just making stuff, you know, here locally, because it's on demand, just in time, made that day, shipped out. But it was a great experience. I learned a lot about the other side of the internet, like, not just media and ads and, you know, audio, video, and social networks, but about, like, commerce and making things, and, like, bits to atoms. But as I kind of saw where that was going, I got really back excited about that sort of social platform world and connecting everybody. And Facebook was getting pretty large, about 500 people in early 2008. And I managed to jump on board and got to work on Facebook Connect and got to kind of help lead that launch and got to really see, like, how do we now take this power of, like, a social network, which by that, you know, 2008, Facebook was getting larger. It was still actually much smaller than MySpace when I joined. And we'd go to other companies and be like, please put Facebook buttons on your site. And they'd be like, Facebook? OK, you cute little thing. That's just for college kids. Why would we do that?

Right.

or a website that doesn't have Facebook integrated in some form is much, much harder, which is a pretty good testament. But that was incredible learning. And I got to work with so many other companies and help make their sites social and help make their sites viral. But I was working with everybody else, and I was like, I still want to just go work on a product, and fell in love with Twitter. And it took me about 6 months to get to know the Twitter folks before they decided to hire me. In the fall of 2009, there was about 80 people in the company. And Evan Williams was CEO at the time. He and Jason Goldman, who's the head of product, said, hey, everybody signs up, but nobody sticks around. Can you help us fix that? And, you know, I'd gotten to do a bunch of viral stuff at LinkedIn. At Facebook, I've been working on the Connect platform, you know. And, you know, so I worked on problems more than I worked on growth. Like, it wasn't like, oh, I'm a growth guy or a growth hacker, but it was the most important problem for all of these companies I'd worked on was like, how do we get bigger? And so at Twitter, I got to kind of put that in action and said, how do we actually just make Twitter bigger? I'll talk a little about some of the stuff that we did to do that. But that's been kind of my career. Twitter went through its whole It's a whole own sort of soap opera. There's books about it. And I ended up getting kicked out in the summer of 2011 and thought, man, I've gotten to see these amazing journeys along the way. I can keep doing that, and I just might be this guy who's like in Groundhog Day, where I'm like company after company after company. I work for a couple years, then it gets bigger, and then I get ADD and decide to leave, or something else happens, like what happened at Twitter. Or maybe there's a job for somebody like me who's actually helped a bunch of companies do this and gotten to see the insides of some pretty amazing stories and said, oh, let me, try my hand at this venture capital thing. And I'd known Reid Hoffman, who's one of my partners at Greylock. And, you know, I just started talking to some Greylock folks, and they said, hey, why don't you come hang out here? And, you know, fast forward 3 years, and I'm a partner, and I've done investments in Medium and Jelly, which is Medium's Evan Williams publishing platform. Jelly is Bizstone's sort of Q&A app, and they're building Super now. SmartThings, which is in the connected home space. And they're trying to build a platform that connects all the things. Just like I'd spent the last 10 years connecting people, I got really excited about the Internet of Things. What happens when you connect all the things? We actually ended up selling that company to Samsung. But I'm still looking at a lot of new investments in that kind of connected hardware, connecting things space. And then one stealth company. And then I'm also very involved in Nextdoor, which Greylock has an investment in. I spend a lot of time with that company. I'm very excited about what they can do. But that's, you know, still about connecting people locally in their neighborhoods, and building much tighter neighborhoods, and letting people communicate and share.

So that's a little about my past. Here's the thing. Listen to me. If you don't have a terms of service, you're gonna get— and a privacy policy, especially in this day and age— you are gonna get hit up by all these ambulance-chasing law firms. I don't mean to scare you, but the fact is, if you raise a little bit of money, you're now a target. You gotta have the terms of service, you gotta have your privacy policy all buttoned up. Now, this can cost a lot of money. That's why you wanna try SnapTerms. SnapTerms.com is, uh, a great way to get started in having a simple, affordable, fast terms of service up and running. We use it. This Week in Startups startups.com/legal. This Week in Startups.com/legal, and you can see our Snapterms. And it starts at only $299, and they can customize it. If you're commerce, if you're, uh, I don't know, uh, social network, whatever, they can take the terms of service and just plug and play little different parts of it. In other words, you're gonna get this done today, quickly, for just a quick hit on your credit card of like $299 or a little bit more, and you get 10% off if you use the promo code TWIST. They can even do custom solutions if you have something more complex. Go to snapterms.com and avoid avoid costly litigation, have a proper terms of service and a proper privacy policy. This is serious stuff. You want to get it done, and you want to get it done early and make it clean and tight. Go ahead and try snapterms.com. Use the promo code TWIST, and follow and thank @SnapTerms on Twitter. Okay, let's get back to this amazing program.

If we think about growth, we forget that, like, growth has always been really hard. And it's still hard today. It's actually easier today than it ever was. But this is what growth looked like in 1995. This great company named AOL made a lot of CDs, and they sent a lot of CDs out to everywhere. And like, this was state of the art for growth. And in fact, it was revolutionary. And AOL got really, really, really big at the time because of, you know, a heck of a lot of CDs. And we think about growth today, you know, we think, oh, virality, it's just perfect. I tell one person, they tell 3 people, each of them tells 6 people, boom, like the whole world works virally. It's basically true, but it's really, really hard to engineer, and you really only engineer it with a good product. And I meet way too many people who think growth is this, oh, let me just get featured in the App Store. Please, Apple, feature me. And they think that that'll be how to grow. And like, that's totally not the answer either. It's not about getting the article in TechCrunch. It's sometimes about getting Jason's first tweet. But, you know, that only gets you started. Like, growth is not about getting featured in anything. Growth is about building something so powerful that— I call it the atomic unit of growth. is one person telling another person about the product that gets the second person so excited they can't wait to go try it. And, you know, now with mobile, you know, it used to be I'd tell you about it over lunch, and you'd have to, like, write it down, go back to your computer, type in a URL, maybe you'd sign up or try it, or, you know, send an email, and it would be there when you got to the computer. Now, if I'm telling you over lunch, you can be like, hold on, download the app on your phone, get it all set up right there.

Yeah.

That's kind of the atomic unit, and you want to, like, do all your things to, like, plan for that so that it really works. So when I think about growth, I think there's, like, 4 kind of key questions that you think about for like the over— the whole kind of arc of your growth cycle of how you actually get people. You know, the first thing, we talked about this a lot tonight, is that problem. Like, what do people want? How do you simply describe the problem that they have, and that why your product actually just solves that problem for them? You know, we talked about a couple of things to help you go out more in the world, and like being able to actually like everybody— if you're like, would you like to go out more and have more fun, or stay at home on more Friday nights? Like, everybody would say, I'd love to go out more. You know?

Mm-hmm.

Whether you have kids, or parents, or whatever it is, or, you know, you're, You're single or whatever, or you're in a relationship, you'd always like to go out more, and yet, you know. So that's kind of the first thing is like really understand that problem and then go, my product solves that. It gives you that, you know, buy this, this, and this. The second problem we have is, let's say you actually have an app that solves that problem. I call it inception. How do we actually get this bug in your head that there is some product out there that will solve that problem, will incept you, and will actually get you excited to go try it? And the third thing is adoption. Okay, you're incepted, you're excited, you're going to come try it. How do you now get going? How does the product actually get you from that moment you've heard about it and you're actually there to try it to actually get going? And I'll spend a little bit of time talking about that. That's what I talked about at the Launch conference, and we had a good conversation there. And the last one is retention. Once I've got you to actually try it and get going, how do we now make it a habit? And so let's talk about kind of the problems that people have. And I think, you know, we look at Google right now, and it's like this beautiful Spartan product, and you just know that when you go to Google, You have a problem, you want to find something on the internet, you type it into Magic Box, and it'll give you a bunch of links that tell you that. But, but, you know, I see so many products and developers that are like, well, Google's designed this way, so I'm going to design my product this way. What you have to remember is in the early days of a product, no one actually knows the problem that they have, and you really have to design it to help them solve that problem. You know, and this is search the web using Google. You didn't know what you're doing. You want to search the web. There's 25 million pages. I'm feeling lucky. I want 10 results. You just, you know, get Google updates. Like, but we laugh, but when you're brand new, you actually have to explain more. You have to educate people. This beautiful aesthetic design, like, that doesn't help people know that you solved that problem. It doesn't actually explain it to them. And so, I don't know, I just think it's really important to realize that, like, the problem that we're solving, we look at these companies that have done an amazing job solving them, and we're so used to how they're designed that we forget that, like, everyone had to start somewhere like this. Or even better is Uber, right? Like, people forget, like, I mean, Uber is incredible now. Like, I need to get somewhere. Like, I was a little bit late here because I was like, oh, I can keep running late because I'll just call an Uber and it'll just get me here right on time. And it took a little bit longer to come, and it was surge pricing and bad traffic out there. But, you know, at the very beginning, like, look, it says one-click car service. This is actually from the— this is the first picture ever posted to Uber's blog. So that's how they're like marketing themselves back then, you know, one-click car service, you know, get started. You know, and then they used to run their own group Like, they were running deals. They had to, like, really do all this stuff to say, hey, you probably need a ride. And, you know, we're gonna run discounts, we're gonna do everything else. So I just wanna remind everybody that, like, you know, these companies that, like, feel iconic today, like, they started by figuring out their problem and figuring out how to describe it, and really just trying to do one thing really well. And they were kinda ugly, and they did a good job sort of describing it. You know, and then, and I'll talk a little about the Twitter story, 'cause I think Twitter is sort of this often maligned company of, like, they don't know what they're solving. And it, like, keeps coming back over and over again. But I think, you know, Part of the reason that Twitter got as big as I think it is right now, and it's this real big force in the world, is we did a lot of hard work to help kind of identify that core problem that we thought Twitter solved, and got at least a lot more people excited about it. But when I joined the company, homepage looked— this was like the homepage before. But we actually asked you. We're like, what is Twitter? Because we didn't really know. But it described itself as a service for friends and family to communicate and stay connected through the exchange of quick, frequent answers to one simple question, what are you doing? And if you think about that, that sounds like Facebook. It doesn't actually sound like a problem most people have. And it was actually confusing. And most of the power of Twitter actually wasn't being able to communicate with your friends, family, and coworkers. It was actually being able to communicate with the world and figure out what's going on. So we didn't even do a good job of the problem that maybe Twitter really did solve, kind of helping everybody else. And then it changed, the website changed to this. Share and discover what's happening right now with anywhere in the world. So that's a very active thing. You can share, you can talk, you can broadcast out there, which again isn't really the problem that Twitter solves for most people. And even more, we had a big search box there. if you search on Twitter, that's like the worst thing you can do because that's not sharing, and it's actually finding a bunch of random stuff where if you don't know what's going on on Twitter, you are so freaking lost. You know, so that's really the problem. So, you know, if you typed in I don't get—

I get drunk.

We actually got— we actually beat I don't get drunk, I get awesome in Google suggest for I don't get. So you can just tell that like we had real issues. Twitter kind of helped people just— but I think it's really about understanding the problem that we solved. So this exercise, and we said, look, who is Twitter for? Like, who really has the problem that we're solving at Twitter? Is it someone who tweets? Like, no one really knows they want to tweet. That's not where we started. Is it someone who wants to stay up to date with friends, which is what that old homepage said? Like, no, that really isn't what Twitter was about either. Is it someone who cares about real-time news? And people started to go, oh yeah, that's kind of closer, yeah, of course. And then we asked around the room, who cares about real-time news? Not a single hand was raised. Because nobody thinks about themselves as like caring about real-time news. And then it was like, you know, then of course this is the beauty of what the press said about us. Someone wants to know what their friends had for breakfast. Which is like, again, totally not, totally not us. Someone who wants to tell everyone what they had for breakfast, which was like the even worse thing at Twitter. So we put all this together. It was really clear that we were kind of in a state where we couldn't really tell you like who wanted Twitter and what was it for. And we came up with this phrase that I really liked. Someone wants to know what's happening in their world. Oops. Sorry.

Okay.

Dropped the clicker. Someone who wants to know what's happening in their world. We realized, like, if you ask everybody else, like, do you want to know what's happening in your world? Do you want to know what's happening in your world? Like, that initial statement, like, everybody has that problem. They would love that. And if we can then say, well, Twitter is the best way to tell you what's happening in your world, that's great. Now, we still have a lot to prove on that. But if you think about the people who really are most passionate and love Twitter and use it all the time, there's a very small segment who are active broadcasters. Most people will say it's something around this of, like, I just, I stay connected. I know what's going on all the time. I know what's happening. And I think that's the power. But it took us a lot of exercise. And when you think about each one of your products, like, what is it? Who is it for? What do they want to do? And what is it that you are the, I want to do that, this is what you should do. And then Dick tweeted out this mission in February 2011. Our mission: instantly connect people everywhere to what's most meaningful to them. And that kind of sets a mission statement for the company. I don't think that's the same thing as, like, the human problem exactly that we're solving. It's a little more fluffy than that. But I think it's important to kind of have these mission statements. It's like, if we achieve instantly connecting people everywhere to what's most meaningful meaningful to them, we've really achieved this really massive, important thing for what Twitter can be in the world. And if you look at kind of where the Twitter homepage got to, this is like vintage 2013. It says, welcome to Twitter. Find out what's happening right now with people and organizations you care about. And all of a sudden, you show up at the homepage. We tell you what that promise is and that problem that you might have, and you start doing it. And then from the entire signup flow, we try to convert you on that. And so we're in a much different place than staying connected with friends, sharing what's going on, a big search box. And so that's kind of the initial way that you just think about, like, at the macro level, for each one of your companies, that 30-second pitch, the 30-second pitch over lunch, that like, hey, do you ever care about this? Do you ever want to do this? This product is the best for you. Here's what it does in the simplest way. That's where you got to start, because that's kind of that atomic unit of sharing that causes all growth to happen sort of after that. And so kind of once you're already there, now let's talk about how to get that atomic unit to like spread a little bit better. I think, you know, the key is really to have it spread, you know. And so, you know, where I learned a a lot and really cut my teeth at LinkedIn was on viral invites. And we used to spend a lot of time testing the virality of our invitations. And it was kind of a crazy set of early exercises because we had to figure out how to make LinkedIn viral. We had to figure out every person who signs up, how do we get you to invite more people to your network? How do we get that person to invite more people to their network? And on and on and on. And, you know, like, the really interesting thing we did was how we tested. We didn't just test, hey, which invitation has the best response? Which invitation has the best click-through? Which invitation has the best feedback? Mm-hm. But it was really which invitation leads to the most long-term virality, or at least one more level down. And so we would actually test, like, which invitation do you send that actually increases the likelihood that everybody else who gets them invites more people? So how many 2 generations away are we actually getting instead of one generation? I'm just going to read you the text of this. This is, by the way, what LinkedIn looked like back then, which is like a— again, we've come a long way in design. But it says, so I found you while I was looking around the network. Let's connect directly. I'm happy to help you with requests and forward things incoming to you. It'll probably make both of our networks bigger. And that last sentence, it will probably make both of our networks bigger, sounds dumb, kind of a throwaway. It's a little bit awkward. Turns out that the number of people who invited after they received this invitation and clicked through to confirm were higher than when that sentence wasn't there, because we've already incepted you with this idea that a bigger network is better. Hey, you should invite more people to add to your bigger network. Isn't like, there's like this like crazy silly learning that like the words you use and the things you want to kind of incept and implant in people actually pays off in sort of downstream behavior, and that is really important to measure downstream behavior. Too many people do a funnel analysis and they're like, oh, my funnel got to here, where the reality thing is like, did they do something 2 days later? Which funnel actually got them more likely to do that or got another behavior? You know, obviously SEO is great, and people, it's funny now that like nobody thinks about SEO-driven growth, because people think SEO is totally kind of dead. But there's a bunch of companies between 2005 and 2010 that got really, really big because of SEO. Yelp, TripAdvisor are 2 very large public companies that were totally formed on the back of SEO. But the real trick to SEO is you're tapping into the nice intent stream that Google, and maybe just Google as the master search engine, have created. And if you can figure out where to tune in and convert people who land on you to come use your product instead, that's great. What Yelp managed to do was every time you would look for a small business, Yelp would always come up first, second, or third. You'd click through and you'd be like, that's a better page than everything else. And after doing that like 100 times, you would finally go, I should just go to Yelp directly. And then, hey, Yelp's so good, hey, they have a mobile app now, I should just put Yelp on my phone. And so you kind of like that, you know, that 100 times you would land on a page before you do it, or TripAdvisor, you're like looking for a hotel, or hotels in, you know, Barbados, and you'd end up there and you'd be like, oh, that page is pretty good, why don't I just go there first? And so you really want to think about, like, as you're making content, and, you know, I think it's actually going to become more important in the mobile web, not less. We're kind of all focused on Facebook and Twitter distribution right now, and I think search is about to go through some massive under— you know, like, massive overhauls, and, like, it's going to be really reinvented on mobile over the next couple of years. Figuring out what content you have that catches people, that they keep hitting, that they realize it's a better experience just to come here first or to come to your app is really important. And you need to track it. You need to go, I've seen this person 11 times. You know, the 12th time, hey, I've seen you 11 times on a whole bunch of stuff. I think you probably like restaurants in San Francisco. Maybe you should just download my app. And you know what? You'll start to convert people one by one by one, and then these other viral invites and other techniques, you know, will start to work. So I think there's a lot to do about SEO, but you have to think about it right, as you're tapping the intent stream, you're trying to convert people to come to you first, versus just, oh yay, Google gives me traffic, because that doesn't ever lead to real long-term success.

I am so happy to read an ad for InVision, because I use InVision every day. Yes, I'm launching new products constantly. Inside.com is one of my latest products. It's a news app. I have a very talented designer named Chad. I've got a very talented product manager named, uh, Andrew. I have a very talented, uh, web architect doing our API named Blake. I have Adam, who is my CTO. I've got all these constituents, including my investors, like Roelof Botha from Sequoia, Matt Coffin, uh, John Miller from AOL fame, I've got all these constituents. And we're constantly talking about how to make the product better. And how do you do that? Well, you show prototypes. You show mockups. You show the next version. And what happens? Amazing discussions ensue. A prototype is worth a million words. Well, what happens when you send a prototype on an email chain or in a chat room? You lose all those great comments and all that great insights that your team has, from investors, to designers, to product managers, to technical people, to editorial people, to marketing, PR people. It all gets lost. It goes into the ether, and you get into the reply-all nightmare. You don't want that to happen. You want to use InVision app. And let me explain to you how InVision works. Your designer makes some mockups. They put it in InVision. They invite people to a link. You take the link. You say, add to my desktop. You add it to your desktop. Then you take your phone out. You click on the icon, and it loads magically the next version of whatever you're working on, and it looks like an actual app. And when you click on stuff, it goes to the next image in the mockup. This is a revolution, because instead of having the developers waste weeks building stuff, you can have the designer build a bunch of different experiences, send them to you, and for a 10th, a 20th, a 100th of the cost of your developers building it, you can actually get a feel for what this app is gonna be like. That's the magic of InVision. Go to inversionapp.com/twist and sign up. inversionapp.com/twist. You get a 90-day, uh, starter plan for free, and you can do all kinds of things, like put comments and You can, uh, put sketch commenting on it, real-time comments. You can do all these really great things. And if me using it every day is not enough for you to try this, how about the fact that Airbnb, Evernote, MTV, Adobe, Box, and Zendesk are using it every day? Those are companies that are much more successful, at the current time, than Inside.com. But there's a reason they're successful. The reason those companies are so successful is solely because they use InVision App. If you use InVision App, I will guarantee you that your company will be worth $1 billion. I'm kidding. But it is a big part of getting there. If you want to build a unicorn, you need to use Envision App.

Period.

End of story. Envision App is amazing. I use it every day. And when they, uh, started advertising on the program, I was like, this is the greatest thing ever, because I use this product every day. I don't even have to look at the copy. You see me glance down here and look at the copy on the page. I'm like, I know all this. And plus, I know how the product works. I can tell you all the details. There's a little button that says comments on or off. So you can look at the designs with the comments on, comments or the comments off. When there's a bunch of comments, I can come in and say, wow, those are all great comments. Here's what we're going to do. Thank you for the feedback, constituents. Then I hit resolve. Discussion over. That's the magic. I could have like 8 or 9 or 10 threads going on about different aspects of our product, and then boop, boop, boop, beep, beep, beep, boop. I resolve them all, and everybody can move forward. It's made me at least 50% more efficient in working with my designer, and 100% more efficient when working with the overall team. Go to invisionapp.com/twist and thank @InvisionApp on your Twitter. And if you're building an app, just go there now. It's an amazing product. And thank you to my friends over there for making a great product. Honestly, sincerely, I mean that. I couldn't do what I do without your app.

So social, I mean, this is, this is sort of like, I think, I feel like this is what I'll talk about the least now, because it's sort of everybody gets it. Get other people to tweet stuff out, and they'll land on your page. And then hopefully, they'll download your app, or they'll get invited to it. But I mean, Instagram is like the master of this. Like, Instagram grew almost solely in the early days on the backs of Facebook and Twitter, because every time you would go to Facebook and Twitter, you'd be like, oh, that's a really cool picture. Oh my God, how'd they get that? Oh, they have Instagram. Oh, that's like a much better way to take pictures. And Instagram made it so easy when you signed up to push all your pictures out to Twitter and Facebook, and it created this very, very tight loop, but it became iconic. So every time I went there, I was like, oh, that's an Instagram picture, and I wanna have something like that. And I think I see a lot of people now who like, they post pictures into the feed, or they post other things, and they're not iconic. I couldn't tell you when I've seen, like, the same thing from somebody 5 times. Oh wait, I've seen, you know, a new, you know, event invite from this company, or a new, you know, picture, or picture album that looks very, very special. And so I think there's a lot of work that we can do, you know, in social to kind of make something iconic. It's a little bit harder now. It's a little bit noisier. I do think it's a little bit overdone, and there's just so much going on. And getting me to tweet stuff out is just like, you know, Facebook stuff, like, it's gotten a lot harder. But if you are gonna go that route, make sure it's iconic. Make sure it's something that every time you land on that landing page, you're like, I want something like that. I understand where it's coming from. I understand why it's special, as it features the content. And then PR comes back to my word of mouth stuff in a very deep way. I think Jason's incredible at PR and probably can teach you guys way more lessons than I ever could. But understand that what you say can plant the ideas in the PR's head, and then what they say will plant ideas in everybody else's head in a very important way. this inception concept like totally at scale. And I use this, this is an interesting story for Twitter, just 'cause I lived it. But you guys remember when Conan O'Brien was like fired from NBC and he went on tour? And he like, literally he did like one tweet and sold out like the concert tour like that afternoon.

Yeah.

And so there was all this great press that said, hey, Conan O'Brien with one tweet sold out the concert tour. And everybody would go read that article and be like, oh, that's cool. Like Twitter's cool, but I don't have anything to sell, so why would I use it? I was like, wait, that's actually the wrong story. The story should have said, Hey, did you hear that if you were on Twitter from 2 to 4 PM today, you might have had a chance to buy a ticket to Conan O'Brien's tour? And if you weren't on Twitter after, if you didn't get on Twitter, you weren't even on it, after 4 PM it was all gone. That's like, oh, wait a sec. Wait, I actually need to be on Twitter all the time. I need to be checking it. I need to be following people I like, like Conan O'Brien, 'cause I never know when this exclusive thing's gonna come out. And like how you tell the story, and like you actually, the way you talk about your company to everybody, the way you talk about it on your website, the way you talk about it on your about page, that tells the story to everybody else. But it's the way the pitches happen, the way each of the stories that you guys do, that's what gets left in people's heads. That's what's either gonna make them have FOMO, or, you know, like, I need to be on that product, I need to be using that, or I'm totally, you know, totally missing out. You know, I didn't know that I could have gone to see that show, or gone to, you know, see that, you know, fun activity, or, you know, gone to actually eat there when I thought it was totally sold out. So I think there's a lot of different ways to think about it, but you have to own that and tell the story. And the last thing is in person. I think Snapchat, people don't understand, but like Snapchat's almost entirely spread because people are like hanging out with each other, and they're like, you're not on this thing? Like, I totally want to send you pictures. Just download it. And they would download it right there, and they would show each other how to use it, and they would add each other's, you know, names, and they would add each other's things, and then boom, they're friends. They start doing stuff. Like, Snapchat's actually interesting. There's a lot of stuff that's kind of hard to use. It's a little bit opaque if you don't know the secret, if you don't know the Konami code, if you don't know you can turn stuff on in like the super special extra settings. You actually don't get access to some of the features. If you don't know that, like, you can actually swipe up and to the left or whatever, you get the white pen or the black pen. Like, it's actually really, really cool that they've actually built this mobile-first product that just assumes that's how people share stuff. And that's how people talk about it. And that if you do master the secret, you actually want to tell somebody else the secret. And you want to like, oh dude, you didn't know you could do that? Pull out your Snapchat, and you're like, ooh, look at this. And then they're like, oh my god, now I can do this thing. Snapchat's actually engine— like, I don't think they're like actively like, I'm gonna put stuff in the product that does this. But they're like, just because they're, you know, the founders and the team there is just so good at thinking about how does this generation, mobile-first generation, think about and talk about apps and products. They're just building that stuff in from the ground up. I think a lot of people miss that. And it was impossible before mobile, because I could— unless you walked over to your coworker's computer, or you were over at a friend's house, you're like, hold on, show this. That was so rare. And now it happens 100 times a day. So I think it's a really important thing that, again, engineer that kind of stuff into your apps. Think about it. Understand that's how people are to do it, make it awesome. Make it fun. If your friend is there and their phone's there— someone was talking about this earlier— but your friend is there and their phone's there, all of a sudden different stuff can light up, and it's showing you how to use it, and you're learning from them. And you're like— The company SmartThings I was involved with, we never built this feature, but I really wanted to. Because it's very hard to get your home all connected, and lots of cool sensors and everything. But I wanted to, like, if I was over at a friend's house, I could go download the SmartThings app, and he could just, like, boom. Hey, you like my cool connected home? Here, let me just send this over to you. You now can get my home on your phone, and it gives you the shopping list of the devices to get. One click, they'd all show up, and your installer might come out and set it up. Your phone's automatically set up, so it works just like your friend's house, because that's kind of the way that physical things spread in the real world. Now that's how we do it. So I think, if you think about inception, these are just all the different ways in which you actually do it. So now I'll talk a little bit about adoption. This is the next kind of really important piece They're incepted. They know your product solves a problem. They've heard about it from their friend. They're ready to come and try it out. They show up at your front door. Now what? And because I like dog pictures, like, first impressions totally matter. Like, this is like the one moment that you have in order to get somebody using your product. And this is when, like, if you don't catch them now, you're never gonna be able to catch it. You're like, much, much harder to catch them again, because they're there. They're actually ready. Like, I mean, how often do you just download an app when you're like on the go and you're like running between things? You're like, hold on, let me go to the App Store and download this app. No, like by the time you actually are gonna use an app for the first time, use a website, you're actually sitting still. You're kind of ready to try something. You're gonna take a look. Oh yeah, I might have an email confirmation, SMS. You kind of like, you're used to these things. So you're actually ready, prepared with time. Whereas like most of the time we're checking our phones, we're kind of on the go, you're notified. So like this is, this is the moment. And then the other thing to think about is like all of our products, we think, oh my God, they're so easy. It's like really easy to use. Like even— I was going to say something, and I realized Jason had given me all that guidance not to. I was going to say something specific, but I will edit out. But like all of our products, we think they're like, oh yeah, like Twitter, you just sign up, you follow some people, oh my God, it's like great. Like this stuff is like rocket science. Like Othello, it's like a minute to learn, a lifetime to master. And you're like, oh my God, it's just a minute to learn. And then you look at the instructions. Actually, it is a minute to learn. Like, you just put blue tiles and flip them over, black tiles and flip them over to the white tiles. But nothing is easy to learn, and we totally take it for granted. Like, we're building very complex, very powerful software and technology. We have to dumb it down. We have to make it easy to learn. And so I like to call kind of all signup flows should be learn flows. It's not signup. It's not about getting you in the product as fast as possible. It's not getting you to any experience as fast as possible. It's about getting you to— the person's there to take the time and there to learn it. And so, like, you know, when somebody signs up, like, here's the mindset. They're, they're, my friend sent me an invite email, Kim Kardashian tweeted about this cool new app, Jason tweeted about it maybe, I saw a feature in the App Store, everyone else is on it, I should finally try it because I keep hearing about it from all my friends. Hey, it's a free trial, I'll go give it a shot. So, like, you're not really like, oh my god, I'm like so ready to go and and, you know, find out what's happening in the world. You're always in a slightly different mindset, but this is your moment. You have their attention. You have more of it than you ever will again. So I'm gonna come back to my Twitter story. This is Twitter homepage when I joined the company. We had a nice, green sign up now button, and we had a big old search box, which took people into the netherland of, like, death, where they were searching on Twitter when they were brand new. And they would see all these great terms at the bottom, like American Idol and Justin Bieber, and they'd click on those, and they'd be like, Justin Bieber, I love Justin Bieber! Like 100 tweets, and, like, that did absolutely nothing to teach you of, like, the power of Twitter and finding out what's happening in the world. And so, and then the next thing you did was like, hey, your friends are here. And I, you know, I already went into this whole thing of how Twitter's not really about your friends, but that was the first thing we asked. And so you're like, okay, wait, it's about my friends, but I have Facebook for my friends. What's really going on here? And they go, oh, there's more random people. Go follow them. And they were all selected. So you could do finish or skip, which most people did skip. But the number of people who just hit finish and are following exactly 20 people is kind of ludicrous. Jason, were you on the suggested follower list?

I was not.

Oh, this is a great story.

I remember I was the number 2 or 3 user in those early days.

Obama and Scoble.

This guy Obama was like trying to get the nomination. And I offered a MacBook Air that had just come out if I became number 1.

Yeah.

And then additionally, I offered $250,000 to be on it for 2 years as a joke on Twitter when Evan Williams and I were having coffee. Because I was like, this thing is worth— like, I had an idea that every Twitter follower was worth $1.

Yeah.

So I was like, I think I could get 250,000 followers if I was on that list.

Yeah. And the answer is you could have gotten 250,000. But those were not Probably all collectively. But most of it— because they mostly came through here, hit finish, and had no fucking idea what was going on.

Yeah.

And so they'd end up on a screen. If they clicked skip, they'd end up on this beautiful— we call it the white screen of death. And like, this is like the perfect social feed. Or they'd end up on a screen that looks something like this, but you have no idea who any of these people are, because you sure didn't select them at all. And you were just totally confused at what was going on. There's a big box to shout something into the world, but you had nothing to say. So a lot of people would say, trying Twitter, trying my first Twitter.

Yeah, totally.

And so this was kind of the era of like, oh my God, it's amazing how many people got through all of this and still had the kind of nirvana experience that like Twitter is when you have it set up right. When you kind of have Twitter tuned, it's pretty incredible. But we had a big old leaky bucket, and it was terrible. And so as I tell everybody, like, learn Flow, signing up, it's your best chance. Teach your product. Like, you can actually help them understand it at this moment in time. And if you don't If you don't give them the dignity of like, I'm gonna spend some time with you to teach you my product, they don't necessarily need to give you the dignity of like actually using your product or really trying it. And if you really think about it, the people who just want to try it out really, really quick and bail anyway may get bored because of a signup flow. And so you're gonna lose them anyway, because if you don't give them a long signup flow and then they don't figure anything out, they're gone anyway. People who actually care and are interested and want to spend time, they're incepted properly, they're ready, they're in the moment, Like, take it longer, take it slow. You might have a really good chance to actually teach them how to do it, and they might have a great experience over time. So this is how we redid Twitter kind of during my era. It's evolved a little bit since then, but it's actually still pretty consistent. You know, we started over. You know, I already talked about the message. Find out what's happening right now with people and organizations you care about. And then, you know, at our time we said, look, the first step is teach a little bit. Now, I generally hate most Most apps and everything I try have like 4 screens you swipe through, and then you hit OK, and then you're like left in the app. I hate reading, and most people hate reading, and in fact all the tests and user studies and data I've gathered from developers and people I talk to on those is like nobody reads those, and the actual time spent on those is like de minimis. And yet we still do this on the very first screen, because we're basically saying, hold on, we're going to teach you a little bit. Pause, breathe, it's OK. But every step after this has you take actually conscious actions, because I believe like the key to signup flow is making you make conscious decisions. Step by step by step. And as you make those—

Yes.

As you make those conscious decisions, you learn just by kind of activating your brain to think about something. So we just tell you what a tweet is, and we basically kind of get you set up, and most people probably hit OK pretty quickly. And we lie and say, oh, it's just going to take 60 seconds, just to kind of make you feel like it'll go fast. But if it really took you 60 seconds, you probably didn't get the hang of Twitter. So this is the real key, is these conscious decisions. Get people to make decisions. And instead of auto-following a bunch of people if you hit get finished, we just made you go through a list and find people you might randomly know. And we did our best that they were celebrities. And I get a lot of flak that, like, why'd you have celebrities there? And well, it's like, because nobody wants to sign up in Minnesota and follow Jason Calacanis if they aren't already into tech and are already finding him and knowing him. But when you do learn how great Jason or other people are at Twitter, you can actually learn a lot by following them. But you don't start by attaching to that. You attach to people you know. It tends to be celebrities, sports teams, politicians, et cetera.

And so we show you people.

you can follow. You start kind of creating this list of like, oh my God, I didn't know that person was here. We then try to give you some more framing and say, okay, actually, so I've given you some individuals, I've taught you how to follow. But actually, the best thing is you realize all these categories are here. And you might, again, not find the person you want. You can search if you need to. But there's categories of music, news, sports, technology, fashion, design. Whoa, I didn't know all these categories. I can find anybody in any of these things on Twitter. So you really start exposing how big kind of the metaverse can actually be. And then finally, we ask for friends. And we say, okay, friends are important, but they're like a final step for us at Twitter. And that really matters. We actually get more people completing the friends activity when it's later in the flow than earlier. Something I learned way in the early days at LinkedIn is we kind of kept reordering the flow because it's such a big ask. It's like, go and give me all your contacts. Go and give me all your friends. Unless it's a specific app that I know I'm supposed to message my friends, it's a pretty big sort of like commitment to do that. Most people are resistant because they worry everyone's going to get spammed. And so, you know, we put it later in the flow and we're like, now go find some friends too. And they're like, oh my God, wait, I can add Barack Obama and my friends? That's That's pretty cool. And then you end up on a Twitter screen like this where you've actually manually followed each one of these people, so you at least have a clue of who they are, even if you don't know everything. And we're like suggesting you to follow more, and we're kind of keeping you in the flow. So this is kind of where Twitter got to in like 2011 and '12, which was like a massive set of changes. And really, that was part of the big ramp from 10 million or so active users up to what's now 280-something million monthly active users, over 100 million every day, probably much more than that. I don't know their actual numbers now. And it's just incredible that it got there all because we did teaching. And so I've said this a couple times, but the best learning comes from making conscious decisions. And so I'll give you another good example of an app called Lumosity. Who's tried Lumosity here? Okay, it's a bunch of people. So Lumosity will make you smarter. That's all you need to know. I think that's the inception moment. Get this Lumosity app, it'll make you smarter. You're like, okay, I don't really buy that. Sounds like snake oil. Like, like, but okay, I'll try it. Okay, so then you say, go sign up. It says, challenge your brain with scientifically designed training. Okay, that's a pretty good message. I'm kind of getting this idea that, like, okay, it's scientifically designed. It sounds a little bit more active than just makes your brain smarter. That sounds like total snake oil. Let's get started. Okay, cool. So the first question it says is memory. Select all the aspects of memory that you would like to train. Well, I'd like to lose fewer objects. I'd like to remember people's names. Learn new subjects. this quickly, keep track of multiple ideas. At this moment, you're like, oh shit, of course I want to do like all of these things, and like I'm really worried about my memory now. And then it goes on. There's like, look, there's like 5 steps here. Attention. Oh my God, I'd love to avoid distractions, concentrate better, improve productivity, focus longer on important tasks. You're like, oh my God, like this thing's going to help me do all this. Oh my God, I'm like using this like every day, and now I'm scared that like my brain is going to decay if I'm not using Lumosity. And then speed. I can speed up my processes, adapt to changing environments, react quickly, make better decisions under stress. Like, oh my God, like, like, what is this thing? This is like gonna like cure me of everything. Flexibility, problem-solving, calculate figures in your head. I mean, at this point, like, it's just beating you over the head with like, you have to choose which of these things you want to do. Like, I can't imagine anybody who goes through this flow doesn't click every single one, right? But no, I've like consciously told the I want to improve all of these things. So like, okay, create an account. Like, create an account, sign up, name, password. Like, people are not used to this. Like, the conversion rates on these things are pretty consistent across the board of every product depending on how you've got people there and how enthusiastic they are at this point. Okay, and then this asks you for a little bit more like education, how you found Lumosity. But I'm so like afraid that my brain is not smart enough that I'm like actually gonna tell them more information now than I normally would another product. And then like, let's begin, start testing, begin your workout. And like now, like, oh my God, like at this moment, you can see how this onboarding flow got me from like, oh, it'll make you smarter, feels like snake oil, to like, how can I not use this every day? Or like, my brain's gonna decay across like 25 different attributes. And so I just, I think that's like a really important way to think about like how the power of onboarding really does take you from that moment of problem and other things. And look, I'm gonna show Foursquare as well, just to, because I think going through these these things kind of in repetition actually really useful. And Foursquare's new user flow is actually pretty darn good. I think Foursquare now suffers a little bit from like, it's hard to get to this later. It's hard to reinvent things where a bunch of people have a certain attachment to it. And it's still like not 100% like at the point where everybody sort of goes, I need that, I want that, and Foursquare delivers it. But it's actually a pretty darn good product, you know, from where it's come. So, you know, their nice message is Foursquare helps you find places you'll love anywhere. in the world. Like, that sounds pretty compelling. You know, is it travel? It does leave you a little bit like, is it a travel thing? Is it a daily thing? No? Okay. So Foursquare is always on the lookout for the spots you like. Okay, that sounds pretty good. You know, it has this button at the bottom called Neat, which is like, I don't know, to me that kind of bugs me because I'm like, I don't really want to tell you neat. Like, that kind of puts words in my mouth, but that's fine. Foursquare is always on the lookout for the spots you like. I love a good fish taco. Will keep an eye out for places that have awesome ones. This is good, like, it's a very specific example. Fish taco, like, who doesn't like— like, I might not like fish tacos, but like, that's a— oh, I can imagine somebody really wants that, or they really want okonomiyaki, or they really want something like that. And like, the moment that you start getting, oh, it's got that specificity, that's kind of cool. And so it says, sounds good, which is like, again, like, I don't know, I don't love it putting words in my mouth. Okay, let's start making Foursquare yours. Tap a few things you're into, you better add more of these and remove the limit. So like, this is actually pretty good. I think there's a lot— been a lot of telling as opposed to engaging. It starts engaging now, but I wonder if they could have done more of the telling more simply, or even found ways to make that like, which of these 3 things do you like better? Because Foursquare has great examples for each one of them. We'll show you how many are in your city right now. And like fish tacos, you know, mop with, you know, like kung pao chicken, or Or, you know, like a greasy hamburger. Like, which of these things do you like more? And you actually say one, and then you do it, and they're like, oh my God, it could actually help me find that kind of specialty. Okay, so now they say they want to use your current location, and like, that's pretty good. And they're kind of messaging on the bottom to tap okay. Like, if you guys have— everybody on iOS, the tactic is ask for the permission before you pop up the iOS screen to ask for the permission, because otherwise then it's really hard for the person to change it later. Although I just saw for the very first time a few hours ago, you can now deep link directly into the settings. settings page for your app. So if you're like, oh hey, you don't have location on, click here, it can go right there. So Secret, I got the new Secret today, and I actually did the hotlink directly into that to like turn on something that I hadn't, that I'd not turned on before for Secret.

Thank you.

And it was like, oh wait, I can actually just change the setting instead of like trying to find it in the iOS screen that's like, no offense Apple, but terrible. So it's great to see that. So now Foursquare gets more interesting, it gets more engaging. You know, up to people to decide if it's a great UI or a messy UI, but it's actually pretty neat. If you guys have played with it, you just start going through and you type all the things you like. And some of them are generic, like Korean. Some of them are like guacamole. Some of them are like Reuben sandwiches. Some of them are like, you know, Bloody Marys or casinos or comfort food. But like the more you do it, you're like, oh, I like these things. And like, oh, I don't like these things. It's actually kind of a fun, fun game. And so you're kind of playing a little game. But of course, you're getting hungry and thinking about foods. Then it goes back and has you turn on notifications after you finish the game. And the really cool thing is if they could actually show you contextual things based on what you just said you liked. And you're like, hey, you might like these things. If you want to know when you're nearby, they could do even more to kind of motivate it. Now they do sign up, because that's normal. And they do this other neat thing. I actually like this experience. And again, this is part of my, it's OK to take time versus be really fast. They go and say, hold on while we find places you'll nearby. They're actually doing some gathering. And it's actually okay to say, I'm going to do some thinking on your behalf. You know, Google got us really trained in a world where like every millisecond matters for search. And you know what? That completely mattered back in 2002 when we were on dial-up modems and internet was often spotty and slow and everything else, and every millisecond mattered. Now, you know, I'm not actually sure that every millisecond matters in quite the same way. Like, there is a general threshold for speed. Sometimes you'd be like, hold on, let me go do something on your behalf, and then I'll be right back, you may actually create more trust and more connection with the user. So I just want to like, these are things to experiment with and play with. And I actually didn't mind that Foursquare was like, hold on, we're going to go get a bunch of stuff. Because it actually came back with pretty good results for me. Like I could start clicking through, and then the results were beautiful, and they were tailored to me, and they matched the expectations of all the foods I said I liked, and they would always tell me why. And it says, we're always looking for things here, so you can tap here anytime to tell us more favorite things, which helps you get back into that sort game mode. So it's like a really interesting flow that I don't know if they have it perfect, but I think it's really interesting to kind of think about a lot of the decisions they've made, as they just did a really big reinvention. So just to summarize my lessons on learn flows. Look, I think the best learn flows take their time, break the product down into steps. They ask questions instead of just explaining. They start with the easy questions first. Like on Twitter, it's easier to follow things than to go import all your friends. And they put the user in the product only when when they really think the user's ready to go into the product and get the product really ready for them. But you can always skip through. But just know that if you're skipping through, you're not going to really get the hang of it. And if the product lets you skip through, you just stand up and like, hey, you didn't really do much. We'll do our best to show you something useful. And I think that's totally fine. So the last thing I'll talk about a little bit— I'm going to speed through this because this could be a whole talk the same length— is about habits and tracking and retention, how you start to know if this stuff's working at all. And so, look, the real key is how do you actually tell if someone's coming back? Like, the only thing that matters in a product is not how many people you got in the front door, but how many people came back the next day, and the next week, and the next month. How many people did it become a habit for? How many people did it just like decide that this product is now meaningful in their lives, solving a problem? You know, if you're a movies app, or like a weekends app, like, you don't really care if people use it every day. But you sure as heck expect them, you know, once a week or so to like check what's going on on the weekend. You know, like I always say, like Facebook's kind of daily usage, what's up with my friends? Like a movie site should be like every Thursday or Friday, like what movies are gonna be out this weekend, what do I want to go see? Because you're doing that planning on like Thursday or Friday, and you know, look, some people will do it on Tuesdays and Wednesdays, but like you really want to see somebody on kind of a weekly cycle in those cases. You know, some things maybe more of a monthly cycle. Travel is a really hard category to break into. Into, because how often do we fly, you know, for personal travel? That might be more on a quarterly or, you know, bi-yearly cycle, or, you know, twice a year cycle. And so that's why it's really hard to break into a new category like that without something else. And so, you know, this is what we did at Twitter, but I work with so many companies, encourage them to do the same thing, which is why I say, like, what is your core metric that you're measuring, whether it's a login or a post or a share or a read or whatever, and figure out how many people do it month over month. And you can also just do this week over week. But it's how many times did you do an activity or did you visit the product in the first month, and what's the percent likelihood that you visited that many times or more in the second month? And what you can tell is if somebody uses a product N times in one month and they use it N or more times in the second month, they're now habituated to it. You can even tell this in month 2. You can actually tell this in week 2. And so it's really important, because most people sign up and they're lookie-loos and they go right away. And that's fine. We really want to find those people who are not, who kind of keep using it on a consistent basis. And so we did this exercise, and we ended up with a line that looked like this for Twitter, which is based on how many times you did it in the first month, what's the likelihood you did it at least that many times in the second month. And in those early days of Twitter when I joined, you know, fall of 2009, the number looked to be about 7. If you came to Twitter like about twice a week in a given month, you were pretty likely to use it about twice a week in the next month. And that just meant that, like, you'd found some habit in your life to do it. Now, there were people who did it, like, almost every day every single month, was actually, like, this neat sort of mode at the end of the curve. But even once I found you'd gotten that habit, you had done that. And then most people just didn't get that habit at all. And, like, most people completely didn't. And so I called those users core, by the way. Just to explain this diagram. Those were the core users who really got the habit of Twitter. And I said, okay, how many people who sign up become core in their first month? I call this the kite diagram, which like we just mapped out what states people went into. We said like 20% of new people become core, and then if they become core, 90% of them stick around month over month. 40% of them go to cold, which I actually called comatose, but I thought cold sounds better now. Which is they just don't come back. Twitter at all, and then occasionally get resurrected. And how many of them become in this bucket we call casual, which means they're coming 1 to 6 times in a month. They don't really have the habit of Twitter in their lives. And so a lot of them just fade out into that colder comatose the subsequent month. And so by doing this, we understood that to change Twitter, we don't need more people to come and use Twitter a little bit more. We need to figure out just how to get this single bucket up. How many people who sign up get into that core bucket quickly through the first month? that, we're going to increase our overall usage in a really meaningful way. And so we then did another exercise to try to figure out why. And we went through kind of a bunch of data and actually a bunch of surveys or like interviews of people. And like, don't underestimate the interviews, especially in the early days. Like, every one of you should be calling your users like pretty often and just like get on the phone with like some of them. Like, why'd you try it? How'd you hear about it? What do you think of it? What could I do better? And like, you're gonna learn so much, because they're gonna like give you those words that then matter in inception, and in your press, and in everything else. We did this great exercise where we called people who had signed up, gone away for 30 days or more, and then come back and were active again. And we said, why'd you sign up in the first place? What didn't we do the first time? What got you to come back? And when you came back, like, what got you to stick? And the story was actually pretty consistent for Twitter back in those sort of late 2009, early 2010 days. They had heard from somebody that Twitter was this awesome place to like shout into the world, and they were using it for their business, they were using it because they loved this activity, and they were just tweeting all the time. And they'd be like, but it sounds like I need to be on Twitter because everybody's on Twitter because the press is talking about it all the time. So they'd come sign up, they'd follow random stuff, and they'd be like, this is ridiculous and I have nothing to say. So they'd leave. And then at some point they would hear from one of their favorite— like their pastor would say, hey, I'm tweeting a bunch of biblical stuff every week and news about the church, you should come to Twitter and follow me. Or they'd hear from a celebrity, like one of their favorite celebrities is like, yeah man, I'm shouting out on Twitter all the time. And like Shaquille O'Neal did this thing where he's like, I'm in the airport. Anyone who can come and touch me will get a picture with me, or like whatever it was. But like, that was like this great story. And like people were like, oh, I can go follow them on Twitter and hear from them and interact with them in like a real way. That's pretty cool. So they'd hear from somebody else and they'd be like, oh, I can go back to Twitter and listen. Oh, and I can listen to stuff I care about. And then they would come back to Twitter, find those things eventually, and really get hooked on Twitter. And so it was like, it just taught us like how powerful the listening message was. And then we also found it in When we got someone to follow more than 30 accounts, when we got them to have a Twitter graph where about 2/3 of who they followed did not follow them back, and about 1/3 who they followed did follow them back, then people were really sticky on Twitter over time. And that was, that was like unique to Twitter. Like, this data doesn't, like, doesn't apply to any other product. It really depends on exactly your product. It might be they added their phone number and they put it on their desktop and their phone, which is like a Dropbox stat. Or, you know, it might be like they added 7 friends, or, you know, what, you know, they used it 3 times, or they put it on their home screen. I mean, there's a lot of things that can be the key catalyst, but you want to figure out what those are and really understand them. You know, and then the last thing I say to everybody for everything you're doing in growth is be creative, because the first and most important thing with everything in product and growth is we're all creating the future. There's no like, this person did it this way, so I should do it this way. It's always unique to your product, and it's only your creativity that will actually come up with the right way to talk to another human being that will, in a meaningful way, that will matter to them. But use your data, because the best part of being creative is you can have 10 ideas and then you can learn softly which ones work or not. Build real features. Like, growth hacking is not slamming a share button on, you know, every, you know, 17 places in your product. Growth hacking is not, please invite more friends, you know, in random places. I'll just put a bigger invite button and that will cause more growth to happen. Let's worry about building these meaningful features that I've talked about that really get stuff going. And then ship, test, and learn. Because if you're not doing that, nothing's going to get better. So anyway, that's just the long thoughts on growth.

Thank you.