Twitter (with Dick Costolo)

Yeah, I'd prefer the one single take. So, you know, but we'll see what happens. It doesn't matter. Welcome to Season 7, Episode 5 of Acquired, the podcast about great technology companies and the stories and playbooks behind them. I'm Ben Gilbert, and I'm the co-founder of Pioneer Square Labs, a startup studio and venture capital firm in Seattle. And I'm David Rosenthal, and I am an independent advisor to startups and angel investor based in San Francisco. And we are your hosts. Today, we tell the story of a company that has changed every single one of our lives, Twitter. Whether you use the product all day, every day, like I know many of you out there do, ostensibly as part of your jobs, or you have just seen the occasional

discomforting tweet from a world leader, there is no doubt that Twitter impacts us all, especially heading into this historic U.S. presidential election. But how did these 140-character messages that grew out of the SMS protocol eventually play such an outsized role in our society? And conversely, why is it that despite being the heartbeat of the world, Twitter still pales in comparison, both in user numbers and in revenue, to its juggernaut social media cousin, Facebook? So a lot of ink has been spilled on Twitter's founding and its early days of musical chairs between various founders and board members. And most of our listeners know all about that.

So as Ben and I were reflecting on what the acquired way to do a Twitter episode would be, we realized that there actually was this pretty significant fork in history. You might say it turned on a knife point where starting in about 2010, it could have been Twitter that ended up owning Instagram, not Facebook. And the conversation we'd be having today about mobile, consumer, and social could all look very different. Indeed. Well, we have the most authoritative source possible to help us tell that story, especially this period of Twitter. The former CEO who generated the first dollar of revenue, scaled it to a real business, and took it public, Dick Costolo. So welcome to Acquired, Dick.

Thanks for having me. Great to be here. Pleasure to have you. Well, listeners, as always, if you love acquired and want to hone your own craft of company building, you should join the community of Acquired Limited Partners. You'll get access to the LP show, where we dive deeper into the fundamentals of company building and investing, in addition to our monthly LP calls, where we talk with all of you directly, and of course, our book club and Zoom calls with the authors.

Recently, we had Emily Chang on to discuss her book, Brotopia, that is now available in the LP feed if you missed that, and just awesome to have Emily on. So fun to see so many of you on there with us. Looking forward to the next one. If you aren't already a limited partner, you can click the link in the show notes or go to acquired.fm slash LP, and all new listeners get a seven-day free trial. All right, listeners. Now is a great time to talk about a new partner of ours here on Acquired, Lagora, the agentic operating system that is redefining how the world's best legal teams work.

Yep. It's sort of obvious that AI is going to completely change the legal industry. I bet most of you listening have dropped a contract into some sort of AI chatbot out there. Lagora took that insight and asked the question, what if you really built something with that power from the ground up for the legal industry? So the founders did exactly what great founders do, operate with obsessive customer focus. They embedded inside a massive law firm for months.

They sat with the lawyers just watching how the work really gets done. And that's how you get features that customers love, like tabular review, where you drop in a folder of hundreds of contracts and it pulls every key term into a grid a lawyer can actually work with. Lagora's bet here is interesting. Since it lets each lawyer handle more complexity, any given person can increase the quality of their work and do higher value work. And this means that the pie can grow even as each individual task takes less time.

And they recently launched Lagora Agent, offering greater intelligence and performance. The agent lets lawyers set an objective. Then it can handle the planning and the execution and delivery of the final product. Legal teams get to maintain full control and transparency since they're still involved where judgment is required. And Lagora works where you already work. You can use it within Microsoft Word while redlining or drafting. The early Lagora numbers essentially speak for themselves. When they have a head-to-head pilot with their top competitor, they win 70% of the time. Lagora now has over 100,000 lawyers on the platform from 1,200 legal teams in 50 countries. And crazily, they went from 1 million to 100 million in ARR in about 18 months.

Truly insane numbers. And that is the real test. Plenty of things demo well, but the question is whether a busy associate actually reaches for it during crunch time, or whether a partner trusts it before going into a conversation with a major client. If your legal team wants to check it out, whether you're a law firm or you're in-house at a company, you can learn more at lagora.com slash acquired and just tell them that Ben and David sent you. All right, David, take us into Twitter.

Well, before we go into Twitter, we're actually going to spend a little bit of time in another era, a simpler time, if you will, of Web 2.0, of which Dick and another key Twitter person, another former Twitter CEO, Ev Williams, were critical parts. Dick, lots of people know, I think probably most listeners on our show know your background. You studied computer science at the University of Michigan, which my dad, if he's listening, will be very upset because he went to state, but that's okay. We'll make an exception.

And I'm very upset, Dick. I'm so sorry about the last, what, 15 years of football as a Buckeye myself. We don't talk about that anymore. What rivalry? And then after Michigan, you made the very natural career choice to throw away your computer science degree and go into stand-up comedy at Second City in Chicago. You have such amazing stories from that with Steve Carell and Tina Fey and I get so many other great places. We can do talk about that a lot in the Recode episodes with Kara. But after that, when you came back to tech, you first, this would have been what, like the early 90s, right? You went to Anderson Consulting?

Yep. Early 90s. That's correct. Way back, dating myself significantly. But yes. Right before it was about to all take off. Then after a few years at Anderson, you left, what was it that you set up? You started several companies first. FeedBurner actually came a little later, right? Yeah, that's right. We started a company, just basically a web consultancy. When Netscape first took off, well, I should say Netscape was my first exposure to it. The Mozilla browser and then the Netscape browser. When that really took off, this is before the Netscape IPO, but when that stuff started to happen, it became obvious to me that the web was this really extensible thing and place. And that that was where most of the innovation was going to occur in the immediate

future, which was true, but not universally understood at the time. Started a little web consultancy, a little web design and technology consultancy called Burning Door Network to Media. And that was the beginning of it. FeedBurner was actually the third or fourth company we created. Was it the first one that was spy on it, right? So that was the second one. The idea behind spy on it was just, look, there are all these things happening over the web and you kind of got to go check them out to see when they change or see something happen. We just called them spies. You should be able to set little spies on things that come and alert you and find you when something on the internet changes in a way that's interesting to

you. Super cool little early alerts idea that worked really, really well and sold that to a company called 724 Solutions, which is a public company in Canada. The reason I wanted to bring up spy on it was I'm wondering if it was sort of like a predecessor to RSS. I mean, I remember going and checking the Engadget blogs every day, like, oh, has anything changed? Like, have they posted new stories? People were just starting to think about like, hey, you can push updates to people on the internet. Is that kind of how you were thinking about it?

Yeah. Both spy on it and FeedBurner were really the brainchild of one of my co-founders, Eric Lunt. He had come up with the ideas for both of those companies. And that's correct. The idea behind FeedBurner was just, look, there's going to be this new world of syndication that's important as people stop going to all these different websites to see if something's happened and start aggregating it or pushing it to individual places. And the idea behind FeedBurner was it'll be important to sit between publication and subscription because there will be all sorts of interesting things that happen that intersection of publication and subscription and we should do that. So that was the idea behind FeedBurner, basically managing syndication feeds for publishers and doing things to them and

manipulating them to make the subscription experience better and then pushing them out. So long-winded way of saying, yes, those two companies were very much related in the way we thought about the future of how content was going to be consumed. It's just ironic for... We have a lot of younger listeners who probably don't really know what RSS is. And it's best explained, I think, as a podcast player for text. It's an app that you open with all these different feeds in it.

That is exactly right. When you said and did all sorts of interesting things to it, when you manipulated these feeds, was the primary sort of monetary value driver the fact that you could insert ads into feeds? Well, that was the primary monetization driver. You know, publishers obviously make money. Most of them did before there was paywalls on anything by people consuming ads on their sites. And if the content starts to be syndicated and there aren't any ads in the feeds, then publishers aren't going to make any money and that's a problem. So that was the monetization idea. But there was just a bunch of other stuff we did to normalize the way they could show up in these things that were called readers,

RSS readers, that were basically where all the different content got aggregated. Just stuff to make publishers' lives easier, be able to understand where their subscribers were, how many of them there were, how many of them had actually read articles, all that stuff. And of course, you say reader, again, for our maybe approaching middle-aged listeners like Ben and me, they'll think of Google Reader. You, FeedBerder, of course, ends up at Google, as does Ev's company Blogger. I mean, what was that like?

So I met Ev in Web 1.0. I'll make this super short story. When this company, Miller Freeman, was a conference company and they had this web design summit conference in the late 90s, Ev and I actually met through that. I think he was on their advisory board or something. And I taught a couple sessions at the first one or two of those. And we didn't keep in touch much. When Google started looking at FeedBerner in, gosh, I guess it was 2006, it was over a year before they actually acquired it. Ev and Biz and those guys were at Google having acquired Blogger a couple of years earlier, or a year or so earlier. And that's how I sort of got reacquainted with Ev. So actually, by the time

we got to Google FeedBerner, that is, in 2007, and I was working for Susan Newadjiki, who was running ads at the time and is now the CEO of YouTube, most of the content advertising and AdSense strategy had been baked. And FeedBerner was, look, if content syndication becomes important, we're going to need to be a player there on the ads front. So that was the logic behind the acquisition. But by the time we actually got there, Ev was in mid-2007, Ev was gone and doing, obviously, ODO slash turned into Twitter. That was 2006 when they started pivoted from ODO to Twitter. And then 2007 when it started, the like sort of Ashton Kutcher, CNN stuff started happening.

That raised a million followers. Dick, are you keeping in touch with them at this point? Or are you pretty like heads down on what you're building at Google? We're keeping in touch with them at this point. I ended up putting some money into the, I think it was the seed round in Twitter that Union Square Ventures, Fred Wilson, led. I put like a personal check into the seed round in Twitter. I think that's going to work out for me.

So that meant you didn't need an equity grant when you joined as COO, right? Because you were already... Yeah. I really screwed that up. We'll come back to all this in a minute, but you end up doing Twitter's first actually revenue deal is a data deal with Google on this front. Yeah, that's right. Google and Microsoft. I think we closed them both on the same day or launched them on the same day or something too. But yes, that's correct. The first revenue deal at Twitter was a data deal with Google.

So in 2009, you finally make the jump from angel investor to executive. You join Twitter as COO. And then in 2010, you became the CEO. We were just talking about acquisitions of FeedBurner and Blogger back at Google. This has been chronicled quite a bit, but there was a lot of acquisition discussion going on at Twitter at the time, right? The first, I think, was before you joined when Facebook tried to buy Twitter for 500 million. This was years before any revenue.

And then there was a second round right after you joined. What was that discussion like? They're always the same. Acquisition discussions, it's funny. It's like 16-year-olds dating. And what I mean by that is, hey, we would be interested in a strategic discussion. Would you be interested in a strategic discussion? But otherwise, we're not interested in a strategic discussion. It's like that. I'm not really exaggerating. It really is like 16-year-olds trying to see if they should go out on a date or not and then not want to be embarrassed if the other person says no. And then, of course, everyone says, we're not for sale. And you may be surprised and maybe it wouldn't be. In most cases, these things are like they are in movies where there's some sophisticated plan of attack. And,

you know, it's just sort of more informal conversations. We've been doing this, working on this data deal. The stuff you're doing is obviously super important. We can see how it fits into our company in all sorts of interesting ways. Let's at least have the conversation. You have the conversation. And obviously, it all comes down to, as it always does, look, do we think the probabilistic outcome of us doing this on our own is better than calling it a day here?

And significantly so. So that'll be worth the work we're going to do over the next few years. And at every step, we thought it was. You know, at least when I was with the company, when we were having those discussions, at every time there was an acquisition offer, we felt like the probabilistic outcome of us being able to do something a lot better than that. And with more significant returns over time, we should do it on our own.

I don't know if they still do it, but Biz, obviously Biz Stone and Peter Fenton used to come guest teach a class at Stanford GSB in Peter Wendell's venture capital class. And they'd tell this story of that day. I told my wife afterwards and she was like, this is the epitome of tech. The way Biz told it was he and Ev drove down and met with Zuck at the, I think they were in Menlo Park by that point in time. I think they'd left Palo Alto campus. And the discussion was for between 500 million and a billion-ish dollars to sell the company, which by the way, we ran the math on that. And if it were 500 million in Facebook stock back then in 2008, Facebook was valued at 15 billion. That would be $27 billion

today in Facebook stock. Twitter would still be doing, made the right decision by not taking that offer. He joked that on the way back, they were discussing and Ev said something like, well, if we're worth 1 billion, we're worth 10 billion. And it's so true. That's a very Biz Stone thing to say. He's got this sort of very Zen way about him. That doesn't surprise me. The more interesting way to look at this was Zuck was then and is still incredibly smart as in many things and as an acquirer. I mean, at this point in time, Twitter was growing faster than Facebook.

It would have been, this was a little later that Twitter hit 50 million MAUs, but the growth was accelerating significantly during these years when you showed up a little bit after these discussions. So there was still no revenue at the company. I'm curious to hear your perspective on what it was like building it, both in terms of how you first did those distribution deals with Google and Microsoft to monetize, but then also building out the ad platform. We've talked about a bunch on this show, whether it's Google or Facebook, there aren't that many people that have built out ad platforms at scale, like you kind of set out to do when you arrived. Like how did you think about even

starting to do that? I'll separate the two things. The two data deals with Google and Microsoft, I got in there, I think September, maybe August 29th, September 1st was my first day. And we had those two data deals done at the very beginning of October. And we're just the first priority, like let's get these things done. There've been discussions going on back and forth for a number of months now. They've kind of been dragging on. I'll go get those wrapped up. So we went and did those two data deals.

But at the same time, those were sort of a fixed amount for a fixed period of time. And who knows whether we're going to have the leverage to do those same deals again, two years down the road. So we knew we needed a more scalable strategy. It was pretty quickly decided to pursue an advertising strategy. So the big challenge with Twitter and building an ad network around Twitter was unlike Facebook, for example, the Twitter content went all over the place. It went to twitter.com. It went to third party clients. It went to SMS alerts. It went to ESPN.com and the New York Times and on and on and on. So we realized right away, all right, we need an advertising format

that goes everywhere the tweets go. From there, you quickly realized, well, the one thing that can go anywhere a tweet can go is another tweet. So we'll make the ads tweets and we'll just inject them into the feed, whether it's a syndication feed or a third party feed or our own site. And that was really the birth of the native, probably much to this dismay of a bunch of your listeners, it was the birth of the feed based ad that ended up being what's used in all these products today, Facebook and Pinterest.

Oh, I didn't realize that Twitter pioneered that before Facebook. Yeah, Facebook's ads were in the like AdSense ads, they were off to the side on Facebook.com. They weren't in the feed, Instagram, etc. Necessity is the mother invention. The reason we invented that is because we had to. There was no demilitarized zone of tweets, the sidebar, where to put stuff. There was just the feed. The beauty of that innovation was, well, gosh, we'll be able to measure engagement with the ad in those same ways you measure engagement with a tweet. And so we had this visiting professor at Stanford, Ashish Goel. And Ashish really came up with the idea of using the tweet ad engagement as that'll be the way we charge. Instead of cost per

impression or cost per click, we'll charge based on cost per engagement, they click and open the tweet, or they like it, or they retweet it, or they reply to it. Those are great indicators of engagement. And he was the sort of the father of thinking about the way we built the ad model and the monetization engine behind it and charged for it. That was all him. He was the thinker on that. And how do you make that management decision that we're pioneering something new here, we have this visiting professor who's a genius, we are going to sort of...

Wait, how come he's the genius? I just described to you 90 things that we came up with. Now Ashish is the genius. No, Ashish is... Oh, I thought you said that he was the father of the... Ashish is a genius. Ashish has got the awesome characteristic of being a very soft talker. You're always kind of feeling like, I think I just missed something really important there that he said, but I already asked him to repeat himself. And now I don't want to look like an idiot.

Anyway, I'm joking. He really was sort of the founding father of the way we thought about the ad model inside the company and the way we charged for it. So interesting to think back to this point. I already sort of exclaimed earlier, I was surprised that this predated Facebook's now multi, multi hundred billion dollar cumulative revenue over time ad format. Twitter really pioneered this. It's important to remember Facebook was just garbage on mobile at this time. It was well before the Instagram acquisition. Everyone was basically worried that Facebook's moat wouldn't translate to mobile.

Yeah. Well, they were wrong. And if you would, there was a lot of money to be made buying Facebook at 19 when it crashed after the IPO because they thought they wouldn't figure out mobile. Betting against Mark has always proven to be not a very good strategy. There was obviously the mobile product issues of Facebook. I want to get into in a minute, all the companies you were buying at Twitter and bringing the Twitter native app itself was a acquisition, right?

Yep. Eight bits. Yep. Lauren Brichter's app. That's correct. So obviously there was the product issues, but also I think you raise an interesting point I hadn't thought about. It's also just the nature of the ad unit itself. Facebook had done that big Microsoft deal. It was essentially still just display advertising on the site at this point, right? Yes. I think they probably charged based on clicks. It was AdSense style was the ads were over in the sidebar. You know, the news feed was in the middle and then on the right hand side of facebook.com you'd have those bar of ads. Yep.

Wow. You tell that to the kids today and they don't believe you. When you made the decision to pursue Ashish's vision of this ad model. I think it was Dick's vision from what I heard. Yeah. Based on engagement. There's like a ton of pieces you got to go build out. That's like a heavy engineering lift. There's obviously the product aspect of that and then the data feed aspect of that with all the third party applications. And then there's also the, now you got to go spin up teams to convince advertisers to do business a different way, right? What was your like Gantt chart of planning all this?

I can't remember that far back. I knew that I needed someone on the go to market side, meaning the sales side to really build that whole organization. I didn't know anything about building a sales organization. I didn't know anything about building what sales ops was, what sales fine. I didn't know anything about that stuff. So before Google had purchased a feed burner, we had been in discussions with another company and that company was Fox Interactive in Los Angeles. Adam Bain was running Fox Interactive and had built their whole ad stack and ad platform at Fox Interactive. Adam is president down there. He both understands the technology and probably one of the best business development networker relationship salespeople I've ever met in my life.

So I'm going to go try to convince him to come join us. So I spent about four or five months trying to traveling back and forth to LA, trying to convince Adam to come up and be president of global revenue at Twitter. And he would say, let me think about it. He was super hard to convince, but I finally convinced him to do it. I remember, and of course, now we're partners on the investing side. I remember telling the board, I got the perfect person. I keep in mind, it's very Silicon Valley board.

Not all the investors on it were Silicon Valley. There was Fred Wilson from Union Square Ventures in New York, but a sort of traditional tech heavy mindset of these are the kinds of people who do these jobs. And these are the kinds of people who are good at those jobs. Obviously, we should go to Google and get one of their people like Facebook did. And I was like, no, no, no, I got the guy. He runs Fox Interactive. And Evan and everyone else were like, what? We're going to hire someone from Fox to do this. But we did. And that was a great decision. And Adam built one of the great go-to-market organizations in, I would say, in the last 20 years. I mean, that team was...

So many of the leaders on that team have gone on to run other organizations and has developed tons and tons of great leaders. And that team was amazing. So that was on the go-to-market side. Once he got in, I kind of like, you take the ball and run with it. I don't know how to do this and you do. And just to underscore that, I mean, people know about Adam, and obviously he's your partner, but he did an absolutely incredible job. I mean, that first year of revenue in 2010 was mostly, I think, the Google and Microsoft deals. But then the first year of ad revenue in 2011, Twitter did over $100 million in revenue and then 3x that the next year. I mean,

it's interesting, right? The Twitter revenue story is this incredible story. Yeah. Most people don't realize we went from literally $0 in revenue before I got there, not close to $0, but $0 to a run rate of over $2 billion a year when I left in July 2015. So five, six years later, $0 to $2 billion. I think that's probably only, I don't know, Amazon, Facebook, Google, maybe Spotify. You'd have to go check. But there aren't many that went from $0 to $2 faster than that. The point being right up there and with a new ad format, and really from a challenger position, all those other companies were kind of like, you got to buy from me because I'm number one and I'm the biggest. And we did it from the

challenger position. Facebook always being the 900-pound gorilla and us trying to fight for survival. So it was super impressive what he was able to do. The guy's tireless. I remember one time, and I'll talk about the product and engineering side in a minute, but I remember one time in probably 2012, we were at a sales conference in Chicago. We had a sales conference with a bunch of marketers the next morning, on a Wednesday morning or something. And Adam and I are in the hotel lobby at midnight going through the deck that we're going to use the next day. Go up and go to bed, see you at seven in the morning. Come down at seven in the morning, and he's still sitting in the same chair

wearing what he was wearing last night, working on the deck. And I was like, dude, you got to speak for an hour. How are you going to do that? And he just powered through it. It was amazing. So he's just one of those people that can go and go and go. All right, listeners. Now is a great time to tell you about a longtime friend of the show, Vanta. AI has scrambled the whole security picture. It used to be that you proved that you were secure once a year, an audit or a static PDF, then everyone would nod and you're done.

But in an AI-first world, that doesn't hold up anymore. Yep. Your risk surface changes every week now. A vendor turns on an AI feature or someone writes in a new model without telling IT, and your posture is different than it was last week, let alone at your last audit. Vanta's own research found that around 70% of companies have this quote unquote shadow AI running with no security review at all. Right. And that's where Vanta comes in. They're the leading agentic trust platform, meaning they've built the thing that closes the gap. And the way that they close that gap is Vanta agent. Think of it as a GRC engineer. That's governance, risk and compliance, except that it's software and it doesn't sleep. It finds the issues, drafts the fixes,

and cuts the time that you'd spend on vendor assessments in half. In half. Which is exactly why more than 16,000 companies today run on Vanta. Companies like Ramp, Cursor, and Snowflake all stay audit ready and catch the risks that crop up between audits across every vendor, every AI tool, the whole environment. And that's the real value. Trust has to be continuous now, which is why Vanta automates your security, your compliance, and the work to earn and prove trust. We're huge fans of Vanta over here. And literally hundreds of acquired listeners have become Vanta customers at their companies over the years. So you can get $1,000 off Vanta at vanta.com slash acquired. That's V-A-N-T-A dot com slash acquired for $1,000 off. And just tell them that Ben and David sent you.

On product and engineering, like today we take this thing kind of for granted. Anyone who's bought ads on the internet, there's a self-serve portal that you can go to and get incredibly granular on who you want to target. And then it all just kind of works. And it gets displayed to them in the way that you expect. And there's probably only like 10 if you include the Reddits and the Snapchats. And obviously everybody has different levels of granularity and different formats of what the ads can look like.

Twitter was one of the first three. How do you do that when it's a technology that's only been built before by like Google and other people who have copied Google and hired Google's team away? The same way. That was one where we definitely knew we had to go get people from Google involved. And we did. The person who ended up running ads engineering, Alex Redder, and built that, we recruited out of Google. And then went and found other people who had done things like yield management and all of that. All of the crazy intricacies of auction-based advertising pricing and had done that at Google. We went and got some of those people. It's just hard. And the learning curve on it is tough. And especially on some of the yield management stuff. So on product and

engineering, we went and got engineers who had done some of this stuff before at Google. And on the product side, there were a handful of product managers. Anamitra was the first one who's now an investor, early stage VC, was the PM on the very first ads product that promoted tweets inside Twitter search. And then Kevin Wheel, who had been on the team. Actually, Kevin Wheel was at the company before I got there and had been on the data science team. Kevin Wheel then came over to the product side. And obviously, later on, years later, the vice president of product at Twitter. And then after I left Twitter, he went over to Instagram and was VP product there.

I'm curious if you're willing to talk about it. What was it like recruiting all those folks? How did those conversations go? I mean, I assume they were, Google was paying them extremely well. It was hard because Twitter engineering, despite having some really extraordinary people on the team, Kaylee Torgeson running the data science team was extraordinary when I got there. And there were a bunch of people. I mean, I'll go through all the specific names. There were just a bunch of extraordinary engineers on the team, some of whom are Evan Weaver, who's a CEO of FaunaDB now. There's a bunch of great killer engineers on the team when I got there. But despite that, Twitter engineering had a horrible reputation. Why? Because of the fail whale. The site couldn't keep the site up. Go to try to

recruit engineers to come work at Twitter. And they're like, what's the matter with you boneheads? You can't even keep your website up and running. I'm going to come like the most famous HTTP 500 return code in the history of, you know. Then you're like, it's all built on Ruby on Rails. So it's just hard. But we did it. Kudos to the people who took the risk and came in and worked on it because it was a slog. Migrating off of Ruby took forever. And there were corridors where we couldn't do any new product work. We just had to refactor the back end because it was such a train wreck. Because it was pieced together. It was cobbled together in the early days just to

keep it up and running and keep up with demand. So that's just the way these things go. You've told this story many times before, but it's too good not to share. Can you talk about when the Russian president stopped by the office? Yeah, Medvedev came by the office. He's going to write this historic tweet to Obama. Surprise, surprise, the site crashed. So we're like, oh man, one of the engineers fake wrote a back end for the tweet that Medvedev sent from the iPad on stage at the Twitter headquarters and rewrote it and inserted it into the database. And, you know, look, it worked. I mean, it was madness. The next morning in the New York Times, there's this big headline, President Medvedev, historic communication with

Barack Obama on Twitter. And meanwhile, at the time, you're like, oh my God, you've got to be kidding me. We literally hardcoded that. Literally, literally hardcoding the tweet that he wrote. Stupid stuff like that. Anyway, yeah, that kind of stuff happened all the time. And there was truckloads of heroic work by engineers who would have much rather been doing other things, refactoring back ends and doing all the work to get us off the monorail, we used to call it, because it was a monolithic back end into a more robust service-based architecture while obviously still scaling the whole company and users. So that was real work. And it was hard to convince people to come do it, but we did.

To my mind, at least, there's one other piece, maybe two other pieces, but one other piece in particular I want to cover to all this as you're building this engine. And that's the third-party piece of this. If the advertising model for Twitter is tweets, people consume Twitter all over the place. It was obvious to me when I was CEO, for a couple reasons, and I'll go into them, and they're not well understood, but it doesn't matter, that we had to own the user experience.

You can't, despite what anyone else wants to say, can't build an advertising business in which a bunch of other people own the user experience, because it doesn't make any sense. You need to be a non-profit at that point and just decide, we're not going to be, I'm being serious, I'm not being facetious. Totally, totally. I mean, it makes a ton of sense. Yeah. You can't do all the work to build the service, run the service, de-spam the feed, do all the user customer service work with the hundreds of people that are responding to user requests, then send out all that content to a third party that takes the content, de-spammed, the clean, de-spammed real-time feed, and puts their own monetization around it. It's stupid.

You wouldn't take the New York Times and hand it to, oh, if you're in Chicago, then you don't want to have the New York Times. You can just take it and sell all the ads yourself and keep all the money. Well, the company's out of business like a day later. So it was obvious to me we had the own user experience. And at the same time, I think it was right before I became CEO, there was a company that started realizing, look, if Twitter is not going to go on their own network, we should.

TweetDeck, right? This is the Idea Lab, guys. Yeah, Bill Gross. Bill realized smartly, look, if you guys aren't going to own your own network, I'm going to go aggregate it myself. And so he started going out and buying a bunch of the third party clients, maybe bought two or three. I don't remember how many of them he bought. And I started to realize like, oh, Bill's going to go build the network and then go, great. I now own all the users and I'm going to monetize them however I want and hold us hostage.

So I was like, okay, we can't let these third parties do this anymore. And that was the beginning of the thinking of third parties differently. I wanted to have a platform in which third parties built into Twitter instead of taking our stuff and building away from Twitter. And so we'd had prototypes for a number of kinds of platforms. In the first couple of years, I was CEO around Twitter cards that never really materialized or came to fruition the way I'd wanted them to. Thesis was always, let's have a platform on which people build into what we're doing, not take our stuff and build off of what we're doing. That's a real platform. And that's where you get network effects.

Is the counterfactual to that? Well, this is why RSS didn't succeed because all those publishers own the content that got plumbed into someone else's experience and they couldn't, in a first party way, control the ads and the engagement on the ads and the data coming back from the ads. No, I think the reason RSS didn't succeed is just because Facebook and Twitter displaced them. Just became easier for everyone in the world to use these things instead of RSS readers.

Makes sense. I remember so viscerally when, for me personally, that light bulb went off and I was like, oh, I had been a religious Google reader, RSS user before then. I was like, wait a minute, I don't actually need that anymore. I can just be on Twitter. I'm curious, maybe reflecting back on it today, there's an economic opportunity that comes from being a platform that developers can build experiences on. And you can think of that as maybe like the Tencent type opportunity. And there's an opportunity that comes from being the greatest place to buy ads in the world, which Facebook and Google basically have a duopoly on.

Twitter also makes $3.5 billion in revenue on, but not $70 billion or not like $100 billion. How do you think about, was that the right call to build an advertising business versus being a platform company? Yes, because the road to being a platform company for Twitter at the time, based on the way the API had been leveraged to date, 95% of which was to build people going and building their own clients and taking the stuff out of Twitter and moving it away from Twitter and doing whatever they wanted with it. The road from that to building a platform where developers could build into Twitter was far.

Having gotten the ad network and ads business up and running in the short time we did, and having struggled with building out the platform of Twitter cards the way we had theorized was possible and could be awesome, I would say that in hindsight that proved to be the right decision. Yeah, it's hard to imagine going from that to a WeChat-like experience. It was just too long a road. Yeah, it's also the most common application you could build upon Twitter's platform is a Twitter client.

The vast majority of sort of apps that people could conceptualize were that. It wasn't like it was Windows, it was Twitter. Yeah, yeah, yeah, correct. Twitter cards, this is the perfect kind of segue to talk about the other thing that was going on at this time. It's so fun talking about all this because so many hugely important both decisions and events happen for the internet and what shapes everything today right around this time. So Instagram basically launched on Twitter, right?

And I believe was using Twitter cards. Is that right? Yes, and they used our API. They used the API to build the network, the graph. Yeah. Yeah, we were close to Kevin and probably is annoyed by this. I remember everyone calling him Mikey. I'm sure he's like now like, hey, I stopped going by Mikey 10 years ago. Jerk. When Kevin had been at Google with you. Yeah, yeah, yeah. Right. Were you there all at the same time?

Yeah, I think we may have. He was in corp dev, if I remember correctly. And we may have crossed paths once or twice, but not tons. But when they started, Jess Varelli, who is in corp dev at Twitter and was there before I was there, and Kevin were close. So we were close to them from the get-go and liked what they were doing and knew it was unique and different. And there were a bunch of things that there was one camp that was like, there were already filter companies.

Hipstamatic was doing it. There were a couple other ones. But it was pretty clear that these guys had, it wasn't just the filter. It was the, you don't have to think about the aspect ratio or the this or the that. It's just a square. It's always going to be a square. And we're going to make it look great. Just point and click. And we're going to make the photo look awesome. They're all going to be a square.

It's all going to show up the same size in the feed. And that was a genius innovation. And it was just the two of them for a while. I mean, I remember even when we tried to buy the company, probably getting ahead of where you want to be. But March 2012, when I tried to buy Instagram, it was before they had raised the $500 million valuation round from, I think, Greylock. Just before that. It was March 2012, meeting with Kevin and Mike and just me and me and a couple members of my board and me and Jack and Peter Fenton and stuff like that.

Just couldn't convince them or get it over the finish line. A month later, I was in Tokyo, April. They went and raised the money instead. They're like, look, we're going to go it alone. We're going to raise this round from Greylock. And what was the dollar amount you were talking about with them? A lot. It was over 10% of the company. It was well over 10% of the company. Whatever the valuation was at that time of Twitter, it was like a lot.

Maybe like $600 million, something like that. But I think if I remember correctly, we made it so like it's $600 million to the cap table. But Kevin and Mike get a lot more because we're going to throw in a bunch of stuff just for them. Well, the crazy thing is that actually would have been a much better deal than what they did. Because when you went public the next year, 10% of Twitter would have been probably more than where Facebook stock was trading.

I think those guys would tell you things worked out just fine for them. I mean, yeah. I don't know. Anyway, I remember just trying and trying. And they came back and said, we're going to go do this raise instead. I was like, good luck. You're going to crush it. And then I was in Tokyo in April. And I woke up and had a text message like, hey, Instagram just sold to Facebook for a billion dollars. And I remember thinking like, oh, man, that's rough for us.

Because the beauty of Instagram had been, well, we had text, but Facebook always had photos. And the beauty of Facebook was everybody pours all of their photos into here. I just remember thinking in February and March, and I've been working with Ali Rogani, my CFO on the acquisition and the numbers and the valuation of the deal and how we're going to structure it and everything. And Ali was also really pushing the board that we got to do this, which was great.

And we got approval from the board to do it. And we just felt like, look, we've got a lock on text and they have a lock on photos. And if we can get Instagram, it's going to be a fair playing field. Yes, they've got this just extraordinary number of users and can crush us in so many ways. But this will really give us a beachhead. Keep in mind at the time, I might be wrong about the numbers.

At the time that we're talking about, Instagram is like 30 million MAU. It's not 500. It's not 200. It's like 30 to 35. And Twitter is like 140 million MAU at this point. These are big numbers. No one would say this now. But if you go back and read the articles at the time of the Facebook acquisition for a billion dollars, people are like, is Facebook insane? Are they crazy giving them a billion dollars? They've got no revenue and only 30 million users.

But we knew we're really in a pinch now. These guys are going to be able to really going to be able to outflank us and bring real leverage to bear against us. And they did. Wasn't too long after that, that they stopped allowing Instagram photos to appear inside the feed and Twitter. And we were just the day of the acquisition. We knew like, oh, man, that's really going to be a problem for us. Two things happen.

I don't remember the order which they happen. The first you mentioned where, and I think to this day still, when you post an Instagram link in Twitter, it frustratingly is just a URL. And you're like, come on, guys, can't you get along? But the other thing is that the social graph stopped working. Instagram no longer could use Twitter's API to help you find your friends to follow. How did both of those go down? Well, I got a call one day, I think it was from Mark.

And Mark said, hey, just want to give you the heads up. The Instagram team has decided that Instagram photos need to be viewed on Instagram. So the insert the Instagram photo into the Twitter feed is no longer going to work. And it's just going to be a link. I think you said something like they're going to shut that off tomorrow. And I was like, that's a bummer. That's really, really bad. It's bad for the... I love the phrasing too.

Like, oh, the team has decided. It's a bad user experience. And I don't think you should do it. But more urgently, if you turn it off like four hours from now or five hours from now or whatever the number was, it's just going to break Twitter. So we need like 36 hours to go make it so that it just doesn't break when you post an Instagram link and at least show the link. We hustled and did that work.

Engineers stayed late that night and we did a bunch of work. And people are constantly calling me and emailing me that week like, why did you do this? I didn't do it. Everyone knows that it's a better experience when the photo that you wanted to post actually shows up. So nothing I can do about it. I can't go over there and show them all the mail and say, look at all the people you've made sad. I think if I remember correctly, keep in mind, we this whole time don't ever have access to the Facebook API that we were blocked from using it.

You can't go use the social graph like fat chance. So at this point, we're like, well, why are we going to be the schmucks that keep letting everybody use our stuff when we're blocked from using all theirs? So I shut that off. But then it didn't matter because they had the Facebook social graph. It is so interesting reflecting back on it because Instagram really has a perfect fit with Twitter. The asymmetric follower model instead of the friend model.

The world would have been very different. It would have been fun to see how it sorted out because it would have been a, I don't know if the word fair fight is the right way to put it. But we felt like we had a beachhead and an angle to compete if we had that text over here and photos over there with the same follower model and the same way of thinking about the asymmetry. And, you know, you could have imagined, you know, I don't know if Kevin's probably like, want to stab me the next time he sees me.

If I say this, Instagram still doesn't really have rebroadcast, right? Or retweet or regram or whatever you want to call it. So Jack and I had been talking about, it'd be great to talk to Kevin about, you start to think about interjecting things like that. And you really have these services that mirror each other. That'd be really, really cool. And there are all these just ways we would have been able to be a lot more competitive.

And that would have been fun. But say lovey. Well, that's one of the reasons we really wanted to do this episode and focus on this time period in particular is there is this alternate universe. A couple of people wake up on a different side of the bed one morning or, you know, a butterfly flaps its wings in Australia somewhere. And Twitter owns not only Instagram, but also effectively TikTok as well, because it was shortly after this that Vine happened, right?

Yeah, that's right. So Jack had met Dom Hoffman, one of the three founders of Vine, and invested in it. This is pre-launch. Keep in mind, this company hasn't launched yet. And Jack shows it to me. And Dom comes into the office with the co-founders and immediately was like, I get it. This is genius. The innovations that they made that, again, are in so many things now today and are in everything like the Instagram video and everything else of the video is just going to keep playing.

It doesn't stop. And then you have to hit play again if you want to see it. It just sub seven seconds. I remember Dom's other like sort of genius was it's not exactly this many seconds, even though we say it's this many seconds. It's just short of that because that's the right length. And I was like, OK, well, you're crazy, but I love it. And he was like, and it just replays automatically. Those guys had come up with a bunch of cool, innovative ideas, like you say, precursor to TikTok.

And we bought it. It launched. It was number one in the App Store on iOS for like eight weeks in a row. Which makes you guys the best venture capitalists of all time. You spotted a product before it launched and before it had traction. Spotted a product before it launched, bought the company, launched it, goes to number one in the App Store. We didn't get the Android client out quickly enough. And I think as a result, one of the things that happened was that when Instagram launched short video, seemingly in reaction to Vine, I didn't think it was going to hurt us as badly as it did.

And one of the reasons I think it hurt us as badly as it did is because, keep in mind at the time, if you had Vine and you had an iPhone, the cool kids had Vine. And the people that had Android phones were like, this sucks because I got an Android phone. And then Instagram comes out with short form videos on both iOS and Android. And you all of a sudden have all the Android Instagram users going like, Vine sucks.

Instagram is way better. That was certainly, certainly helpful to them and hurt us more than we thought it would. And then just the power of the ability to bring numbers to bear across that network on multiple products was just as bad as we thought it would be when they first bought Instagram. We knew it would be bad. And then it was as bad as or worse than we thought it would be. And they were quick. You can go look at the historical charts.

Usage just started dropping off and faster than I would ever would have suspected it. Man, that is the best encapsulation of basically the story of consumer technology over the last decade on the Internet. You really can't outrun someone once they have a lead and they have either network effects going among a consumer user base or they have look at Netflix, you know, enough users that are paying that they can amortize the costs of paying for content.

It all keeps coming back to the same thing, which is once one of these services hits escape velocity and becomes dominant, it's so hard to ever compete with them on anything close to the same playing field. You just said the key point, though. The one place where I always felt, and this led into the kind of stuff that we did in 2013 and 2014, the one place I always felt that we had an advantage that we didn't leverage significantly enough was syndication.

You can't view Facebook content anywhere but Facebook. And you can view Twitter content, billions of tweets per day, all over the web on ESPN.com and NewYorkTimes.com and in blogs and in blah, blah, blah, blah, blah. And it's everywhere. You don't have to go to Twitter to see it. It's literally everywhere. So I always felt, and this was the idea behind Twitter moments was, look, if we can package up things that happen on Twitter into these kind of bundles as they happen, whether it's around the sort of predictable things like the Oscars or the Super Bowl or the blah, blah, 200 of these things a year, whether it's around those or it's around this crazy thing just happened.

And here are the eight people who are there and their perspectives and all these crazy other tweets about it. And you can make bundles of those things that then get syndicated like all the other tweets get syndicated. And you could start to think about long-term injecting ads into those bundles. That was a place where I felt Facebook can't compete there because they won't let their content be viewed off Facebook. The comment you just made, Ben, about an even playing field was I always felt like Twitter's advantage or opportunity for a beachhead in competition against Facebook was in syndication because Facebook wouldn't go there.

So we would have free reign to do that. After I left, that wasn't pursued as aggressively for whatever reason. I know Jack has said, you know, look, the company was doing too many things and had to pare back what they were doing. And that's worked out for them, obviously. But that was my thinking behind that. Was Periscope part of the thinking around that syndication idea? We were constantly looking for beachheads. What are they not doing that we could go do that will be at least hard for them to do?

And Periscope was another company, again, that Kayvon Bakepore had founded. Once again, we bought that before it launched. And Kayvon was working on launching it at Twitter. Meerkat launches. And like, oh boy, Kayvon was a week or two away from launching Periscope. When Meerkat launched, it didn't freak me out. It was just like, ah, that's a bummer. My thought was, oh, it's a bummer. Here's something Facebook can go by if they see Periscope and want to compete against it is what was going through my mind.

Again, you just constantly have the 900-pound gorilla on your brain. And you're trying to think of ways to outflank them, outrun them, or both. And then Periscope, the idea behind Periscope was, well, here's a way we can try to get a little bit of a flank against them around live content. And Twitter is all about live content. And those two things together should be awesome. That was the thesis there. And again, great acquisition. It worked and took off.

And then, of course, as you know, Facebook launched Facebook Live, some kind of interface where you can just tap on it and get the likes and see them bubble up. They really ratcheted up the volume of live Facebook Lives. If you went live on Facebook, the volume of people who would see that was enormous. When you've got that network, you can just beat the crap out of the new kid on the block. So it happened for the second time there with Periscope.

It's crazy. It's so crazy. I mean, there's a food chain here, right? And when you look at Twitter launching Periscope after Meerkat, I mean, it was obvious two weeks after Meerkat launched. It was the Silicon Valley darling for two weeks. It was all anyone could write about. It was the next Twitter. And then suddenly Periscope comes out and you're just like, oh, I feel so sorry for those guys. They're going to get crushed. And they did.

Yeah. And then Facebook Live comes out and you're like, now I feel sorry for these guys. All right, listeners. Now is a great time to thank our longtime friend of the show, ServiceNow. If you are running a large enterprise, AI agents are likely spread across every team and deploying them is no longer the hard part. Yeah. The hard part is knowing what permissions they have, what employees are using them for, or what decisions AI is making.

AI security for an enterprise at scale is not a small concern. Like the risks are real. And the challenge with AI is governing it, securing it, measuring it, and making sure that it actually delivers value. That is why ServiceNow built the AI control tower. Yep. AI control tower gives enterprises a single place to see, manage, govern, and optimize AI across the entire business. And it works with any AI, not just theirs. Every device on your network, every permission across every system, every AI agent visible and secure in one place.

And ServiceNow can do this because they've spent more than 20 years building the operational backbone of the enterprise. The workflows, governance, approvals, security controls, and institutional knowledge that power how work actually gets done across IT, HR, customer service, finance, and security. ServiceNow already runs more than 100 billion workflows annually and trillions of transactions for more than 85% of the Fortune 500. So when companies need a place to govern AI at enterprise scale, they're building on a platform at the center of how their business already operates.

And in a future that isn't going to be one AI, it's going to be thousands of AI agents working across every function of the company. But the question is, who's managing them all? So if you're trying to turn AI ambition into real business outcomes and make it work safely, securely, at scale, go check out ServiceNow.com slash acquired and tell them that Ben and David sent you. All these dynamics are true. The Musical.ly story is interesting because they actually gutted it out.

They were growing and things were going well, but it was a tough slog until the ByteDance acquisition. And then when ByteDance poured so much advertising dollars into the rebranded TikTok, that's finally when it kind of hit escape velocity, right? 100%. Yes. Would it have been possible for Twitter to do something like that with one of these products to just say like, we're going to go raise boatload of capital? Would have been a very high stakes move.

But did that ever cross your guys' minds? No. The things that crossed our minds and that we worked on were big acquisition ideas. To make a long story short, again, this is the kind of thing that became obvious to LinkedIn. They're like, okay, we're subscale. We're just going to be here at subscale unless we do something huge. And then they went and did the Microsoft deal. So we were constantly trying to find that similarly. Like what's the thing to go do that changes the equation?

To make a long story short, and there were lots of reasons for all the different ones that we pursued didn't happen or that we thought of didn't happen. For lots of different reasons, we just weren't able to find that right fit there. But we were constantly thinking about in addition to the how do you outflank these guys or go somewhere they can't go, not sell the company, but what do we go buy or merge with or go borrow a bunch of money from the bankers and buy this thing?

Kind of along the lines you just mentioned that makes this a different playing field. And we just were never able to come to the right answer there. Did you guys ever think about podcasts? There was Twitter music, but any other audio? That's sort of the obvious. It's not video. It's not text. There were some people inside the company that, of course, would build little things during hack weeks and stuff around that. But I would say no, generally speaking, no.

I mean, obviously, the ultimate irony since the company was spun out of a podcasting company, ODO. So that would have been amazing. That would have been amazing and probably the way the movie should have ended. Yeah. Well, there was a, I got to imagine, pretty bright spot in all this in the IPO process. Yeah, that's a crazy process. Yeah, the IPO is nuts. We put the price of like, I'm going to get the specifics here wrong.

But it was $26 a share that you priced at on. Yeah, I remember that. The S1 had $13 to $16 a share on it. We printed the S1 and before the roadshow, it was $13 to $16 a share. Keep in mind, the company is 7 years old at this point. It started in 2006. It's now 2013. The company is 7 years old. It's been around for 7 years. We print $13 to $16 a share on the S1.

Go out on the roadshow. Crazy demand. 70 times oversubscribed or something like that. So the night before, we're like, okay, we're going to price this at $26. Wow. All right. Well, so the price was about $13 a week. 7 years to get to $13. 2 weeks to get to $26. So I'm on the morning of the IPO. You're on the floor of the New York Stock Exchange. I'm actually on CNBC. I'm live on CNBC. And I'm being asked some question by Jim Cramer or Carl Quintanilla.

And I hear over my shoulder, I hear one of the market maker guys go, we got $13 million at $47, which is like, you know, meaning he's got like demand for he's got demand for $13 to $47 or whatever the number was. And Goldman Sachs was the lead left banker. And so Goldman guys were going to open the stock. And they weren't going to open it until there was sort of a balance of a certain number of shares on both sides of the equation, on both sides of the trade.

Took a long time. It was almost 11 o'clock, I think, before it opened. Maybe I'm 1130. But anyway, I hear in my ear, and I remember thinking like, oh, that's funny. I didn't know there were two companies that could go public on the stock exchange the same day. And one of them's, I mean, obviously, that wasn't the exact thought I had in my brain. But I remember just thinking like, what could that possibly mean? We're not going to like open it.

And we were. And then we did. So I remember that day, immediately after it opened, we got on the subway to go up to Midtown, the Twitter New York office in Midtown to talk to the employees there. So I'm on the subway with CFO, Mike Gupta, and our banker, two or three other people on my staff, some of my team on my team were on the subway. This guy looks at me, he's like, aren't you the Twitter CEO?

I was like, yeah. He's like, what are you doing on the subway? New York office, this will be a lot faster. He's like, huh. I get off, I go into the New York office, and I talked to the team there. And Biz and Evan and Jack and I go and talk to the team there. And then we were like, we got to go. We got to go back to San Francisco. We're not going to like do the traditional stay in New York and party.

We got to go back and talk to the team. I got on stage. I got back to Twitter headquarters. Everyone's freaking out, of course. And I got on stage and was just, I said something along the lines of, it took us seven years, seven years. And you know how hard it was to build a company that was worth $13 a share. While I was gone the last two weeks, good job, whatever you all did, because you doubled the value of the company while I was gone.

So I don't know, whatever you were doing, like that was amazing. But even more amazing was that while I was gone this morning on the New York Stock Exchange, you almost again doubled the value of the company just sitting here drinking champagne all day. So good job. Everyone's laughing. And I said, listen, the reason I say that is for the first time, we now have a stock that won't just go up. Once a year, it'll go up when there's a financing.

It's going to go up and down every day. And you all know sitting here that you didn't do anything today to increase the value, intrinsic value of the company, the worth of the company by almost 100%. You know you didn't. So just remember when the stock goes up for whatever reason or the stock goes down for whatever reason, that's now disassociated from what happened inside the Twitter offices on these days. Keep your head down because that's going to happen.

There are going to be days when the stock gets hammered and people are going to look around and go, what happened? And it's not going to be what happened. It's just disassociated from the focus and goals we have to have inside the company on a day-to-day basis. And you're just going to have to build resilience to that and be mentally tough about it. I got off stage and I thought, man, I really like steeled the company for the day that stock goes down a lot for the first time.

And of course, as you might imagine, when the stock went down a lot for the first time, people were like, what's going on? You know, it's like I just had this conversation. But of course, the people who are starting to multiply their options by the stock price and assuming it's never going to be below that and I'm going to be able to buy this kind of house. So when something happens, they get upset and they worry about it.

But anyway, I thought I was doing this really genius ninja move and it didn't really work. Well, we should say it would have been hard for anyone with what the stock was doing to instill any sort of measuredness. It went up to 75 before the end of the year, which was, you know, That's so crazy. Multiple on 75 to what we were doing in revenue and even just EBITDA at the time was bananas. You talked about that $40 whisper in your ear while you're on CNBC and then it almost doubles again in the next six weeks.

It's just wild. And then, of course, it has a six month crash back down to basically the IPO price or a little bit below from there because there's just so much hype around the company at that point. I think one of the interesting things that happened was we're pretty confident about our financials. We were just like, look, the work we got to do is driving MAU growth. It's work. And I sat on CNBC that morning and they were like, oh, congratulations, a really well run IPO.

It's like, man, we got a lot of work to do. And I said it during the roadshow over and over again. I think I said, man, we got a lot of work to do like four times that morning on TV. And then we did. And I thought we were going to get more credit for our outsized financial results. People would go like, OK, they didn't grow users as much as we wanted to this quarter, but boy, they really crushed it on the financial results.

And so we're going to give them some leeway here. It became clear after the first quarterly call that we were going to be judged almost exclusively on user numbers quarter to quarter on the big MAU number that was in the S1. And we talked about all the time and compared to Facebook's MAUs and all that stuff. Yeah. I mean, I can imagine the narrative of people that were buying the stock at those prices was, hey, this is the next Facebook story.

This is going to get to a billion, two billion users, be it that same. And that's the only way you would justify that, right? Yeah. I mean, we always had a goal. I mean, I'm sure the company still does value inside the company to reach every person on the planet. Twitter will be at its best when everyone is on it. Certainly something we wanted to have happen. It was just harder. Again, people even at all hands meetings at Twitter would, people in charge of like growth on the product team at Twitter would get up at all hands meetings and type into Google.

I don't get how to or I don't get. And the first suggestion would be how to use Twitter. It's just hard for people to get until they get it. It's one of those things that once you get it, you get it. And until you get it, you're like, why does anyone use this thing? So it's not intuitive for most people. So it's always been more of a slog. It's one of the reasons I was again, I was like, well, the beauty of syndication is you don't have to get Twitter or log in and create a follower graph to leverage it and use it and get use out of it and get value out of it.

So let's go build into that. Let's go build moments and distribute those things and syndicate them far and wide and make a lot of money that way off the billions of people who see tweets and syndication. It was another sort of rationale behind the desire for syndication to be a more powerful force and monetized force inside the company. Yeah, Dick, if we just speak really abstractly for a moment, and I just... We have been speaking abstractly.

Well, I could ask about the financials if we want to go the other direction. I'm kidding. I'm kidding. What other reasons are there other than the products for whatever reason is hard to use and always has been hard to use? Twitter, when you left, was about 300 million users. It was more than that, but yes, give or take. Okay. And now, at least last reported MAU numbers a year ago, a year and change ago, it was about 330 million.

So it's a product that in large part, in terms of MAU numbers, appears to have basically topped out. Now, the daily monetizable or MDAU is the monetizable daily active users continues to climb at a great clip and the team's doing a great job executing on that. Why is it that Twitter doesn't have Facebook's DAU or MAU opportunity? The beauty of the Facebook network is that it's your friends and contacts. Well, when they first built Facebook and there's nobody's graph to go use, what did they use?

They used Gmail. And if you'll remember, when you were not at a university and you started to use Facebook, they said, you want to import your contacts from Gmail? Yeah. Oh, great. Type in your Gmail password here. And then they went and scraped your Gmail contacts and built your graph. So the Facebook graph is your friends and your family and the people you talk to every day. The Twitter graph is the economist and that one dude who's super hilarious at ESPN that writes about the Milwaukee Bucks.

He's not in your phone. He's not someone you email on Gmail. So there is no like starter kit for your graph for the interest graph on Twitter like there was for Facebook. The beauty of that is every time someone says they built a Twitter killer or Twitter competitor, it was easy for us to shrug it off because the network is super resilient because the nodes in the network, the connections in the network were like really unique and hard to build and took a long, long time.

So to think you're going to go rehydrate that network in a week, you're not going to, it's not going to happen. On the other hand, if you've got a network like WhatsApp, you remember shortly after Facebook acquired WhatsApp crash for like eight hours, Telegram added 13 million users in that like six hours. Why? Because it's just your contacts. You import them from your phone and you've rehydrated the exact same network. In a matter of minutes. So those networks are less resilient over time and can be hit in the side of the face in a moment because they're so easy to rehydrate.

The beauty of a LinkedIn or a Twitter is there are difficult networks to build. There aren't other network node types like them. They're much more resilient to competition. Yeah. Twitter's the only social platform I think that exposes bidirectional. You can go to any Twitter account and see what accounts that Twitter account follows. You certainly can't do that on Facebook. You can't do that on LinkedIn. To your point. Okay, great. You extract that data like it's not going to get you anywhere.

Yeah. Dick, I have a follow up question. So it's sort of a similar question, but on a different axis. It's this sort of comparing Twitter to Facebook. And we should first acknowledge, oh my God, Twitter's a company that has 300 something million users. That's amazing. And they do three and a half billion dollars a year in revenue right now. And I can't fathom how difficult it is to build an enterprise that does that. And at the same time, when you compare it to Facebook, the amount that Twitter is able to generate in revenue on any given user.

And you can look at this number globally. You can look at this number scope specifically to the US is lower. Is there sort of an intrinsic reason about why that is or will always or is more naturally the case? It's a good question. I'm not sure what the answer to that is, to be honest. I don't know what the upper bound on either one of those is. Obviously, it's the case that on Facebook, you can micro target, you know, most everything about those users and advertiser, you know, where they live, you know, where they went to school, you know, what high school they went to, you know.

Relationship statuses and on and on and on and on and sex and blah, blah, blah, blah, blah. And on Twitter, you don't. You don't know that you have an interest graph, but you don't have a detailed demographic understanding your users. The ability to micro target by demographics gives them extraordinary pricing power. But who knows? It's hard to know where either of those tops out, I guess is the short answer. And the only reason for the differences that I can think of are the interest targeting versus the demographic targeting.

But it's possible that you could come up with a world in which nothing really changes about the way you think about either user base, but the average revenue per user or the value of a user flip flops because of some other innovation. I just don't know what the answer to that is. Yeah, it's interesting because it's funny. Like, we don't know any demographic information about our listeners, but we do know what they like. And from that, we can extrapolate that they're an incredibly valuable audience.

I don't need to know demographic information to know that everybody who listens to us skews toward founders, skews towards executives, skews towards investors. Knowing that means that they're a much more valuable audience than if I knew the demographic stuff. I've heard that argument before that Facebook has more information so you can target more granularly so you can charge higher amounts because you can guarantee that someone's getting exactly the customer they think they're getting. But I always come back to this thing where, gosh, if you can know the things people care about, there's something that for certain use cases is much more valuable about that.

I'll put you in touch with Matt Torello, the head of global revenue at Twitter, and he would love to talk to you. I'm sure he has thoughts. Ben could be a tear quote on the Twitter sponsorship kit. It's so great to dive into all these business aspects of the history of Twitter. I think we'd be remiss, though. We're recording this a week before the 2020 U.S. presidential election. There's kind of also this parallel story at Twitter, mostly while you were there, Dick, of politics.

And I guess particularly, maybe we can scope it to U.S. politics on Twitter. It must have just been wild, right? Like, you know, during this time, shortly before you joined, Barack Obama gets elected in 2008 as president. Large part thanks to Twitter. There's the Twitter town hall. Donald Trump accuses Barack Obama of not being a U.S. citizen, not having a valid birth certificate on Twitter. Like, this is all happening on the platform. Like, what was that like for you running the company?

I have to tell you that it's like the Medvedev thing. People are like, wow, you were there when Medvedev, like, tweeted at Barack Obama. You're like, yeah, that's not what we were thinking about that day. We were thinking about the fact that the site was down. It's kind of always like that. You read these company histories or you read people's biographies. There's always this mythology of we knew when we knew then that this was exactly where we were headed.

My experience is that that's all nonsense or largely nonsense. It just is. We were always surprised. I'll give you an example. I'm in my Monday morning staff meeting shortly after becoming CEO. It's like November 2010. My assistant comes in the office. Leadership team meeting. You got to like meet for 90 minutes and go through all this sort of more strategic topics. And it's an expensive meeting because you got the 10 people with the most reports in the company.

My assistant comes in the meeting and goes, hey, this phone call for you. All by way of saying like, okay, why are you even telling me that? Take a message. I'm not going to interrupt this meeting for like a phone call. And she goes, it's the White House. My first thought was like, I didn't do anything. What do they call me? What's the White House calling me for? I was like, first of all, it was like Zuckerberg.

No, I was like, first of all, like, how do you know it's the White House? Did they call and say it's the White House? Like, how do you really know it's the White House? She's like, no, it's Valerie Jarrett. She wants to talk to you. And then my second thought was, I didn't do anything. What does the White House want to talk to me for? It's my Midwestern guilty conscience. They wanted us to come to the White House and with some other tech leaders and talk about these kinds of things you guys were just mentioning.

And I remember thinking like, man, I got a computer science degree. What do you want to talk to me about this stuff? After it happens, you realize it and you start to think about it. You're like, oh, I got to start actually being more thoughtful about these things and understanding how we think about intellectual property policy and how we were blocked in China. And who should we talk to in the US government about that? But as it's happening, you just don't kind of realize whether or not it's a momentous thing or just, well, that was weird.

You don't sort of realize it until later and you look over your shoulder and the impact you had or that the service had is sort of the wake that's behind the boat. You don't recognize it while you're in it. Yeah, that's fascinating. I can only imagine that every day you just wake up and your primary job is, I have all these stakeholders to manage. We have all these goals and how do I hit them and avoid all the company killing things that are constantly coming my way, including my own decisions.

I might decide something wrong or I might not do something or somebody on my team might not do something well. And that could be catastrophic for us. And all I'm trying to do is do what we said we were going to do. There's own goals. And by own goal, I mean, you know, it's a soccer metaphor where you're like, what just happened? The old, the vice president of HR comes by and is like, you're not going to believe what just happened.

And you're like, you've got to be kidding me. There's just always... You're like Snoop Dogg in our office? Yeah. For example, I was never, fortunately, never actually there for most of those things, fortunately or unfortunately. But all that stuff, trying to balance and manage. And I was just always trying to tell myself, and it's hard because you want to like, wait a minute, what just happened? And I got to go focus on that. But you're just constantly trying to remind yourself, like, what's the highest leverage thing I could be working on right now?

Not dive bombing people and getting into the weeds. But it's hard because, you know, you feel like this crazy thing that's super messed up and shouldn't be. And I can go show them how to fix it. But you're right. And on top of all that stuff, there are these self-inflicted wounds. There are at every company. All right, listeners. Now is a great time to talk about one of our favorite companies, Statsig. Yes, longtime acquired partner. There is a reason why the best product teams at companies like OpenAI and Notion, Atlassian, Figma, Rippling, Brex, and more rely on Statsig, whether they are iterating on their core product features or shipping AI-powered experiences at scale.

Yep. In the crazy speed of today's AI world, shipping fast is just table stakes now. It's basically trivial to build and deploy your app constantly. The real advantage is how quickly you learn what changes actually created value for customers and how fast you can use that signal to guide what you ship next. Whether it's a feature tweak, a pricing change, a performance improvement, or an AI update like a model change or prompt adjustment, they're not relying on instinct.

They're measuring what actually moved engagement, retention, and ultimately revenue. And as more teams build with AI, that learning loop becomes even more important. Building with LLMs introduces non-determinism into your product experience. The same input doesn't always produce the same output. And behavior can shift in subtle ways in real-world use. So doing offline evals will give you part of the picture, but you can really only understand the impact once your product is live with real users, and then you can measure how their behavior actually changes.

It's very different than the way that you would ship features in a pre-AI world where you knew exactly what the software was going to do in production. Yeah, exactly. So this is where Statsig comes in. It brings experimentation, feature flags, and product analytics into one unified system so teams can ship safely, test rigorously, and directly link what they changed to how users actually behaved. The result is a tighter feedback loop and learning that compounds over time so you don't just ship more, you ship better.

So if you want to make learning your competitive advantage, whether you're building new AI experiences or just evolving your existing core product, go to statsig.com slash acquired to get started. What's an example? I don't want to directly ask you what are some self-inflicted wounds, but what are some things that you remember deciding or you remember the company deciding to do something? And in retrospect, do you wish you had done it differently? Oh, just the way we communicated with third-party developers.

You know, we tried to like tiptoe around it or like, hey, you can't do this anymore. But you can still do this. And communicating by trying to not upset people is stupid. And it should have just been like, look, we got to own the user experience or we're going to get crushed as a business. So get over it. People would have been mad and hated me, but so what? And it would have been clear and simple and direct.

And it's what ended up happening anyway. And it needed to happen and was the right decision. So we should have just said that instead of, but you can still do this and this and this. And then everyone's like, this is just nonsense. So stuff like that. Some of the self-inflicted wounds are like HR stuff. I'm not shrugging it off. It's important. That's what I'm saying. You're like working on this strategic thing or thinking about this thing.

And then someone comes in with some HR issue and you're like, oh, wow, that's super important. I can't blow that off. I got to go deal with that now. But wow, it would have been great if I could have kept working on this thing I was thinking about and was making progress on. You're constantly trying to balance all that stuff and not be the person that says, hey, I don't care about that HR thing. I'm working on this really important thing.

Who cares about that? That's just also the wrong way to think about things. You have to balance the forest fires with the forestry management stuff. Right. Well, it's interesting. Reflecting on one piece of forestry management, we talked about the growth of the importance of Twitter in the political sphere. Simultaneously, there's also more and more people suffering with personal attacks on Twitter and for the lightest way to put it, probably trolling or a heavier way to put it, probably hate speech.

This is an extremely difficult problem to solve. Is there anything you can point to along the way where you're like, I wish we had done some policy differently? Yeah. The thing that people bring up is, oh, you wrote this note in whatever it was, 2014 to the team. That gets sort of overplayed. As early as 2010, I was like, this is nonsense. Why are we allowing this stuff on the platform? And the defense to, there was a combination of both a naivete around, we're the free speech wing and the free speech party.

But there was also this sort of like, well, because it's a slippery slope. And if you suspend Bob, then what are you going to do the next time Janet says something like that? Or if you're going to suspend Bob for X, what are you going to do about when Janet says X prime? And I always just kind of said like, well, who cares? We'll cross that bridge when we come to it. This guy's clearly like threatening violence against this person, but in a way that doesn't specifically violate our terms of service, which is stupid because our terms of service say, in addition to all the specific reasons, you can be kicked off.

The terms of service say, we can remove your account for any reason or no reason. So just tell them we removed you for no reason. Yeah, you're an at will user. So I just always had wished we were like more aggressive on that front while understanding the other side of the coin, which was and the pushback I got, the heavy pushback I got, which was, well, well, well, if we do X, then what about X prime?

I just think that all of the platforms have waited too long to say, well, actually, we do need to go do X and we'll figure out X prime when it happens. And is there anything I think pundits talk about this a lot where they're like, well, in order to maintain their status as agnostic platform, that's not taking sides and actually being a media company. If they arbit what is said on their platform too much, then they'll be viewed as an arbiter instead of an independent. And I think this wades into Section 230 territory.

But fair enough. However, the problem is the companies already arbit what you can do on the platform. If you post a video of beating your dog, guess what? It gets taken down. Why? Because the company decides that people need to see that garbage. So you're already editing and moderating what happens on the platform to say like, oh, no, we're the phone company. No, you're not. We take stuff down every day. I just have always found that to be a sort of a fourth down time to punt, not the right way to think about it.

So Dick, speaking of content and moderation on Twitter and taking down tweets, you recently took one down of your own volition. Tell us about that. And what were you thinking? Boy, did that go sideways. Holy smokes. First of all, I should know better, you'd think, than to use sarcasm on the platform. But I was making a sarcastic response to Parker and Jason Kalkanis in this whole back and forth about the Brian Armstrong's post about separating the mission of the company from social activism and social causes. I think Parker made some comment about, I'm not going to remember his exact response to Jason, but it was something like, look, if you want to really want to go do that stuff, you go to a nonprofit.

And so I just fired off this sarcastic response. I was also like doing four other things that day and getting ready to entertain a bunch of friends the following day. So I fired off this sarcastic response. I was like, I'm just going to cause an argument. I don't need to deal with this right now. I'm just going to not pay attention to Twitter for a little while. Boy, was that a bad decision. I should have used the pitchfork analogy. And then the headlines would have been, Costolo incites farmers to revolt.

What I was trying to do that went horribly wrong was just make the observation that, look, if you think you can separate the social contract from the economic contract in society, then don't be surprised when the pitchforks come out. That went wrong quickly. A bunch of people were like, you should explain it. You get back on there and explain it. I'm like, how do you explain sarcasm? That seems like a bad idea. So I just kind of left it up there and like, ah, people will settle down. And then I'm laughing because I started getting these violent threats on my Instagram.

pictures of tomatoes in my house in Napa Valley. And people were like, I hope these tomatoes burn in the fire. And how about if I line you up in front of those apples and shoot? I was like, wow, wow, this is really bad. Finally, the next day was like, all right, I just take the whole thing down. But man, what a mistake that was. Whoops. I'll bet. Do you feel that there's some sort of reckoning coming in some sense of too much polarization?

The thing that it obviously reinforced for me is, man, people are on edge right now. So I get it. Like people are just on edge. I think it's probably going to get a lot worse before it gets better, which is unfortunate. And hopefully it does get better after some distance from the election. Yeah. Well, knock on all sorts of wood there. All right, David, now I'm going to take us into grading here. And rather than doing our normal grading thing of either looking at sort of the big company acquiring small company, which we sort of did already, thinking about what would have happened if they had bought Instagram or perhaps painting what is an A plus CF looks like in the

future. I think it's actually more interesting on this episode to evaluate how Twitter has managed their impact of the platform on society. For a lot of you out there, this is going to be a squishier way of grading. Dick, we wanted to ask you first, how would you sort of rate Twitter's ability to manage the impact of the company on the world? While I was CEO, I wish we had done a lot more and been a lot more aggressive.

We did some stuff. We removed this alt-right, I guess you would call him troll, even though he didn't specifically violate. He violated the sentiment of a number of our policies than any one specific policy. And so we removed his account for those reasons. I can't remember his last name, Chuck something, but wish we had been a lot more aggressive. So I'd give myself a C on that. I was the CEO of the company and despite pushback, could have said, I hear everybody's opinions, but we're going to do it anyway. So I give myself a C. Since then, I think the company is really leading the charge in a number of ways and on a number of fronts around political ads. They were

one of the first to say they weren't going to take certain political ads, even when everyone else stood firm. And then on moderating nonsense content or fake news or whatever you want to call it, misinformation. I think they've done an increasingly great job in the last little while. Well, companies always going to come under fire for whether they do too much or do too little, but I think they've done an increasingly great job on that front. So I give them an A over the course of the last year. It's interesting, right? I mean, obviously there's still much work to be done, but over the last year, two years, it does feel like Twitter is the only one that is even doing some of this work. Facebook, not so much.

Or at least pushing the envelope. I think Facebook ends up doing things that they need to do for PR reasons. To our earlier conversation, I mean, maybe this does open up a seam again to have Twitter be something that Facebook's not going to do. Facebook's just not going to do that because of how their business is run. Yeah. Well, there's certainly a whole can of worms there. David, I pretty much agree with you, but I only think there's been even a B effort even recently. The reason I'm not higher is that there's just so much hate speech that comes from what seems to be pretty simple heuristics to catch and manage. And I know that's always easier said than done, but it just feels like Twitter has

kind of shied away from that fight. They seem more emboldened now, but still definitely afraid of sort of the business impact or perceived business impact that could come from any of that, which I don't know, maybe rightfully so. But anyway, I'm a B. Great to dive into this stuff. Listeners, we'd love to hear your thoughts too, especially how you would grade Twitter on this aspect. Join us at acquired.fm slash slack. And David, into our carve out?

On to carve outs. And I think we, all of us, probably Dick included, have just one carve out for all of you today. And it's a big one for everybody who is a US citizen, go vote. It's the most important thing you can do. And there's never been a more important time to do it. Amen. Well, that about brings us home. Thanks so much to Dick for coming on. You guys are awesome. Thanks for the time. I appreciate it. It was fun.

And listeners, if you aren't subscribed and you like what you hear, you should. And if you liked this episode and you have a friend that you frequently talk with Twitter about or a co-worker, or you're interested in just some of the sort of business concepts discussed here, you should absolutely share this episode with said friend or co-worker. Share it on social media. I think we may have talked about a few interesting websites upon which you can do so.

As always, if you love acquired and want to hone your own craft of company building, you should become an acquired LP. You get access to the LP show with awesome interviews with people talking about the nitty-gritty of company building, be it raising money, be it figuring out your pricing strategy, hiring, retaining great employees. David, I'm about to start making stuff up and then hopefully we can do some episodes just so that we can fill it in in the future.

But I think there is no better place on the internet to learn about the craft of company building and to join a community of people who are obsessed with just that. So if you aren't already a member, you can go to acquired.fm slash LP, get that seven-day free trial. And you can also join us in the Slack. We're hanging out at acquired.fm slash Slack. With that, listeners, we will see you next time. See you next time.