The Exchanges

Q phase of like, actually, we can get glitzy people who don't really relate so much to our core, which is startups Venture, growth, and we had kind of famous people on, and the North Star misled us, and it was the wrong North Star. If I say to you, when do you think you had the wrong North Star or wrong goal structure, and what did you learn from that?

A I'll be honest, I think Mark is extraordinarily consistent. The biggest place, which I talk about in the book, is before I joined, it felt like the North Star wasn't clear within the leadership of the company. And that was the famous time when Mark nearly sold the company to Yahoo because it felt like the North star to a lot of the executives was it's a startup. We need to make it big and exit and become millionaires. And for Mark, obviously the North star was connect the world online, do what he felt would make the world, you know, a better place. And that was at odds. And I think that led to a lot of the executive team saying, I want to sell to Yahoo and Mark saying, I want to stay independent. And we turned over the entire executive team. And then out of that fire, which I think really forged the company, we came to much more clarity on what the North Star was, so everyone knew what they were signing up for.

AI assessment note: “out of that fire, which I think really forged the company, we came to much more clarity”

Q You said to him once, and he says it many, many times a day, he said, retention is king. Okay, so retention is king, and you go then from hang on a minute, maybe all guns blazing on acquisition is not the right approach and we should switch focus to retention. How does that change what you do? How does that change product decisions? How does that change strategic decisions?

A Well, I'll go, I'll go back to eBay for this one. There was a major strategic change we made at eBay when I was there, and that was moving from confirmed registered users, so users who registered and then confirmed their, we had this concept of the funnel from registered users, people who just registered, confirmed registered users, you confirmed your communications token, phone number, or email, activated confirmed registered users, you'd placed a bid or you bought an item or you listed an item. The affiliate program when I was an affiliate for eBay was based on confirmed registered users. And when you focused on confirmed registered users, you landed people on the registration page because you wanted them to register. So you search on Google for trampoline, we will land you on the registration page. You search on Google for kettle, we will land you on the registration page. We don't care. Registration page. When we flipped it to activated confirmed registered users, it completely changed the way that the affiliate business worked and the entire online marketing business worked. And this is what drives retention. If you sign up for eBay and you do an action, bid, buy, or list, your long-term retention goes up tremendously. So what it changed was instead of just having the keyword list prioritized by how many registrations you'd get, You would look at the keyword list prioritiz…

AI assessment note: “you would change the landing page from the registration page to the search results page”

Q phase of like, actually, we can get glitzy people who don't really relate so much to our core, which is startups Venture, growth, and we had kind of famous people on, and the North Star misled us, and it was the wrong North Star. If I say to you, when do you think you had the wrong North Star or wrong goal structure, and what did you learn from that?

A I'll be honest, I think Mark is extraordinarily consistent. The biggest place, which I talk about in the book, is before I joined, it felt like the North Star wasn't clear within the leadership of the company. And that was the famous time when Mark nearly sold the company to Yahoo because it felt like the North star to a lot of the executives was it's a startup. We need to make it big and exit and become millionaires. And for Mark, obviously the North star was connect the world online, do what he felt would make the world, you know, a better place. And that was at odds. And I think that led to a lot of the executive team saying, I want to sell to Yahoo and Mark saying, I want to stay independent. And we turned over the entire executive team. And then out of that fire, which I think really forged the company, we came to much more clarity on what the North Star was, so everyone knew what they were signing up for.

AI assessment note: “out of that fire, which I think really forged the company, we came to much more clarity”

Q And the Coke moving to online, they still are doing brand marketing with that then, aren't they? Because they're not like, they don't expect me to go and click, oh, I'm going to buy a 12 pack through this click through. It's still brown.

A And their online is brilliant. There were two campaigns they did that I thought were absolutely fantastic. One was called Coke creation, where creators were being asked to play with the Coke logo and Coke was co-creating with anyone who wanted to do something cool with the Coke logo. And if you look into it, there are some really cool things that people did. I mean, I, I'm a sucker for flowers and there was one that was all flowers, but it was incredibly creative. And the other one they did was share a Coke. Which I think, if I remember correctly, revolves around the Cokes that have people's names on them, and people posting pictures of them with their Coke with their name on, which a year or so ago you saw a lot of online. That was a deliberate move by Coke. And so, yeah, those are brilliant brand campaigns. They've moved from the box of television and thinking about the 62nd slot or the 32nd slot and the tyranny thereof, and they've thought outside the box. And it's been a really successful thing, especially in places like Brazil and India, these big growth markets for them.

AI assessment note: “And so, yeah, those are brilliant brand campaigns.”

Q And the Coke moving to online, they still are doing brand marketing with that then, aren't they? Because they're not like, they don't expect me to go and click, oh, I'm going to buy a 12 pack through this click through. It's still brown.

A And their online is brilliant. There were two campaigns they did that I thought were absolutely fantastic. One was called Coke creation, where creators were being asked to play with the Coke logo and Coke was co-creating with anyone who wanted to do something cool with the Coke logo. And if you look into it, there are some really cool things that people did. I mean, I, I'm a sucker for flowers and there was one that was all flowers, but it was incredibly creative. And the other one they did was share a Coke. Which I think, if I remember correctly, revolves around the Cokes that have people's names on them, and people posting pictures of them with their Coke with their name on, which a year or so ago you saw a lot of online. That was a deliberate move by Coke. And so, yeah, those are brilliant brand campaigns. They've moved from the box of television and thinking about the 62nd slot or the 32nd slot and the tyranny thereof, and they've thought outside the box. And it's been a really successful thing, especially in places like Brazil and India, these big growth markets for them.

AI assessment note: “And so, yeah, those are brilliant brand campaigns.”

Q You said to him once, and he says it many, many times a day, he said, retention is king. Okay, so retention is king, and you go then from hang on a minute, maybe all guns blazing on acquisition is not the right approach and we should switch focus to retention. How does that change what you do? How does that change product decisions? How does that change strategic decisions?

A Well, I'll go, I'll go back to eBay for this one. There was a major strategic change we made at eBay when I was there, and that was moving from confirmed registered users, so users who registered and then confirmed their, we had this concept of the funnel from registered users, people who just registered, confirmed registered users, you confirmed your communications token, phone number, or email, activated confirmed registered users, you'd placed a bid or you bought an item or you listed an item. The affiliate program when I was an affiliate for eBay was based on confirmed registered users. And when you focused on confirmed registered users, you landed people on the registration page because you wanted them to register. So you search on Google for trampoline, we will land you on the registration page. You search on Google for kettle, we will land you on the registration page. We don't care. Registration page. When we flipped it to activated confirmed registered users, it completely changed the way that the affiliate business worked and the entire online marketing business worked. And this is what drives retention. If you sign up for eBay and you do an action, bid, buy, or list, your long-term retention goes up tremendously. So what it changed was instead of just having the keyword list prioritized by how many registrations you'd get, You would look at the keyword list prioritiz…

AI assessment note: “you would change the landing page from the registration page to the search results page”

Q That's not my charm, my friend. Uh, but my point being, Is it really still friending? Is it really still following? Given the fact that actually Reels is so dominant, time on side is so dominant, is it not attention?

A I think it's all of the above. Where do you get your, so you look at TikTok and what they're doing right now. They're focusing very heavily on messaging. They're focusing very heavily on friend graph. Um, so clearly it's important. They're bringing it back in Twitter, Snapchat, these companies focus on following, like how often does YouTube prompt you to still follow an account? I mean, I'm sure you want people to press the little notification budget button. If you haven't, please do. Um, so It's part of the mix. It isn't a hundred percent anymore, and you shouldn't treat it like it's a hundred percent anymore. Um, but it's definitely part of the mix, and where do you get that first bit of data on what to show to people? It's who they follow, and who they friend, and what they like.

AI assessment note: “I think it's all of the above.”

Q said about the move from direct to self-serve being so transformational, and I was like, huh, walk me through that, and when he told it to me, it was kind of so momentous, but also relatively untold. So tell me about the move from direct to self-serve, which was very much part of your Role and, and experience. Why did you make the move and how did that go down?

A Well, I mean, firstly, you, you had Brian on here who was my closest partner and really led to the work. So you've already kind of heard the story on the podcast. Brian Hale, um, told it well. It was, it was, we actually have a lovely blend. It isn't like direct sales to self-serve. It's more like making sure that people can click the button to boost an ad. I think Brian told you the promote Like we can give the post a promotion. Is it now a VP versus boost post? People knew what it was. So it was just like really basic stuff about presenting to small businesses. Hey, you can buy a lightweight ad by boosting this post as a sponsored story, putting it in people's way when they're using their own page saying, oh, buy an ad. Very, very basic stuff like that that mattered. There was, there was one funny, we, we mucked up a translation. This was the first wave of this work. And we said, instead of advertise, when we translated it to French, we said, create an ad. And in France, we had a much larger rate of advertiser growth, like really much rather a larger rate, like 30, 40%. We were doing an analytics test with a, with a vendor, and they found that France was far higher, and we found it was this mistranslation, so we changed everywhere on the site. Instead of saying advertise, we changed it to create an ad, and we step changed up advertiser growth. So we got all these people self-…

AI assessment note: “we changed everywhere on the site. Instead of saying advertise, we changed it to create an ad”

Q Should we have psychologists in our product teams and in our brand teams? When you say about creating an ad being that transformational moment and actually making a very significant shift, Should we have psychologists and people who understand human behavior much, much better integrated into our teams?

A Maybe. I mean, the lady who used to head our content design, Alicia, who now she's head of design and creative and content design for me for a third of the company. Amazing lady. She has a psychology degree. And she was an incredible friend and partner in actually getting these tests run. Um, at the same time, if we were American, we would say we want a lot of times at bat. We're not looking for home runs. We just want a lot of opportunities at bat. And so I also think you should just be willing to try a lot. If you think ahead of time, you can predict what people are going to respond to with a button color, a size of button, the words you put on the button, the flow that you've created. Like Brian gave you examples of this. There are flows where we made them longer. By giving people more targeting options where we got a higher conversion rate and a higher lifetime value. Whereas the growth standard is shorten the flow and take out steps. So I think you need to try more experiments because you are not a genius. And even if you are, you can't predict what every user is going to do or think based on what you say to them.

AI assessment note: “Maybe. I mean, the lady who used to head our content design... She has a psychology degree.”

Q We're gonna get to it. Um, I, I want to ask about the Facebook Messenger split. We did touch on it. Like, talk to me about that, because that seemed like a pretty pivotal product moment for the company and for the product. Why did you do it, and how did that go down?

A Yeah, I mean, it's a long story and, and, and probably too long for this podcast, but fundamentally in the shift to mobile, we'd actually done very well with messaging on the web. Um, the product that was built there, I think Chris Putnam built the original version of it. I can't remember. It was, I think it was called Chirp. Anyway, we, we built this chat on the web, integrated, and it did really well. Like, it grew tremendously, and it was growing, growing, growing, and then mobile came along We launched mobile messenger. We were tracking it and mobile messages were growing. Great. But we were measuring it the wrong way. We were actually losing market share because the web was declining. Faster than mobile was growing for us, and then WhatsApp and iMessage and all of these things were coming along, and actually the shift of platforms meant even though we were growing mobile messaging, we were growing it slower than the industry, and we were losing market share in the shift to mobile. So that was when we sort of went, oh, actually this strategy isn't working of the hybrid. We looked at what was correlated with people using messaging the most, and it was them using the Messenger app, and we investigated with user research. Why? And it was because they knew that when people had the messenger app, they would respond instantly to their messages. Whereas when they were messaging in…

AI assessment note: “fundamentally in the shift to mobile, we'd actually done very well with messaging”

Q Are these really risks? If you actually think about it, like reels, if it didn't work, it's not a sink the boat decision in a Bezos world, is it?

A I mean, I think like it is to some extent because of the limited resources, like you only get to make a few bets, and if you make the wrong bet and you miss the next wave, you're in real trouble as a tech company. Look at how Microsoft languished for a very, very long time. Um, so I, I do actually, I mean, it's not gonna, this is the point of absolutes. Is it gonna sink your company? No. But it really matters you don't miss lots of different waves as a company. Um, but when you're a bigger and more mature company, you can afford to do some longer term experiments. So we did it in these countries, and we watched how those countries evolved while we perfected the design and the methodology for the rollout. Personally, I think we took too long, and if I went back and did it again, I would have moved a lot faster, which is weird for a meta executive to say, but I actually think we moved too slowly, and in many ways it was my fault. Because I wanted to measure everything, but the data was completely clear, and then we rolled it out globally, and even though we had that small initial dip, we could argue it to everyone whose metrics got hurt, and say the dip is worth it, because look at Romania, look at Portugal, look at the Czech Republic, look at Hungary.

AI assessment note: “Is it gonna sink your company? No. But it really matters you don't miss”

Q Do you see what I mean? So I'm just trying to understand if actually speed does matter, you learn more from being second, you can perfect more, and actually, is it sometimes best?

A I think it's good to get out and iterate. Personally, I think generally it's okay to be second, and we're totally fine being second. Catching up on the basics, and then trying to out-innovate. Like step, step, step. Like launch, catch up, overtake with innovation. I think is a very, very good model. You don't want to be third or fourth, and actually when someone's got to the point of product market fit, you really need to move fast. And the other thing that move fast gives you, and again, um, the lady from Growth Hackers was, was awesome. She was saying, like, you get to get more experiments in. I think Brian said, I think, look, I think you had multiple people in here say, The, the pace of learning really, really matters, and so no, getting out shipping and trying, the first version of Reels failed. We tried to put it in the stories spot, like we tried to integrate it that way, it hard failed. Getting a couple of revs on that, we were out before YouTube, we were out before anyone except TikTok with our competitor there, faster than Snap, who also cloned and copied TikTok despite everything they've said about that, um, but like we were out, um, faster than everyone else, and we got a couple, our first iteration failed, Second and third, we learnt a lot. Now we're actually innovating and adding really cool things that people aren't trying other than us. We wouldn't be here if we…

AI assessment note: “Personally, I think generally it's okay to be second, and we're totally fine being second.”

Q said about the move from direct to self-serve being so transformational, and I was like, huh, walk me through that, and when he told it to me, it was kind of so momentous, but also relatively untold. So tell me about the move from direct to self-serve, which was very much part of your Role and, and experience. Why did you make the move and how did that go down?

A Well, I mean, firstly, you, you had Brian on here who was my closest partner and really led to the work. So you've already kind of heard the story on the podcast. Brian Hale, um, told it well. It was, it was, we actually have a lovely blend. It isn't like direct sales to self-serve. It's more like making sure that people can click the button to boost an ad. I think Brian told you the promote Like we can give the post a promotion. Is it now a VP versus boost post? People knew what it was. So it was just like really basic stuff about presenting to small businesses. Hey, you can buy a lightweight ad by boosting this post as a sponsored story, putting it in people's way when they're using their own page saying, oh, buy an ad. Very, very basic stuff like that that mattered. There was, there was one funny, we, we mucked up a translation. This was the first wave of this work. And we said, instead of advertise, when we translated it to French, we said, create an ad. And in France, we had a much larger rate of advertiser growth, like really much rather a larger rate, like 30, 40%. We were doing an analytics test with a, with a vendor, and they found that France was far higher, and we found it was this mistranslation, so we changed everywhere on the site. Instead of saying advertise, we changed it to create an ad, and we step changed up advertiser growth. So we got all these people self-…

AI assessment note: “we changed everywhere on the site... and we step changed up advertiser growth.”

Q What do you mean a metric never perfectly describes a goal? Because I often try and we do OKRs as teams. Um, and it's, it's actually very challenging to ensure that they are very measurable. They tie exactly to the goal. How do you mean about that kind of metric and tied to the goal?

A Every metric has a flaw. Active users has fake accounts like Elon Musk has been telling us with bots, but it matters. You have to have the counter metric. So you can easily do something like registrations, right? Like we write the beginning of our, our interview. Registered users, confirmed registered users, active confirmed registered users, monthly active users. There's been this journey of the industry over the last 20 years. It all started with everyone claiming about registered users, but it had the negative side effect of them not being confirmed. Confirmed, they weren't active. Activated, they weren't retained. And so actually, like, every goal has, uh, the goal is to have lots of people use your service. The way people have measured it is from registered users through to monthly active users through to Now we even try and measure people. There are flaws at every step of the way, um, and if you don't acknowledge and understand the flaws, you will succumb to good arts law, where it's not a good measure anymore.

AI assessment note: “Every metric has a flaw. Active users has fake accounts”

Q to really make everyone feel like they can come out when they feel they want to and feel welcome in whatever way they are? Like, I think there's, you know, people talk a lot about, you know, I, I feel I can now. I don't actually feel that that's the case at all in a lot of companies. Do you feel that we have a long way to go still?

A Yeah, I mean, I think that the trans community has it the hardest. So like, to be completely blunt, as a white cis gay man, I've got it, you know, Relatively easy compared to many people in the community. Um, at the same time, I've said that on stage before, and people have corrected me that everyone has their own journey, and so it's not that it's straightforward. You know, look around. How many boards do you serve on where there's another gay board member? How many investor conferences do you ever see someone gay on stage? Seniors, how many people have you had on this show who are openly gay, who are a senior executive at a tech company? There's still a gap, and that alone should tell you that there is still a problem, and You know, I'm, I'm very lucky. I work at a company. Mark had a gay board member with Peter. He had a gay co-founder with Chris. Like, I know he supports our community, and I feel very grateful for it. Cheryl too. And I feel that from my colleagues. But it's also, like, still very lonely. Um, in the senior teams I'm on, in the boards I've served on, like, I'm always the only gay guy. So yeah, there's a, there's a gap, and we have a long way to go. And I talk to many LGBTQ people, and again, look, trans people, way, way harder than where I stand. And many of them say to me, I can't be successful in my job, and I can't be out in my work. And I always say, that…

AI assessment note: “So yeah, there's a, there's a gap, and we have a long way to go.”

Q So what, what do we want to have done in the next five years?

A I think that we want to turn reels into a really major monetization engine, not just an engagement engine. I think that we want to have a much firmer understanding of the success that the metaverse will allow for people, um, probably grounded in virtual and mixed reality with augmented reality further on the horizon, but a proof case out there. And I think like, yeah, we want to have put to bed a bunch of the things that we're discussing around social media Uh, in the public sphere where the public realm has to make decisions about what they want on content moderation, what they want on well-being, we need the appropriate laws in place, and then we need to be moving to a place of compliance with those laws, because society has settled some of the decisions about what they want out of social media, and that's ok.

AI assessment note: “we want to turn reels into a really major monetization engine”

Q We're gonna get to it. Um, I, I want to ask about the Facebook Messenger split. We did touch on it. Like, talk to me about that, because that seemed like a pretty pivotal product moment for the company and for the product. Why did you do it, and how did that go down?

A Yeah, I mean, it's a long story and, and, and probably too long for this podcast, but fundamentally in the shift to mobile, we'd actually done very well with messaging on the web. Um, the product that was built there, I think Chris Putnam built the original version of it. I can't remember. It was, I think it was called Chirp. Anyway, we, we built this chat on the web, integrated, and it did really well. Like, it grew tremendously, and it was growing, growing, growing, and then mobile came along We launched mobile messenger. We were tracking it and mobile messages were growing. Great. But we were measuring it the wrong way. We were actually losing market share because the web was declining. Faster than mobile was growing for us, and then WhatsApp and iMessage and all of these things were coming along, and actually the shift of platforms meant even though we were growing mobile messaging, we were growing it slower than the industry, and we were losing market share in the shift to mobile. So that was when we sort of went, oh, actually this strategy isn't working of the hybrid. We looked at what was correlated with people using messaging the most, and it was them using the Messenger app, and we investigated with user research. Why? And it was because they knew that when people had the messenger app, they would respond instantly to their messages. Whereas when they were messaging in…

AI assessment note: “losing market share in the shift to mobile”

Q What can leaders do? And what do you do to create a culture where experimentation is encouraged? And actually, Alex, you know what? That didn't work. That's fine. We've got deadlines and I've got someone, you know, calling time, but that's fine. I'm, I'm pleased we're doing this. How do you create that culture for experimentation?

A Well, our company, I think, um, Shrepp, our former CTO, uh, now senior fellow, and Jay Parikh, who ran the infrastructure team, probably deserve a lot of credit for making Meta a culture where you can make mistakes, and we try not to enforce blame. I mean, look, if you do the same thing five times, you muck up the same five ways, you will have a performance conversation with your manager, but broadly, they've built a culture from engineering where we have these several views when things go wrong, And we try and learn from them, and if you made an honest mistake, that is completely okay, and we want you to step up, do it, say it, and that tries to take blame out the culture and makes people feel safe to experiment. We try and maintain that across the whole company.

AI assessment note: “takes blame out the culture and makes people feel safe to experiment”

Q What do you mean a metric never perfectly describes a goal? Because I often try and we do OKRs as teams. Um, and it's, it's actually very challenging to ensure that they are very measurable. They tie exactly to the goal. How do you mean about that kind of metric and tied to the goal?

A Every metric has a flaw. Active users has fake accounts like Elon Musk has been telling us with bots, but it matters. You have to have the counter metric. So you can easily do something like registrations, right? Like we write the beginning of our, our interview. Registered users, confirmed registered users, active confirmed registered users, monthly active users. There's been this journey of the industry over the last 20 years. It all started with everyone claiming about registered users, but it had the negative side effect of them not being confirmed. Confirmed, they weren't active. Activated, they weren't retained. And so actually, like, every goal has, uh, the goal is to have lots of people use your service. The way people have measured it is from registered users through to monthly active users through to Now we even try and measure people. There are flaws at every step of the way, um, and if you don't acknowledge and understand the flaws, you will succumb to good arts law, where it's not a good measure anymore.

AI assessment note: “Every metric has a flaw.”

Q the Canva. The guys at Canva couldn't raise a dollar. He went to a 150 VCs. Couldn't raise a dollar. And so my, so no, we don't actually, and one person can write a check. Very important for me. Um, but how do I change that? His goal is 20 companies per week to meet. If I put it to investment, one investment per month, he can game that too.

A Yeah, totally. I mean, counter metrics is the way I talk about it. So, you know, you could look at it as 20 companies per week, um, with a benchmark of you need to get at least one of them to investment grade. Like, that could be a counter metric. Like, nothing is perfect. This is where it comes back to, again, my boss is saying, common sense is the least common of all senses. At the very bottom, you need to actually look at it and see if it's being gamed. Like, as an analytics leader, as a product leader, as a company leader, no metric is perfect. No set of counter metrics are perfect. You still need to put logic on the top and common sense on the top.

AI assessment note: “counter metrics is the way I talk about it”

Q with them? Like my mother, she's the light in every room, I always say, and I always want my meeting with someone to be hopefully one of the highlights of their day, because I want them to be happy as well, which is also challenging as a leader. You can't always have that. Thanks, Mum. But I, I really try and be the light in the room. What are yours?

A When I come back earlier to that point on being kind of long-term kind, I actually think that's that, I got that from my parents. My parents, I think my parents do the right thing, even if it's hard for them, and I think I've learned that from them, uh, no matter what the scenario is, like, My mom, my dad, they try and do the smart thing, they try and do the right thing, and I try, and that's very much a value in our family. Like, huge value, so that's huge. And then being patiently right, I, I think that fits with my dad. Um, he's like, he doesn't bang his chest, he isn't sort of the, the, the trying to make a, a big deal of himself, but like, he's quiet, and when he says something, he's generally right, um, and you want to listen. And I've also learned a lot about how to actively listen by having a father who English isn't his first language, and I have a boss now who English isn't his first language, he's brilliant English, both of them, but actually listening carefully, I think I learned that from my dad, um, from that, so, but yeah, I think, I think, like, long-term kind, doing the right thing, if you believe it's the right thing, even if people are annoyed at you, or it's really hard to do it, uh, that is something I've learned from both of my parents.

AI assessment note: “being kind of long-term kind, I actually think that's that, I got that from my parents.”

Q That's not my charm, my friend. Uh, but my point being, Is it really still friending? Is it really still following? Given the fact that actually Reels is so dominant, time on side is so dominant, is it not attention?

A I think it's all of the above. Where do you get your, so you look at TikTok and what they're doing right now. They're focusing very heavily on messaging. They're focusing very heavily on friend graph. Um, so clearly it's important. They're bringing it back in Twitter, Snapchat, these companies focus on following, like how often does YouTube prompt you to still follow an account? I mean, I'm sure you want people to press the little notification budget button. If you haven't, please do. Um, so It's part of the mix. It isn't a hundred percent anymore, and you shouldn't treat it like it's a hundred percent anymore. Um, but it's definitely part of the mix, and where do you get that first bit of data on what to show to people? It's who they follow, and who they friend, and what they like.

AI assessment note: “I think it's all of the above.”

Q What can leaders do? And what do you do to create a culture where experimentation is encouraged? And actually, Alex, you know what? That didn't work. That's fine. We've got deadlines and I've got someone, you know, calling time, but that's fine. I'm, I'm pleased we're doing this. How do you create that culture for experimentation?

A Well, our company, I think, um, Shrepp, our former CTO, uh, now senior fellow, and Jay Parikh, who ran the infrastructure team, probably deserve a lot of credit for making Meta a culture where you can make mistakes, and we try not to enforce blame. I mean, look, if you do the same thing five times, you muck up the same five ways, you will have a performance conversation with your manager, but broadly, they've built a culture from engineering where we have these several views when things go wrong, And we try and learn from them, and if you made an honest mistake, that is completely okay, and we want you to step up, do it, say it, and that tries to take blame out the culture and makes people feel safe to experiment. We try and maintain that across the whole company.

AI assessment note: “we have these several views when things go wrong, And we try and learn”

Q I'm so pleased we went for lunch beforehand, and I feel built a rapport because I can ask a harder question. Why Meta as a name? I think people instantly think of Metaverse, which is transient in terms of consumer excitement. Why, why Meta?

A I think we're mapping out the future that we believe in. Like, we're a company that does connect people. The red thread in our brand, like, Meta is about connecting people in innovative and inspiring ways. At the extreme, that's virtual and mixed and augmented reality experiences, but honestly, that starts with intimate conversations over WhatsApp, and so our, our brands in our branded house are Meta, we connect people in awesome and inspiring ways, innovative. Facebook, it's about social discovery. Um, the first thing we ask you is find your friends, people you may know, pages you may like, groups you should join, social discovery. Instagram is connecting around creativity, whether that's resharing your favorite reel or posting a photo with a, with a, with a filter. And WhatsApp is intimacy through privacy. You can only be as intimate as you can possibly be and share the most important things to you if you know there are multiple layers of privacy protecting you. And so all of these things come together as ways people connect virtually, and what is the metaverse? It is a place where people connect virtually, and the end point of it is an embodied space that ideally goes with us the entire time because of augmented reality, but the steps on the way start with two D interfaces where you are connecting virtually.

AI assessment note: “I think we're mapping out the future that we believe in.”

Q the Canva. The guys at Canva couldn't raise a dollar. He went to a 150 VCs. Couldn't raise a dollar. And so my, so no, we don't actually, and one person can write a check. Very important for me. Um, but how do I change that? His goal is 20 companies per week to meet. If I put it to investment, one investment per month, he can game that too.

A Yeah, totally. I mean, counter metrics is the way I talk about it. So, you know, you could look at it as 20 companies per week, um, with a benchmark of you need to get at least one of them to investment grade. Like, that could be a counter metric. Like, nothing is perfect. This is where it comes back to, again, my boss is saying, common sense is the least common of all senses. At the very bottom, you need to actually look at it and see if it's being gamed. Like, as an analytics leader, as a product leader, as a company leader, no metric is perfect. No set of counter metrics are perfect. You still need to put logic on the top and common sense on the top.

AI assessment note: “counter metrics is the way I talk about it”

Q love him too, and he said how bluntly Zuck holds you as one of the core people in scaling Facebook to a billion, uh, to a billion users, obviously many billions now. Um, and I wanted to ask, what would you say are one or two of the single biggest movers in that scaling to a billion? And I know I'm asking for bastardized moments, but I'm just too interested.

A Well, I think firstly, it's really nice of Mark to say that, and he, and he said it publicly, which is really cool, but you can't A-B test if someone was there, and so you don't know what would have happened, so I always, I always, anyway, I appreciate it a great deal, but you, you don't know what would have, what would have happened, and you can never miss off Javi, and Naomi, and people like Brian Hale, who you've had on the show. Danny Ferrante built our entire model for looking at growth. I mean, Dustin Moskovitz personally built Um, phone number registration. There are so many people, so it's nice, but I, I do also think sometimes it's, it's overstated. I think the critical thing was having a data-inspired growth model. So we really understood what was driving our growth, and the first deep insight, and this was Danny Ferrante did it for us, I think he brought it from Yahoo, and we had something like it at eBay, but it wasn't as advanced. Super simple. Everyone uses it now. Growth accounting. Number of registrations. Plus number of resurrections minus the number of people who churn equals your net growth. Super simple. But when we implemented it, we thought it was registrations minus a bit of churn. What we didn't realize was that the churn number and the resurrection number were massively bigger than the registration number. And so we hadn't realized how big a lever Reduc…

AI assessment note: “I think the critical thing was having a data-inspired growth model.”

Q Are these really risks? If you actually think about it, like reels, if it didn't work, it's not a sink the boat decision in a Bezos world, is it?

A I mean, I think like it is to some extent because of the limited resources, like you only get to make a few bets, and if you make the wrong bet and you miss the next wave, you're in real trouble as a tech company. Look at how Microsoft languished for a very, very long time. Um, so I, I do actually, I mean, it's not gonna, this is the point of absolutes. Is it gonna sink your company? No. But it really matters you don't miss lots of different waves as a company. Um, but when you're a bigger and more mature company, you can afford to do some longer term experiments. So we did it in these countries, and we watched how those countries evolved while we perfected the design and the methodology for the rollout. Personally, I think we took too long, and if I went back and did it again, I would have moved a lot faster, which is weird for a meta executive to say, but I actually think we moved too slowly, and in many ways it was my fault. Because I wanted to measure everything, but the data was completely clear, and then we rolled it out globally, and even though we had that small initial dip, we could argue it to everyone whose metrics got hurt, and say the dip is worth it, because look at Romania, look at Portugal, look at the Czech Republic, look at Hungary.

AI assessment note: “Is it gonna sink your company? No. But it really matters”

Q love him too, and he said how bluntly Zuck holds you as one of the core people in scaling Facebook to a billion, uh, to a billion users, obviously many billions now. Um, and I wanted to ask, what would you say are one or two of the single biggest movers in that scaling to a billion? And I know I'm asking for bastardized moments, but I'm just too interested.

A Well, I think firstly, it's really nice of Mark to say that, and he, and he said it publicly, which is really cool, but you can't A-B test if someone was there, and so you don't know what would have happened, so I always, I always, anyway, I appreciate it a great deal, but you, you don't know what would have, what would have happened, and you can never miss off Javi, and Naomi, and people like Brian Hale, who you've had on the show. Danny Ferrante built our entire model for looking at growth. I mean, Dustin Moskovitz personally built Um, phone number registration. There are so many people, so it's nice, but I, I do also think sometimes it's, it's overstated. I think the critical thing was having a data-inspired growth model. So we really understood what was driving our growth, and the first deep insight, and this was Danny Ferrante did it for us, I think he brought it from Yahoo, and we had something like it at eBay, but it wasn't as advanced. Super simple. Everyone uses it now. Growth accounting. Number of registrations. Plus number of resurrections minus the number of people who churn equals your net growth. Super simple. But when we implemented it, we thought it was registrations minus a bit of churn. What we didn't realize was that the churn number and the resurrection number were massively bigger than the registration number. And so we hadn't realized how big a lever Reduc…

AI assessment note: “I think the critical thing was having a data-inspired growth model.”

Q I'm so sorry. Why does no one in our circles use it then?

A Yeah. I mean, it's just, it's a bit more normie. Well, firstly, it's a huge success in like Japan and Taiwan. Let's start with that. Japan, interestingly, Is a text first country. That's why haikus come from Japan and all that stuff. It's a text first pseudonymous society. And so it's a really great product market fit for Japan. And we've always struggled actually. Twitter is bigger than I think Facebook and Instagram in Japan. Insane. And so threads is now really thriving in Japan. So firstly, segments that really care about this is different. Secondly, you know, in general, I think VCs don't tend to use Facebook very much either. And they all kind of act like Twitter is Facebook when they're talking about Facebook in public. You can kind of hear it. You listen to anyone talk about Facebook who's like a VC or a politician or policy elite, we'd call them, or a highly engaged news audience, we'd also call them. They don't really use Facebook or Instagram. They use Twitter. And they think Twitter is Facebook, which is not true. It is the most engaged community on Twitter, I think, is Facebook. VCs and tech people. Yeah, the normies aren't the VCs and tech people, and they use threads, but VCs and tech people use Twitter.

AI assessment note: “the normies aren't the VCs and tech people, and they use threads”

Q I actually skipped those comments, clearly. Um, but how did you go from paper aeroplanes first to eBay?

A Oh, I, I, so I was doing Natski at Cambridge, specializing in physics, and I was trying to pay my way through university on my own without taking any money from my parents, because we'd, um, you know, we'd gone through a dip as a family, and, um, that mattered to me, and, um, I was running these paper airplane sites. Originally I did GeoCities before you were born, and, um, on GeoCities they had Cape Canaveral area, and I had fun science projects, and I had paper airplanes, and I had Formula One, and paper airplanes did well on AltaVista. The way, so I got super excited about learning how to rank in SEO, and the way you ranked back then was you just said paper airplanes a lot on your page, in white, five pages down, which is how I ranked. So then I built a paper airplane website, and I stopped being just interested in traffic, and became interested in monetizing it. And so I started to learn how to make money online. It was interesting. Google back in the day, they had PageRank, which was using the backlinks. What was the most important phase for backlinks? Yahoo directory. You get a link number one in Yahoo directory, you instantly ranked first in, uh, Google. So I got that link, I ranked first, and then I started trying to make money online. I made it via these ad products, and then I started doing affiliate marketing for eBay. Buying money with my credit, buying ads with my …

AI assessment note: “then I started doing affiliate marketing for eBay.”