Bret Taylor’s Journey Leading Salesforce, Sierra & OpenAI

28:51Yeah. How do you think about that? Let me go back to the Google analogy for a second and then answer the question because I think the nuance matters here. I think there was a narrative that is not completely untrue that Google won because PageRank was better. You know, a lot of search engines were reductively based on analyzing the content of the webpage, which made it really prone to spam and made it really hard to figure out which of these webpages talking about this topic was actually canonical. PageRank and looking at links and basically sort of anchor links between webpages had essentially a notion of reputation and importance and credibility in addition to analyzing content and the combination of those two just made it work.

29:37You could search for something and be relatively likely to hit in the top results something that actually was what you were looking for. So it's a really important technical insight. I think some people say, okay, that was Google won because of Patreon. Yeah, well, there's a couple other things. Like Google also signed deals with all of the major portals around the world. Google powered search on Yahoo for a really long time. And if you went around almost every country and every market, you know, so essentially Google distributed its search engine and sort of like, you know, basically got distribution through a really effective partnership structure.

30:09Had they not, that would have opened the market to a lot of other competitors like InktoMe who ended up acquired by Yahoo to go into that market. Secondly, you know, with the, when I got to Google, I don't believe AdWords existed, but there was banner ads at the top. You could like buy a keyword is my recollection. I might be embellishing history here. And then going to not only cost per click based ads, but the way AdWords ranked it, which was proportional to the click through rate too, which incentivized high quality ads in addition to, you know, optimizing for sort of an auction so people could bid on those keywords, created one of the greatest business models of all time.

30:50It was essentially like mapping the demand of, you know, intent on the internet to the supply of people who could fulfill that intent in a very efficient auction. And so then you look at the enemies of the world that were licensing their technology to portals. And you realize at that point, you know, the Google's business model was just better. It was just generating cash, like it's going out of style. And you just couldn't reproduce that in a B2B context. So you have the partnership strategy, you have the business model, which was really important. If Google's business model had remained selling that yellow enterprise search appliance, it would not be the company it is today.

31:31It wouldn't have been able to finance things like Google Maps that I worked on. You wouldn't have had a business model support that. All of it's important. Business model, distribution strategy. Google's costs were much lower than a lot of competitors, you know, because we built our own data centers. and the cost to serve was like meaningfully lower. All those things mean you can just do different things with your technology or product. Okay, so now let's go to the AI market. How you price matters. Are you charging based on consumption? Are you charging for outcomes? Are you charging just a site license to use the product?

32:08That matters a lot. Because, you know, my intuition and the reason why Sear does outcomes-based pricings is outcomes are valuable. You know, people know the value of an outcome. If you end up charging proportional to technology, you're probably going to end up marginally, just you'll have a thin margin on top of the price of AI, which is going down rapidly, like even as we speak. Similarly, what is your cost to serve? You know, I think that there was a lot of companies that, you know, created, you know, two and a half years ago where they went into places like Kleiner and said, we have all these great AI researchers.

32:44We're going to train our models because we're really smart. Well, it turns out that spending, you know, tens of millions of dollars to train your own model when the foundation model market is sort of commoditizing, turns out to be a really unintelligent use of capital, at least in the applied AI market. Most of those companies have been consolidated now at this point. You just can't have the business model of a pharmaceutical company where you spend all of your capital and license it out. And so, you know, and then similarly, like, what is your go to market model? Is it product led growth? Is it a developer motion like cursor?

33:17Is it an enterprise software motion like Sierra? That also impacts your cost of book. It impacts what customer, what parts of the market you're able to address. And it also impacts things like attrition when new people come around. You know, the reason why Google was able to overtake AltaVista was you just visit a new URL. It's really easy. That's always been Google's argument, you know, for when people say they have a monopoly. It's like, just go to a different search engine. They're not completely wrong on that, right? Right. And, you know, in contrast, you know, if you look at, say, a large ERP deployment, you're, you know, it's hard to rip out an ERP system.

33:55Every large company in the world has had these like multi-year ERP system replacements. And that's an example of like a market that's like slow to address, but very low attrition when you address it. So I think all these things matter. So I don't think like a first mover advantage on its own is a sustainable advantage. And in fact, if you look at the dot-com era, it wasn't always maybe even less likely if you were the first mover. But I don't think it's causal necessarily. I think that what really matters is like what is your sustainable sort of like technology point of view and advantage? What is your go-to-market model?

34:34How do you manage your customer relationships? What is your business model? You know, I think like a pricing model and I think all of these things matter. The fun part about new technology is we didn't have cost-per-click ads in the early days of the internet, and now it's pervasive. Now we can even do conversion-based ads and things like that. I think the exploration going on in the industry on our approach to outcomes-based pricing is really novel. It's new. I think it's a really good idea. We'll talk in 10 years, and I'll tell you whether it was actually a good idea. I don't think it's going to take that long enough.

35:08I think we'll find out soon. And I... Alcome-based pricing meaning? We charge when our agents autonomously resolve the problem and it's free otherwise. And we do that so that at Sierra, our business model is aligned to your outcomes. It's very new. And it's very much like the difference between impression-based ads and cost-per-click ads. These are things that excite me right now is I think we're going to find some key conceptual ideas that are hard for incumbents to copy, maybe give you a different angle than your competitors. And I think just like you can, the narrative I had around Google, which is probably at least half right, you know, on why they succeeded, um, with the benefit of hindsight, we'll know which of those key insights were actually like the meaningful ingredients to winning in these really big markets.

35:57Yeah. It is interesting to see it play out. Like you're riding on the wave of the frontier models, right and so you have open ai and xai and all of them right they're all pushing and we were having this conversation i was having it with um one of my partners ev yesterday he was like you know i don't really use perplexity anymore and i used to live in it and he was like everything is now just in chat gpt you know like they were first and it turns out all of the things that people thought you were going to build on top of ChatGPT, ChatGPT is just going to build in many ways. And now that they have memory, a founder that I was with in New York last week was like, I think this is going to be the stickiest piece of software ever, ever.

36:54If that all plays out, right now, consensus is all of a sudden in the valley that OpenAI is a runaway train because they have all of these first mover advantages. I don't know. We'll see. Super interesting. Well, OpenAI is synonymous with AI for most of the world. It's the Kleenex of the facial tissue market. And you combine that with what, obviously I'm totally biased, but I think by far the best models and by far the best product in ChatGPT built on top of it with memory web search deep research is a mind-blowingly great product i don't know if you use it much but just just amazing it's like having a full-time intern working for me uh it's just amazing research plus memory yeah that's the like killer combo for me yeah was like oh my gosh like you can go do all of this reasoning work and it's coupled with all of my personalized contextualized history.

37:57Yeah. It's like magic. It's amazing. And, uh, yeah, I think broadly, I, I do think chat GPT will be the enduring consumer brand from the AI era. There might be more, um, in general, like when, like a couple of years ago, someone asked me about how I thought the ad market would play out. I think I still have the same opinion, which is, I think there'll probably be one, you know trillion dollar consumer brand coming out of it and like 10 or so uh enterprise software companies um that are less valuable but extremely valuable you know just you can't quite because consumer scale just it just produces so much value and i think i think chat tbt is that brand i mean it is it doesn't mean there can't be more than one it's just uh consumer is harder than than enterprise and it's much more of a winner take all market, not just because of market dynamics before the dynamic of like the Kleenex facial tissue thing.

38:56Like ChatGPT is synonymous with AI. And then if it's a great product, like I think it is, people come back to it. And I have the same experience. Like I always try to use everything just because I'm so curious about this. But for me, I just always go back to ChatGPT because it does, it's just such an effective product. And it's amazing. I mean, it's amazing. I love learning things with it. I love using it as a creative foil. It's a product that has so changed the way I live. It reminds me of Google in that respect. I just couldn't imagine the internet before it. And I used it a hundred times a day and I still use Google, but I think ChatTBT is the most impactful technology product for me of the past two decades easily.

39:36Yeah, I agree. And if it plays out like the Google world, then yeah, there'll probably be$1 trillion company. If you had to give up your smartphone or give up using ChatGPT and all of its competitors, like using an AI system, which would you give up? Post-memory, it has become a very difficult question because, let me put it this way, every day that I spend more time in these systems, let's take open AI as the system that I spend the most time in, every day I get convinced that I've never seen anything like this. Yeah. By the way, you know, we talked about AGI and what's the meaning of it. If you talk to anyone four years ago and you ask them to describe AGI, it's, we're probably beyond what people would have said.

40:25I don't like any of these labels of like agents and AGI. Cause I'm like, even this agentic thing, it's like, there's already a lot of work being done on our behalf. You know, like what's the, where do we, where do we superimpose a line of full autonomy? you know i don't know it's like or when does autonomous for cars it's like it's like are we splitting hairs here um but yeah i think like uh i've never experienced um i've never experienced anything like this yeah where the more that i use it the more that i'm like oh my gosh it did that but then there's also things where i'm like it's re like i will refresh questions with different models and then i will re-ask the same question and get a different answer like there's something fundamentally that i'm not habituated to of how i interact with this thing does that make sense like there's just something it also feels i always give sam and tim so much shit for like the naming it's like you know get rid of 10 models yeah it's like the menu of models at the top.

41:34It's comically bad. But I actually like it right now because the technology is immature. Ideally, you should just be able to go say something that does the right thing, chooses the right model for the right task. I like the fact that it's not just you and me. I'll talk to family members outside of tech and they're using these things. They have opinions on them. They're like, oh, I tried out O3 and here's what I learned about it. I think we're all understanding the contours of of these technologies i remember growing up uh and i wasn't into computer programming until uh later in high school and but i knew like the graphics card in my computer and i knew what games could power it and i you know understand like if i wanted to do this thing like i got i understood like more of the architecture of my computer than i do the architecture of this phone totally because you kind of needed to to like be like to use a pc back then like it's like in the same way in the early days of the automobile like most people knew how to like be a mechanic basically because it would often break down to the side of the road it's clear that like the uh it'll feel like a vestige of the past that the menu of models is like with 100 certainty that's not going to be something we experience five years from now but i also like the fact that like we as users of this technology can kind of experience the contours of and see the chain of that reasoning.

42:57You'd be like, oh, that's interesting. That's kind of gives you kind of an intuition that works. I'll give you one to really kind of, it's going to sound like a not really a thing, but I think it's related. One of the things when Clay and I started Sierra, we were saying the first time an AI agent that is performing customer service on behalf of one of our customers hallucinates something, how are they going to react? Are they going to expect that the technology is perfect 100 % of the time? Because no matter how good, and we're really proud of our guardrails and knock on wood, haven't had a social media moment with our platform yet.

43:27It's still a non-deterministic technology. Like you can't promise that it'll never do something wrong. Even by the time our first agent launched, ChatDBT had become so popular that everybody in the executive team of every company you work with had experienced hallucinations. Like there was the famous like lawyer who made up the case law and the, I think it was Air Canada that like made up a bereavement policy. And so when we were explaining the risks of these technologies, they're like, oh yeah, I get it. It's almost like explaining that occasionally your phone freezes and you need to restart it.

43:58And I was like, wow, in like six months of this technology existing, everybody in the executive team of all these mainstream brands understands like the nuances of its idiosyncrasies and what could go wrong. I think that's great because as society grapples with the impact of AI. For example, what level of decision making an agency should we forward in AI around customer service? That might be a simple question. What about making a financial decision? What about making a decision and something related to national security? You almost need the accumulation of all of this like intuitive understanding of the contours of the technology that isn't just relegated to the technology industry, but actually society writ large so that we can have a conversation about the collective risks.

44:47I always like to say like the main reason GPT-8 will be safe is because GPT-1 through 7 were safe through responsible iterative deployment and we as a society grapple with the impact. Right now it's things like hallucination but there's things like how sycophantic should the could the models be like when does that cross a line of inappropriate jailbreaking all those things. There's so many nuances we need to grapple with as a society when you apply this to like real world situations, I think it's great that you have that silly dropdown selector so you can actually get a sense of like, why are these models different from each other?

45:24How does that form change my opinion of this technology? And I think that will actually accumulate to better societal decision-making over time. It reminds me a lot of, like I grew up in the cloud security industry. My operating career was when AWS was basically serving Capital One and a few other companies like Airbnb and a couple tech companies. And I remember the transition happened over maybe two years where the overwhelmingly obvious value that came from the public cloud was so much more dominant than all of the headlines around S3 buckets. It's like, it reminds me so much of someone leaves an S3 bucket open to the internet, and then they find all sorts of unencrypted data in there.

46:17And there was headline after headline. The DNC in 2016 got hacked with an S3 bucket. It was like one after another. And we like grappled with that as an industry, but it didn't slow down the inevitable at all. Like it was, this wave was coming. and that took two years and this just feels to your point like 60 it's like 4x more compressed of a timeline than the previous yeah in the previous version it's kind of a weird analogy no but it's the same and it's like it's just interesting i think it's such a compelling technology that everyone's like diving into the deep end and understanding it and almost more receptive to its imperfections than i would have imagined yeah because i think about it like computers since you and I started were really fast rules engines basically with storage attached, you know, and if you look at the entire software engineering industry, it's how many nines of reliability can you get and the software development lifecycle and the testing methodology, it was designed to make things more reliable, more predictable.

47:21And there's, I mean, the idea that the same input could produce different outputs and that's the way the system was designed. is like so foreign to those of us who grew up in this industry. And we've now absorbed these new technologies so quickly. And I think it's fun. I think it's interesting. As you said, I think people have really understood the imperfections. And I think it's important that ChatTPT exists so that like everyone in the world has had it say something silly and obviously wrong. It'd be like, huh. So it's like super intelligent in some ways and totally idiotic in others. Totally.

47:56And that's an important intuition for all of us to have when you're using this technology. Yeah. It's also why things are so competitive in the market right now, because it's like, you know, if you went off on an island and spent a week in one of these tools, very quickly, I think you'll come to the same conclusion that we come to, which is somehow this might be still underhyped and underrated after you spend literally a week in it. I agree. I mean, it's to me, I might've borrowed this from, from a talk, but I can't remember which one, but you know, we're basically going from making something scarce and making an abundant.

48:38We did it with power. Uh, we did it with transportation. We did it with, uh, food, at least in most parts of the world. Um, and now we're doing it with intelligence. And I think that if you look at the impact of like electrifying, uh, the world, you look at the impact of, making fruit production the way it is now. What percentage of society worked in agriculture 300 years ago versus today or 200 years ago versus today? And then I just think we haven't actually absorbed how impactful that will be. And I think it'll be largely great. There's definitely going to be some, as with all technologies, downsides.

49:20But how amazing is that? And it's hard for me to imagine a world where I can't you know, go to the supermarket and get the food I want. It was just like, it's hard to imagine to go into a room and not be able to just flip on a light switch to have the lights go on. I think, you know, tens of years from now, it will be hard for people younger than us to imagine a world where intelligence is not abundant. And I think that's like a really kind of mind-blowing thing to think about. Can you tell me about the three martini lunch experiment that you did at Friend Feed? How did you learn about this? I went deep in the archives.

49:58So, oh my God, Paul Vuhite can maybe correct me. But so we were 12 people at Friend Feed, so very tight knit. And we ate lunch together every day. And that was right when Mad Men was really popular. And if you remember that show they just drank like fishes so like it was just like and so we you couldn't help but talk about it just because they're in the office and yeah and we got in a debate we went to there's this Mexican restaurant kind of near we had rented this like kind of cavernous warehouse uh in Mountain View near the Caltrain station and went to our favorite Mexican restaurant we had this like relatively heated debate they're like there's no way like you Paul could like drink three martinis and do anything after lunch.

50:43And like, I can't remember whether it was Casey or Anna or me, but like, we were like, Oh, and I bet I could. So there's only one way to settle that bet. So we went first, uh, in this day and age, it's hard to find a place that serves martinis at lunch. We found a steakhouse in Mountain View. We went, I don't think most people made it through more than two martinis. I'm proud to say I did all three. And then we went back and tried to see who could commit code, which is not, I wouldn't endorse this for future generations. I did, I did get a commit in. I was sleeping on the couch. Oh my gosh. I think Casey won the bet though, but I can't remember the conditions of the challenge.

51:24So I can't believe you knew about this. I think a lot more people have known now, but that was the reason. And it was just a stupid startup bet where we just thought, you know, how productive can you be? I don't remember the commit it was. That is so funny. And how long, like, I guess, do you think of that company as a success or not? Not really, no. But it's interesting because I do think it is a capsule of what makes Silicon Valley so great. I will meet people to this day who will talk about how much they love Front Feed. They'll talk about what we did. I'd say in addition to the like button, which we were at least sort of credited with in Vendee, though a lot of, candidly, a lot of social services had similar affordances, but because we did it in the feed and we're sort of a Silicon Valley darling at the time, we get a lot of credit for that.

52:15We also were the first real-time social networks that meant that you didn't have to refresh the page. So as people commented it was streaming and that was really impactful to the to the industry and to this day people will talk about those things and say god you guys were so innovative like congratulations yeah facebook adopted when they bought you they adopted the like button is that right a little bit even before you know like there was well everyone was looking at every like from tumblr to us right right looking at everybody else so some before some after um what was neat about it is like i I always use this analogy, but if you started a hedge fund in New York and it was a financial failure, but you had some really good ideas.

52:56My guess is you don't go to like a cocktail party in Manhattan and they're like, you had some really good ideas. Congratulations on your financially unsuccessful hedge fund. But I don't encounter anyone who's kind of in Silicon Valley who doesn't view FriendFeed as at least partial success because it was a very innovative product and company. we got acquired and, you know, the employees, we all did reasonably well as a consequence of that acquisition. Very grateful to Mark and Facebook for that. And so it kind of turned out, well, like we innovated, we moved the proverbial, you know, we innovated and we changed the way people conceived of social software.

53:33It was a good financial outcome for everyone. So by Silicon Valley standards, it was a success. Even though we were basically popular in Turkey, Italy, in Iran. We had$0 of revenue and 12 employees. And, but that is something that's some good stories and some great stories of a three martini lunch. And, uh, and, uh, I think that's great that even though I don't think objectively it was a successful company by any meaningful measure in Silicon Valley, that is a form of success. And I think it's the reason why folks like me started a second company because we innovated, we clearly had a good product culture.

54:08We clearly had great ideas. And that meant when I went into Peter Fenton, a benchmark the next time, he's like, yeah, I'm in. It was not even a question. And I think that's great because I don't think people always ask, how do you reproduce Silicon Valley and place X? I'll say, what's the next Silicon Valley? Access to capital is important. Access to talent is important. I do think the way Silicon Valley conceives of what is success and failure is very nuanced and very different. And I think the second, not just second chances, but actually just having more nuance with how we think about failure is so uniquely Silicon Valley.

54:48And so it gives, I think it like affords, it makes us take bigger risks in a really positive way. And I really love that part of our culture. I totally agree. When you joined Facebook, did people make money because of the appreciation of the Facebook stock? Both. And it was a decent outcome. It was a decent outcome and certainly Facebook's success afterwards. This was like early 2000s. Yeah. And in fact, it was sort of during the financial crisis. It was a particularly good time to get, you know, Facebook equity. Because it was depressed. Yeah, it was depressed. That makes sense. And so part of it was just like, you know, for 12 people and who didn't have that much money, it was great.

55:27Totally. How old were you? I was 20, 28. How long did it take before you became the CTO of Facebook? I think like a few weeks. A few weeks? Yeah. And how big was the company at that point? I don't know, maybe a thousand? Yeah. A thousand people. Yeah. And you're the CEO of a 12 person, like, by your - startup and then you join Facebook and then within a few weeks, Mark makes you the CTO. First of all, did you like, I don't even know how to approach this line of questioning. Number one, how could that happen? Like how did the, the, how did the, like, uh, how did the, the antibodies of change not swallow that decision up?

56:25So the preceding part of my story was I had been at Google for a little under five years, so four and a half, five years. And I was part of the team that made Google Maps. Can I interrupt you here? Because I think you're being modest. You rewrote Google Maps in a weekend because you thought it sucked. That is a mostly true story. I'm happy to go through that story. So I had experience with a really well-respected technology company at a larger scale and had helped create some pretty meaningful, meaningful products there. So FriendFeed was my first, you know, time in the proverbial arena starting a company.

57:05And so when I got to Facebook, it wasn't just my FriendFeed experience, I think it was like, you know, on my resume, it was also the Google experience. And also, I feel actually really blessed to have ended up there because there's a few idiosyncrasies that were specific to Facebook that I, I think with the benefit of hindsight, I really appreciate. One was it was a very young company. Facebook had just transitioned to let, I think, anyone sign up right around the time that we joined, not just college students. And as a consequence of that, though, most of the people who worked there had used Facebook at college.

57:43And it was relatively hard to convince someone who had never used Facebook before to work there. It's a consumer service. Yeah. People are useless at enterprise software companies. But, like, you know, it's the thing for college kids. Like people didn't, although I think Snap probably went through this too. I imagine I haven't talked to David about it, but I imagine it's similar. You know, when you have something that is like, you feel like it was a, it's a different generation. So despite being, you know, 28, 29, you know, when I became a CTO, I wasn't exactly like inexperienced relatively speaking.

58:16Right. In fact, I was often one of the more tenured people in the room. So that was just a quirk of Facebook that I think ended up helping. And then, you know, FriendFeed, despite not being a successful business by any objective measure, we were thinking about the same problems. And I think Mark is a famously first principles thinker, and I think we respected each other intellectually. And so, I, you know, I built teams at Google and built a startup, and we really were compatible intellectually. So, uh, and then, and the other thing that I, um, really tried to do is like be embraced being a part of Facebook.

58:54Like I, I've been acquired and acquired a lot of companies and that identity shift is really hard for people to make. And I really quickly, it was like, I was grateful to be there. And I like sort of took off my friend feed hat virtual in front of my Facebook one. And I think that probably helped with the transition as well. Like I was, uh, very much a part of the team there. What do you mean that identity shift is hard? I think it's having acquired a lot of companies over my time at Google, Facebook, and Salesforce. One of the hardest things is like you go from being the founder of your company to if you get acquired being an employee.

59:33And I think you see it play out in a lot of different ways. Like we want independence. We want to operate this way. We want to keep our brand. And we want our own hiring standards. We want this. And any one of those things might be fine. The sum of all of them usually is a emotional response to just losing the identity that you have from it being your company. And in particular, the uncomfortable thing that people always talk about is most people don't choose to be acquired. The founders often do. The people who made the decision, but most of the employees made a decision to work for company A, not company B.

1:00:12And so, and often, by the way, if it's an acquisition, they're probably in a related market, might even be slightly competitive. It depends on sort of market. So you not only have employees who didn't choose to work for the acquiring company, but often maybe even made the decision not to work at that company to work at the startup. And so you end up where like you might have a great high functioning team with a really unique technology and all these things. But so much of that success isn't just like some, you know, PowerPoint presentation on the synergies, but actually like, how do you take this company and make it a team?

1:00:48And how do you actually take the individuals there and have them identify in a way that isn't mercenary, but actually, you know, sort of, you know, missionary on the like, I'm a part of this company and happy to be there. and most don't make the transition. I mean, most founders stay at companies for relatively modest amounts of time. That can be okay, by the way. It's sometimes healthy, actually, if you can't make that transition. But, you know, when you ask sort of like the antibodies and all that, I think if I had shown up resistant to the Facebook culture, I think it would have been a different story and I tried not to be.

1:01:24Just recognizing that I respected the company a lot, you know, like it was the, not nearly as big as it is today, but we thought it was like the best company in the space. And I wanted to learn from the culture and the company. And I think that authentic embracing of the culture probably helped. How long were you there for? Not long, three and a half years. That's a decent amount of time. You don't think that's very long? The shortest I've ever been anywhere. Post-acquisition, you're saying? Or just in general? Just in general, yeah. Three and a half years. Yeah. And by the time you - Three years, yeah.

1:01:58But not less than three, but I don't remember the exact amount. And you left to go start another company. that's right and b2b company yeah nonetheless yeah which you've never done before yeah it's a first time for everything and uh and maybe just as an aside you are one of the least known most impressive people in the valley like it's unbelievable meaning like your ratio of accomplishment to well-knownness is insane it's just i i really respect your versatility it's amazing. I appreciate that. Um, and, um, maybe just to put a finer point on that, like it's, it's almost polymathic the way that you can oscillate from CEO of Salesforce, CTO of Facebook to 12 person, three martini launch, B2C company to B2B company.

1:02:49Um, and it's just amazing. It's amazing. I appreciate it. Um, uh, when you go, first of all, did was it just the itch the siren the siren songs of startups that called you out of facebook because you know like being the cto of facebook not meta yet like in 2012 13 was like this like i mean in hindsight like it's like the best job ever i've left a lot of jobs that other people And like probably the most lucrative. Yeah, that's probably right. Yeah, so it was a number of factors. So I'll say the personal part of it. I had a friend, Kevin Gibbs, who had started Google App Engine, which is what became GCP.

1:03:42When I at the tail end of my tenure at Google, Google Maps had an API that was quite popular. It was probably the first, actually, I think it was the first meaningful developer product that Google had that anyone used other than the ads API, which wasn't really, it was used to automate ad stuff, but it wasn't like a developer community, if that makes any sense. But the Google Maps API, everyone from, you know, Yelp to Trulia, everyone was built on it. So every time there was an engineering team who had like a 20 % project related to developer, they would be sent my way. and Kevin had made this prototype of basically letting developers host applications on Google's infrastructure, which I think was called Prometheus internally, if I'm remembering correctly, from a long time ago.

1:04:32I ended up leaving right before that launch, but I was really excited about the project and Kevin and I developed a ton of respect and affection for each other. He used friend feed, he'd send me notes, and then he had reached out to me uh probably the timing wrong call it two and a half years into my facebook tenure he's like i really want to start a company with you and i was like i really want to start a company with you and i was like but i just have more to do here you know i felt a sense of obligation to to mark zuckerberg and facebook you know and and cared about the place too you know you sort of start to develop a little loyalty and then uh he said okay i'll like i can't wait forever but i'll wait a while.

1:05:13And then Facebook went public and a lot of things. And I was like, sort of felt like it was like an okay time to leave. Like I wouldn't have a sense of not doing right by this team that I cared about. And then similarly, we had gone through this really awkward transition to mobile at Facebook. Now I think Facebook's viewed as something really benefiting from mobility. There's a point when I was CTO where our iOS app had a one and a half star rating in the app store. we had made some bad I'll take full account with some bad technology decisions early on we had when Blackberry was popular and the Trio was popular and all these legacy platforms we had this really like right once run anywhere kind of architecture that was made everything work on these devices but it was just like shitty and it was sort of like and then relatively quickly it's almost like the AR market like things moved really fast there it went from oh there's a billion mobile platforms to clearly just Android and iOS are going to end.

1:06:12And so we ended up like rewriting our apps also during my tenure. So I could at least take credit for fixing the shit I played a part in and making, um, and, and made a really high quality experience. Part of that, um, Mark and, and the whole exec team, Shrap, Chris Cox, Cheryl, we, we basically were dogfooting our own products. We used Facebook messenger, like exclusively to communicate. And we, uh, posted in Facebook groups instead of using mailing lists, mainly just so that we were experiencing the pain of using our mobile products like every day. That's the proverbial dog fitting. Like the contours of the technology.

1:06:47Exactly. Exactly. And it was really effective. And we got out of that. We also did reform our business model too with mobile ads. And it was a really like a growth moment for me. And it's nice to like have some mistakes that you like, you know, learn from and not nice, but you know, It's like you learn a lot from those mistakes. So then you come out of that, and I've got this engineer in Kevin who I think is one of the best people I've ever worked with. I was like, the idea of him starting a company that I wasn't a part of just tore me up inside. And then I saw the impact. I just want to work with him.

1:07:22It's like these people that are like your creative foils. Yeah, I have a list of Avengers in my mind that I kind of catalog of like, these are the Avengers. And then the other thing is I'd seen the impact mobile had in our business and a little bit like we're talking about large language models now. I was just thinking about the impact it was going to have on all these other markets. And in particular, because of that dogfooding experience that I had, I was like, wow, literally 0 % of the enterprise software products we're using even work on my phone. um this is and i was like this is kind of interesting as like what will is there an opportunity here to build something for teams and for companies that you know is there an opportunity to like unwedge the the offices of the world and the others that are built on like you know the metaphors of the desktop and do something native here uh probably like quip was actually a reasonably successful product in that space we did well it was a it was a sustainable company unlike FriendFeed, you know, had a lot of customers' revenue.

1:08:20Slack was obviously a much stronger company in the space. But I think a lot of, I don't know if you think of Slack as like a mobile company, but I do think the emergence of like messaging other than email would not have happened if not for push notifications in the mobile phones. I think there was a bunch of us sort of circling around that concept and a number of kind of meaningful companies that came out of space. But I got to start a company with Kevin, which was fun. And sold it for like almost$600 million to - $750. $750. Yeah. To Benioff at Salesforce. Yeah. Him individually. Yeah, you sold a straight-to-bar.

1:08:56Okay, but the stories tell me, and again, whatever I read in the headlines, I'm sure is definitely true, but the stories as I read them were similar to you having your eye on working with Kevin someday. Mark had his eye on working with you someday. You can interview him and ask him that question. I certainly want to work with Mark. I think he shared the affection, but it's a big acquisition. There's a lot that goes in. It's not just that. And how long from acquisition to up the elevator to the top floor of you being the, I guess it was like chief product officer, right? So I operated the Quip business unit for a year.

1:09:42It was really fun. We had a great time. We learned so much. You know, you joked I hadn't done B2B software. Well, you know, we sought advisors. We talked to people like you. We were like, how should we structure our sales team? And then you go into Salesforce, which is, I think, unambiguously like the best go-to-market team in enterprise software. And you're like, oh, man, we are idiots. You know, it's like I don't describe it. It's like you think you're a good painter and then you see DaVinci and you're like, oh. So we were just like that first year, I was like a sponge, just like, you know, rethinking, Kevin and I just rethinking like how we did everything as it relates to bringing our product to market.

1:10:19And it was fun to just, you know, really grow the business actually. And it made, I think, the whole team who worked on Quip just happy that we did the acquisition. It really like put on these like rocket boosters behind our business. And I think Salesforce has done that incredibly with a lot of acquisitions, actually, just like MuleSoft turned into an amazing business after acquisition. So, like, the deal thesis worked, you know, it was really working. And then after a year, Mark asked if I was willing to be chief product officer. The chief product officer, instead of the company, actually just wanted to transition into kind of like a different, more strategic role, less, you know, operating as, you know, like operating these huge teams.

1:10:57It's like a, it's a job, you know. And so I think he wanted to take on a different role. It's definitely a job. And it wasn't the plan at all, but I really was grateful to be a part of Salesforce. I had a really good experience that first year. And I really felt an obligation to Mark. So I was just like, I'm gonna say yes to anything he asked me to do because I wouldn't be here if not for him. And it was very personal for me in that respect. I was like, you say jump, I'll ask, say hi, hi. like him and I was really learning a lot so was chief product officer for a few years I think I can't remember the exact number and then became chief operating officer for a couple years and then then co-CEO okay was the you don't have to answer this but like was the plan to be the CEO not my plan certainly I don't want to speak for Mark like he might have had some plan in his head but but I don't I don't know um I actually not planned on staying there as long as I did not not How long was it?

1:11:55Net? Seven years? Seven years or so. Yeah. Not because I was like, didn't want to be there. I just didn't, I didn't think I would find such a great career there. And I just ended up saying yes to the opportunities that Mark gave me and then feeling a lot of loyalty to that place. And so, and then the pandemic happened right after I became chief operating officer, right after the previous co-CEO left. So I ended up sort of like thrust into the, probably a different flavor of the role for anyone, you know, Salesforce, when I left had about 80 ,000 employees. And so like operating in a company in a global pandemic, that's got, you know, 70 ,000, 60 ,000 employees is like, it's a, it's a complex job to say the least.

1:12:41And so I ended up just in a different, you know, like a lot, a lot of things happened, like a global pandemic that no one predicted. So, So, and then, you know, just it's, so I don't think any of it was like the plan, but you just have to sort of roll the punches. Okay, during all this time, you obviously make fantastic wealth, like generational money. You had mentioned earlier that when you were thrust into the open AI mediation, your wife was the one that actually ultimately encouraged you to go do it, maybe. As you describe your journey, you talk about your plans. It almost seems like you never, it never goes according to plan, but you clearly plan these things out.

1:13:24Like you think about this very intentionally. So I'm curious, like today sitting here, are you thinking about, what are you thinking about intentionally around that plan? Well, I mean, I'll talk first here. Clay, my co-friend and I, like we want to, our plan is to make Sierra an enduring public company and our goal is to have it outlive us. and which I think is hard to do in the tech industry. I mean, just quick aside, Facebook's first campus was Sun Microsystems campus, which had been after sort of a long decline acquired by Oracle. Google's first campus was Silicon Graphics, SGI's campus, which once was a great company and then was essentially renting out its campus because it had declined.

1:14:11There are companies like Microsoft that have gone through multiple technology phases and remain strong. IBM actually is, you know, maybe an example of that too. I think ours did a great job there. It is hard. It's more likely that you actually, your company was created for one year of computing and then like the world moves on. And so that's our goal and our ambition. There's two parts that I was sent to your questions, like how much, you know, you obviously don't, you're not too rigid around those plans. I think that's true. I actually, most of the unhappiest people I know are rigidly following a plan and not observant of their own happiness or observant of the opportunities around them.

1:14:53And I actually think that a big part of life is recognizing when there's a unique opportunity that you didn't plan for and asking yourself the question is like, should I change my plans? Whether that's in your personal life or your professional life. I think it's the quote I always I heard was Eric Schmidt to Sheryl Sandberg. If someone offers you a seat on a rocket ship, don't ask what seat, you know, and I, I like that philosophy of life. Like, you know, I do think especially in Silicon Valley, there's just unique moments and you just have to be self-aware and aware of the market. And then the other thing is like being philanthropic, I really care about philanthropy, but I love to work.

1:15:34There's this Alan Kay quote, the best way to predict the future is to invent it. And that is like my operating principle. And I want to impact the future and I want to help invent it. And I think the idea of sitting on the sidelines and drinking a Mai Tai on a beach doesn't give me joy at all. I want to build. And so after I left Salesforce, by coincidence, since ChatGPT came out right then. And Clay ended up, we ended up having lunch. That's how we decided to start a company together. But I would be building open source software of not running a company right now because I just want to work in the technology and help shape it because it's the most exciting technology of my memory.

1:16:19And I want to play a part in shaping how we all use it. Maybe more of a dumb question, but it seems to me that when people reach your level of success, whatever that is, like they tend to their next company tends to be one that's like a higher calling, right? Like a pursuit that's deeply meaningful to them, right? Let's just say like Brian Armstrong doing New Limit, which is like longevity, right? Or obviously Elon and all the things that he's doing. It's like examples of these things. for you, like, it's like, I don't know, like Quip or Sierra, like, it's not the thing that you probably are like the most personally passionate about ever.

1:17:07How do you think about like, I don't know, there seems to be this trope in Silicon Valley that is work on the thing that you're obsessed with, that you wake up in the morning. I'm just curious, like, what your reaction is to that? Well, essentially, because I am extremely passionate about Sierra. So, you know, I, let me just start with like, I think once you've had success in your career, I think there's some people like Elon, who famously like invested all of his money in his next thing. How amazing is that, right? That take the permission of having, you know, you money and, and, and, and like take more risk.

1:17:52And then there's some who, you know, I think, uh, once you've been successful, it's not particularly fun to fail in public. And so there's another aspect to, I think a lot of folks who come into wealth where you could argue it's a passion project, but I think it's also, I've noticed a lot of people I know, and I won't mention their names, but like they'll, their next thing is decidedly like lower risk. Um, Because they have like more reputation to lose. Yeah. Yeah. Trust me. Like everyone, every entrepreneur knows the feeling of failure because whether even if your thing worked out in the end, you know, when Chaski was making Obama O's at the, you know, DNC, you know, and literally sleeping on an airbed somewhere, you know, it's like, it's fun to talk about now because it worked out.

1:18:41Like I bet it wasn't that fun at the time. And then, you know, and then you have, you see it all the time with the second or third time entrepreneurs, like on social media, people sort of root for their demise. I don't know what it is. Maybe it's jealousy or whatever it is. I always root for entrepreneurs, like even our fiercest competitors at Sierra, like, like, I'm like, it's hard for me to root against an entrepreneur. Like I just like, I, I, it is, I feel a connection to, you know, everyone who's taken the risk to start a company. And I do think that the target on your back is bigger, the expectations are higher.

1:19:18You can raise more money, which raises the stakes of any modest failure, you know, even higher. So I think it's really hard. I think people don't really talk about that very much. And I think, you know, I'm proud to put myself out there and like, yeah, this like, I don't think Sierra will fail at this point. We have a pretty enduring business. But, you know, when we started, it was just, you know, two guys on a whiteboard as any other business starts. And I think that's really hard to do. On the passion side of it, I think that there's the consumer side of AI, there's the AGI, also the research side of AI.

1:19:50I'm really excited. I grew a deep affection for how can you make this next generation technology not just the domain of tech companies. And I actually think that if you look at empowering products that actually like Shopify is to small retailers. Like, you know, now a mom and pop shop can have just as sophisticated of a commerce platform as a big company. Those tool making companies, I grew a lot of affection for. So what I think of Sierra, like why it's so exciting for me is we're helping every company in the world, not just like the tech ones, make their AI agent, like the one that they want.

1:20:29And actually, how do you empower them to do that when they don't have the same sophistication and AI that, you know, say some company here in San Francisco has, that's an incredible challenge. And that's why I bring up things like Shopify. I really admire that company because they serve entrepreneurs. Aren't you on the board? I was on the board. I'm on the board. I think it's stuff like that. So actually I'm super passionate about it and I actually love making it. I love making tools for people and for businesses. And I think Quip was a form of a tool and as is a Cera platform. It's incredibly rewarding.

1:21:02Like seeing, you know, like there's a SiriusXM is one of our customers and there's a team there that maintains Harmony, their AI agent, and they have the title AI architect. And like, I don't know if they would have had that title before us. And now we've empowered them with this incredible new technology. How cool is that? It's really satisfying. Can I revisit briefly the point about like the reputation and like, let me pull that all the way through it for you. Like, you know, a bunch of people as do I, that did something amazing and they will make up excuses why they're not doing something equally huge again.

1:21:39But I think we both know that deep down the real insecurity comes from burning down all of the currency of reputation that you've built. Do you think it's, why do you think that people are scared about losing that? What are you losing in your mind? Like, as you think about, all right, I go to Sierra and it doesn't work. Let's just say. Yeah. Like what, then what? Do you know what I mean? Like what happens? I think we mythologize entrepreneurs. And so, you know, if you've started a successful business, There's this sort of idea of it was your magic touch, your great decision making. And then when you have multi-time entrepreneurs who's done multiple things, they're like, you know, godlike or whatever.

1:22:28And it's not bullshit. Obviously, there's literally to have started two successful companies is so remarkable. Statistically, you deserve the praise. but I do I cringe a little bit at it because I'm like I don't want to dissuade that person from doing a third thing because then all of a sudden you're not batting a thousand anymore was that really you or was it the market conditions or da da da and actually I do think no no I get it it's just humanity it's like the schoolyard but like social media and the mainstream media will go around and collectively like judge, you know, judge these things.

1:23:06And I, um, I, part of the reason I just like cringe at it is like, I, I admire people who take risks and I admire people who start companies and I admire people who, uh, are willing to, uh, put themselves out there to fail. And quite literally it shouldn't matter. You know, you still have your wealth, you still have your family stuff all the other things but 100 you care you know like no one has so little vanity that they don't care when an article comes out saying you know they failed or and and there's a certain amount of person you are a failure you know it starts to become your personality so i get it i it's really hard i thought about it when we started when i started quip let alone sierra and um you have to overcome uh the safety net of you know like before i started sierra that map story existed and you know quip was a high priced acquisition and and then you go and like you know this is the last you're irrelevant again i could be irrelevant again and if it fails that's probably like the the first page of the the book you know yeah sort of thing and then you just got to be like it like this is like this is i want to start a company because it's how you have an impact on the world and i want to do it um but it's really hard to do and i actually i'm like the enabler though.

1:24:24Like I like if folks, I always am like the one, like you've got, you've got this vision for the future, go do it and do it the right way. Like take that risk, build that team. Um, there's a lot of subtle things too, though. Like it's, I think having experience helps in enterprise software a lot. Uh, just, you know, the actually building the machine of a go-to-market engine is something that is, uh, repeatable, uh, across companies. and similarly, having an intricate understanding of how procurement works at companies and how software is bought and sold and the enterprise software landscape. And so I think it's why if you look at Workday, started by, that was Anil's second company and remarkably successful.

1:25:12Both were really successful. And I believe ServiceNow was too. I don't know the whole history of it. So if you look at all these companies, you know, it's different parts of the market, you know, being a repeat entrepreneur has some, like your experience has a lot more value. Consumer nowadays is quite hit driven. I mean, there's some innate sort of viral growth techniques and things like that. But I think it's harder to intentionally just repeat there by like methodologically, you know, kind of replicating, you know, your lessons from the past. Is there anything that has most surprised you that you feel like is not what you expected it to be or let you down having reached the success that you have?

1:25:59Whether it's the street cred that you thought you would have and what that means for your identity and ego, whether that's the money, whether that's the technologist that you thought you would be that you ended up not becoming, whether that's the houses you get to own everywhere. is there something that felt like, is there something that you imagined you daydreamed about, like when you were younger, that now being there, that's absolutely not what I expected? No, I mean, I, my life has exceeded my expectations and I'm very grateful for it. And so, and I have a very open mind. So I probably have been surprised by things, but I'll try to recalibrate.

1:26:41I try not to have a sort of an ossified view of what life or the future will be like. And I'm a very curious person. So I, I, I'm certain I've been surprised at things, but my reaction is not, Oh my gosh, it's like, tell me more, you know, it's, uh, I do think it is something that is, I see him like people like Patrick Coulson, who's got to be like the most curious person I've ever met. I try to like, I might be one 10th of him, but I'm, I have that personality. I, surprise is like, uh, not quite the emotions, like curiosity. And I just, I'm constantly like recalibrating my priors. Maybe one more question and then I'll get you out of here.

1:27:19Like, uh, you had mentioned this idea of sitting on the beach would kind of drive you crazy when you, uh, when you're away from work, does it drive you crazy? Yeah. I'm not driving me crazy. Like if I'm doing something with my family, it's amazing. But if I just have a free weekend. I'll work. And so, and I'll, uh, you know, or I'll, uh, write code or develop a project, uh, or I'll cook a meal. I like, I like building things and doing things. So it's a blessing and a curse. When you cook a meal, do you listen to anything? No. Just quiet. Yeah. I'll chat with my kids or something, but yeah. Yeah.

1:27:55I, I, for me, I like making software for similar reasons, like making meals. I like making things for people and having them enjoy it. And so for me, uh, cooking a meal and eating meals, like a social process, you know, you do it for the people you love and you, you do it as a collective. And, uh, I just love the whole process. Uh, I do too. It's also like active, you know, you're using your hands, you're thinking about what you have to do. Even the grocery store is like part of the process for me, go to the grocery store, get some food. And then, you know, I'm always like, I always do the like, Oh, this one, this one's not as good as usual.

1:28:29You know, like I always Yeah, qualify it. You almost care more than your guests. That's right. I appreciate you doing this. I really do. Thanks for having me. Is CIRA hiring? CIRA's hiring. We're hiring in San Francisco, New York, Atlanta, and London across all departments. So core engineering, agent engineering, go-to-market operations. So all of the above. We're growing like a weed and growing in all markets. Last one. I ask every guest, when you hear the word grit, what do you think of? Like mental resilience. Like that's what I think about. I think the ability to overcome meaningful setbacks.

1:29:11And so when I hear the word, I was like relive my own. But I love the word. Thank you. I can't wait to do this. Maybe not 10 years, like maybe every few years. Yeah, so that's great. You can revisit your progress. Yeah, sounds good. And see how things are going. I appreciate you. Yeah, thank you. That's it for now. If you liked the episode, please leave us a review or go back into the archives where we've done more than 200 episodes with some fantastic folks. This podcast is a Kleiner Perkins production, and I'm Juven. Thanks for listening.