Platforms and Power (with Hamilton Helmer and Chenyi Shi)

Hey, Acquired listeners, we have some fun news for you. Following hot on the news that we are doing a freaking arena show. Arena show! At Climate Pledge Arena in Seattle, we are here tonight to announce the very first guest that will be joining us for that show. Ben, who is it? It is fresh off his full-page profile in the New York Times this weekend, Jim Weber, the CEO of Brooks Running. We wanted to have a really fun local Seattle success story, but I think most people have no idea the magnitude of the success story here or Jim's personal story.

It's just fascinating. We want everything we do for this event to literally check all the acquired boxes. This checks all the acquired boxes. Berkshire Hathaway. Berkshire Hathaway. People who don't know, Brooks Running is a very successful, now standalone division of Berkshire Hathaway. Local Seattle story. Amazing, amazing journey. What they went from, under Jim, they went from like, de minimis small number of millions of revenue to over a billion in revenue a year, competing with Nike and Adidas.

They were like accidentally bought by Berkshire as part of a Fruit of the Loom roll-up, and they were sort of this bland nothing brand. And by unbelievable maniacal focus on making fantastic running products, that is how they became the billion-dollar business that they are today. Not to mention Jim battled cancer along the way. Like, so great. And when you say billion, by the way, we should say billion in revenue. This is a profitable, over-a-billion-dollar revenue-growing company.

This isn't a billion-dollar valuation. That's a billion in cash every year. Not cash flow, but revenue. Yes. So we're very excited to have a conversation with Jim at the event. You should totally come join us. It's May 4th. Doors open at 5 with plenty of time for drinks and mingling throughout the event. You can go to acquire.fm slash arena show or click the link in the show notes to RSVP. We've got a few more details in the previous little mini episode that we released to announce the arena show.

All proceeds will go to Climate Pledge Arena's philanthropy, the One Roof Foundation. It's $20 to attend. And we hope to see you there. Acquired.fm slash arena show or click the link in the show notes. We're going to have some more announcements coming over the next couple of weeks. And it's amazing. People are DMing us saying they're flying in from all over. If you live in Seattle, definitely come. If you don't live in Seattle, Alaska Air has great flights to SeaTac.

This is going to be a huge party. We're so excited. Awesome. Listeners, acquired.fm slash arena show. We'll see you there. Who got the truth? Is it you? Is it you? Is it you? Who got the truth now? Is it you? Is it you? Is it you? Sit me down. Say it straight. Another story on the way. Who got the truth? Welcome to this special episode of Acquired, the podcast about great technology companies and the stories and playbooks behind them.

I'm Ben Gilbert, and I'm the co-founder and managing director of Seattle-based Pioneer Square Labs and our venture fund, PSL Ventures. And I'm David Rosenthal, and I'm an angel investor based in San Francisco. And we are your hosts. Well, today, back by extremely popular demand, we have Hamilton Helmer. We email Hamilton like once a quarter or so, David, and ask, hey, have you found an eighth power yet? And he always says no, but he did finally say, hey, my colleague Chen Yeh and I have been developing a new framework for how executives can apply seven powers for platform businesses, which are way more complicated.

So obviously, we jumped at the chance to get to dig into this with Hamilton since we later found out that by platform, he means this very broadly, like any business that serves as an intermediary to make transactions. So it probably applies to the technology business that you're working on right now. Or investing in. I find it particularly interesting reflecting after the interview, because it's a framework still in progress. So you can start to see some of the like really important principles crystallize for Hamilton and Chen Yeh as they sort of talk through it.

We get to read about these things in business books 20 years after they're finalized and gone through tremendous rigor. And it's just really cool to see it in the infant stage. So fun to have Chen Yeh on, too. She is an incredible rising star. For sure. All right, listeners. Now is a great time to talk about a new partner of ours here on Acquired, Lagora, the agentic operating system that is redefining how the world's best legal teams work.

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

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

And they recently launched Lagora Agent, offering greater intelligence and performance. The agent lets lawyers set an objective. Then it can handle the planning and the execution and delivery of the final product. Legal teams get to maintain full control and transparency since they're still involved where judgment is required. And Lagora works where you already work. You can use it within Microsoft Word while redlining or drafting. The early Lagora numbers essentially speak for themselves. When they have a head-to-head pilot with their top competitor, they win 70% of the time.

Lagora now has over 100,000 lawyers on the platform from 1,200 legal teams in 50 countries. And crazily, they went from 1 million to 100 million in ARR in about 18 months. Truly insane numbers. And that is the real test. Plenty of things demo well, but the question is whether a busy associate actually reaches for it during crunch time or whether a partner trusts it before going into a conversation with a major client. If your legal team wants to check it out, whether you're a law firm or you're in-house at a company, you can learn more at lagora.com slash acquired and just tell them that Ben and David sent you.

All right. With that, David, let's dive into our interview with Hamilton and Chen Yi discussing platforms and power. Seven powers was an attempt to take an understanding of strategy and make it generally available as sort of pattern recognition. Chen Yi, I think, uses that term, which I like a lot. For founders and people that are interested in trying to create great strategies and make their companies successful. And the reason it needed to be decentralized is that there are really two major step changes in the value of a company.

The first is product market fit and the second is getting power. And they're quite distinct and involve different things. And Peter Diehl gets at it in his book with his X and Y axis. You create X value, but you only get to keep Y percentage of it and X and Y are independent variables. And the thing is that each of these involves an invention. And so it means there's a creative activity involved. And when you look at something that involves invention and creativity, it means, guess what?

It's the inventors that do it, right? And so the idea behind seven powers and strategy 3.0 is to put into the hands of those who are capable of inventing that, a way of looking at the world that gives them a little bit more acuity about what will work and what won't. And specifically, what will work and what won't on developing power. Not finding product market fit, but the second piece. Not on product market fit. That's right.

It's very much focused on the second thing. And it should be in those people's hands. And so you have to empower the people that actually can do the invention. So what we've found is that platforms, they often involve different types of power. They're very complicated and idiosyncratic. They're complex. You can look at one and it's hard to tell whether they have power or not. And figuring that out is hard. So pulling that apart is a useful exercise.

Chen Yi, could you tell us maybe two things? One, from your perspective for this work that you all have done and are still in the process of doing, how do you define platforms? And then also, why is this group particularly interesting to y'all? Yeah. I'll address the second question first, which is why platforms are so interesting. I guess it's pretty obvious to us all that platforms are creating tremendous value. There's stats out there that says majority of the most valuable companies today would operate on some form of platform in this model.

So for us, it's both important and it's just intellectually so intriguing to go into them. And I guess the way we define platform, we think of it very broadly and high level. We think of it as an intermediary for transactions. And that's it. I know sometimes we see people sort of think of platforms as being bounded to digital technologies. People equate the term platform as digital platforms. For us, that's actually limiting the scope of the topic because platforms really is more than that.

It's a model with very ancient roots. You know, there's a book that Hamilton and I both really enjoyed reading and learned a lot from. It's called The Matchmakers from Evans and Schmollensy. So both are economists who've thought really deeply about platforms and they took their title from or in reference to the Chinese ancient village matchmakers who would keep a knowledge base of single men and women in the village and paired them up for dates. I love that example because these are people who existed 3,000 years ago.

They have no access to modern technology, but they operate as platforms. So this is sort of the scope we're going after. If we come up with a framework to understand platforms, they should work as well for an Uber or Airbnb of today as they should work for the matchmakers 3,000 years ago. And so in that example, is the sort of keeper of all of the men and women to matchmake, is that the sort of intermediary of transactions that you're thinking of?

Whoever keeps that database is indeed a platform. Yes. And if you think in their mind, 3,000 years ago, they'd be thinking, how do I outcompete this other matchmaker in my village? And you'll be thinking about, okay, do I have a bigger database? And do I have better knowledge about these people on my platform? So it's actually not that dissimilar. Yeah. What are the dimensions that are going to make my platform more successful than another? Exactly. There's been a revolution in knowledge about this kind of business model.

And a lot of that is around the first step, which is product market fit. You make better matches, right? If you've got mobile phones, all of a sudden it's possible to do a lot of stuff that you couldn't do back in the day. But that question is different than the question of, okay, you've created all this value. How do you get to keep some of it for yourself? Which is what power is about. What you're essentially teeing up here is, just to make sure I understand, platforms under this definition of an intermediary of transactions encompasses a lot of business models that we would otherwise refer to more narrowly as like marketplaces or platforms like what Windows offered for application developers to build on top of in the 90s that gave it so much power

could encompass aggregators of sorts because they're all, in a way, an intermediary of transactions. And I think what you're saying is there's a whole lot of work that you've done and there is to do out in the world of applying the seven powers framework specifically to platform businesses because they're different than businesses that aren't platforms. Do I have that right? Yes, that's well put. David should be interviewing you. Chen Yi, can you walk us through your work on how to assess platform power?

I guess even before that, we're not trying to understate. I think we just said, you know, platform is not tied to a particular form technology, but there's an important thing to notice that technology is really important. And there's a particular form that gets reflected in the platform, which is technology lowers transaction costs. And it lowers it so radically that what you see is entirely new markets get created that was not existent before. So our way of framing the relationship between platform technology is that platform is not bounded to a form of technology, but technology is the driver for new platforms to emerge.

And that's something I guess all the entrepreneurs and inventors out there would have a very clear sense of what is the technology trend they're riding and creating new markets with their platforms. It's fascinating. It's like definitionally true, be it by reducing distribution costs or by reducing the friction to make a transaction. Every step change in technology dramatically reduces transaction costs. Shoot, I mean, TikTok using AI, that getting cheap enough and good enough to have that drive the algorithm.

So many examples. Yeah, exactly. The frictions or transaction costs have come in very different tastes, right? It could be the reduced of search costs, the reduced cost to input information, reduced cost to deliver things to your customers. And it's only up to the imagination of entrepreneurs to create new ways to make them happen. It's a great point. Even thinking about the database of the notebook, to go back to the matchmaking example, because I think it's wonderfully concrete.

Somebody has in their head the five single people they know that they can recall that might be good to matchmake someone with, you know, the Yenta of the village back in the day. But if you have a big notebook and you can write down everyone, suddenly your cost of storage went down dramatically and your recall got much less expensive and much more scalable. And then, of course, when you fast forward to today and you have Tinder and Hinge and these sorts of apps, it's all infinitely cheap, not only to store, but to input, to distribute, to incentivize people to load their own information to the database.

There's just all these dimensions of cost that sort of collapse to make everything have less friction. Technology is what opens up whole new potential vistas on product market fit in just exactly the way you described, Ben. But if so, you think of two step changes in value, that's the first step change. The second step change is, okay, how much of it do I get to keep? That's the power step change. It makes that in some ways harder.

That's the paradox. That's the paradox. Because all these things that reduce friction are easily available to lots of people. And so, your very ability to spin up quickly and make something happen, and it seems very powerful, may in fact be the very thing that makes it easy for a competitor to catch you. So, that's one of the paradoxes that you have to be really thoughtful about when you look at power and platforms. I love it. Okay.

So, can you walk us through how you've developed a framework for analyzing power of a platform? It's a very complex problem. And the reason for that is, each platform we've tried to do a case study on, the industry economics is different. There are so many idiosyncratic characteristics that impact the equilibrium state. So, for us, instead of trying to give you a very abstract, high-level framework that's not going to relate to the actual situation, it's probably easier to raise a few questions that we think every operator, when they think through, will find some value in it.

So, I'll throw out the three questions, and we can go through them one by one. So, number one, how is economic value created on your platform? And how does that value change as your platform becomes larger or has more participants attending it? The second question is, how does each group of your customer perceive their economic value from your platform? And how does that change as your platform gets larger? And the third question is, what is preventing your competitors from getting to equivalence in that value proposition?

So, that to us is sort of a comprehensive list of questions that you have to think very carefully about. And after that, you may be able to get some good insights about whether your platform may have power or not. So, the first question is, what's going on here economically? Who's gaining? And where's the money, right? And so, I'll give you an example in Uber. So, what's going on is you have two sides, drivers and passengers, and they're trying to match.

And they're highly heterogeneous because each driver-passenger is time and location stamped. And by having more drivers and more passengers, it makes it possible for Uber to develop more efficient route structures. And what more efficient route structures do, essentially, is to minimize driver downtime. It doesn't change how long it takes to do the drive. That's baked in to who it is and where they want to get to. But it does change how much time you have to wait before you get the next ride.

And then you look at both sides of this and say, okay, why does greater density create an opportunity for value here? And the answer is that it's a better fit. You can pick a driver that's nearer a passenger. And this fit notion is something that's the nature of platforms. So, you're looking at both sides must be highly heterogeneous and you're trying to get a better fit. And that's the nature of the economic value. And then the question is, okay, how does that value that you deliver vary as the participants grow?

Because that's the characteristic of platforms is that's often how they're differentiated by different levels of participation on both sides. And what happens there is that as you are more dense in a specific region, so this is a very geographically bounded economic proposition in the Bay Area, for example. As that density increases, you can decrease the amount of wait time for drivers. However, I'd argue all other people at Uber and Lyft that know this much better than we do, I'd argue that's not a linear function, right?

Right, it has diminishing marginal utility. Right, so it's a negative second derivative. And so, what that means is that the curve flattens. And so, now if you ask the question, how does one provider compare to the other, which is the power question, which is the third thing Chen Yu was getting at. And just to set that up for one moment. So, the first two questions are more about understanding the lay of the land. And then the third question of what prevents competitors from getting to equivalents.

That's the power question. Is that fair? So, it's about both how much value you can create and then how much you can capture from it. So, the second question is not worth asking if you're not creating much value in the first place. So, you actually have to go through all of them pretty carefully. Just for my benefit, and I think maybe listeners, because we went through the questions kind of quickly, can you just remind us of the questions again so we know what we're referring to in the first, second, and third?

Yes. Question number one is how is economic value being created on your platform? And how does that value change as your platform starts to get more participants? Number two is how does each group of your customers perceive their economic value from your platform? And how does that change as your platform scales? And lastly, how do you prevent your competitors from getting to equivalents? I see. So, the first one's really about quantifying value creation and obviously understanding how that changes over time.

The second one is about perception of that value creation by different participants. And then the third one's really about value capture and defending the castle over time. And so, Chen, you said that sort of the power question involved more than just the last one. And so, if you remember that power involves two things, a benefit and a barrier, right? So, the first question is the benefit, and the third question is the barrier. Yeah. Well, actually, maybe first, can we finish Uber and talk about that third question for Uber?

Because that sort of feels like, for Uber at least, and Lyft, that's where the rubber hits the road on. How valuable are they? Yeah. I think a very challenging characteristic of platforms overall is you don't own your customers. Your customers choose to patronize their platform, and they can choose to do the same with another platform. So, this is a scenario that we call multi-homing. And essentially, what we're saying is a lot of the differential value a platform can generate is a result of differential scale they have with their participants.

And multi-homing is what arbitrage out all of that differential value if those platforms don't make profits out of their operations. The result of that is you're creating so much value, but all of those gets arbitraged and the customers get it, but not the owner of the platform itself. So, in the case of, you know, right-sharing business particularly, the things you really want to ask yourself is, what is preventing my customers from also accessing the other platform that's competing with me?

And your customer is referring to both your riders and your drivers. So, it doesn't matter if my platform is right now larger if, technically, both groups of my customers can fictionlessly multi-home on the other platform. So, if you always open two apps and look for whichever one that happens to have that rider that's closest to you, and you always have open two apps to see whichever one gives me the closest rider that I can get matched with, then relative scale does not matter because you're contributing to the same pool of density in your local area.

Makes total sense. Deep-lung listeners have acquired will know this hits very close to Ben's heart with your, what are your early startup weekend projects, right? That's true. At one point, I started a startup that actually got killed by Uber not willing to play ball with us and cutting off our API, but it's effectively a meta-search for ride-sharing to be able to find the fastest, closest ride to you, independent of who is actually providing the ride. Wow.

Yeah. It's an undifferentiated experience to hop in one or the other. Yeah, and it's interesting. The whole meta-search idea is a friction reducer. So, if you can get an overlay that compares two, then it's not hard to see what that does in terms of disintermediating power, right? Meta-search as a category enables your customers to arbitrage away your power with less friction. That's right. Think of Snowflake or something, right? Okay. So, ride-sharing is an example platform where you get to the end of the questions and maybe there's still kind of a question mark about the industry and the companies within it.

Could we walk through an example of a company or an industry where you get to the end and you conclude, oh, wow, this company, this platform has a lot of power? Sure. So, David, before we leave ride-share, think of the moving parts here. One of the moving parts is what does that ride efficiency curve look like? So, as you scale, how quickly does it start to slope down? Because that's going to be critical. If it's a straight line and you're two times relative market share, you're home free.

If it tails off pretty quickly at an early stage with two times relative market share, you could be an attractive relative cost position. And once the business is scaled, it's no longer true because you're into the flat part of the curve, right? So, it's the shape of that curve. It's your size relative to somebody else. Let's say you have that advantage. What's to keep it from not being arbitraged out? So, you want to turn to people like Ben and say, don't offer that app, right?

Which they did. Which they did. Yeah. So, you can fully understand their motivation. You're the enemy, right? So, I think what you're saying is take this example of a market where there is diminishing returns to density. If the diminishing return curve is steep, then it's very unlikely that any individual company can develop power. But if it's gradual or linear, then yes, you still can develop power even within a market with diminishing returns to scale. There's this concept that we call heterogeneity of preferences that matters.

So, what it means is returns to scale is always diminishing. But how quickly it diminishes depends on the space you operate in. So, the case of Uber and Lyft, we are, Ben, as you said, the drive is undifferentiated. All I care about is how far away my driver is or my rider is. So, what they're doing is literally putting dots on the 2D geographical map. And the minute you get to enough density, you're good enough. So, we can think about maybe something with high levels of heterogeneity.

Think about YouTube. The content you watch has so many dimensions that you care about. The language, the theme, the music, the production quality. And that is a space where in order to get to a scale where I'm good enough that this additional piece of content is not going to appeal to my users anymore. That's a much higher level of requirement of critical mass or of scale of content. So, this is something that we think operators should also think through.

Like, what is the level of heterogeneity that your transactions have? And that determines how important scale is to you. Right. That makes so much sense that basically YouTube will continue to compound the economic value of their lead. Because even though there's... I'm going to make up some numbers that seem big. But 100 billion hours on YouTube and only 10 billion hours on competing platforms. The thing I want to watch is so unique to me as a person.

How I feel at that moment. If the personality of the creator is interesting to me. That the fact that there's always going to be slightly better content market fit for me on YouTube than that other platform just means that that curve diminishes very, very slowly. Exactly. And there's a favorable byproduct of that. Which is for something like YouTube's content. It's actually really hard to describe what I'm looking for. Right. How can I, in text, search for the exact content I want?

So, YouTube accumulates this unique set of knowledge about both what you like and watch time of others that attest to the quality of the video. So, that makes the search so much easier on YouTube compared to a competitive platform, even if it's a complete copycat to YouTube. So, YouTube is a good example where the value they provide to me is high. I perceive that value to be high. And then when you think about other constituents also.

I mean, the creators perceive the value to be high. Advertisers perceive the value to be high. And then when you get all the way to that third question of how are they better than competitors, there's lots of ways that they're better than competitors. I'm curious, as you think about YouTube as a platform, what are the ways that they have power versus other video platforms? Because of what Chen Yui was talking about in terms of heterogeneity, if edge cases matter, there's a pretty good chance there's an opportunity for power.

And because when you say edge cases matter, what you're really saying is that curve doesn't flatten out very quickly. Right. Right. And so, it says that even the differences in one of the sides scale remain material in terms of the other side's appreciation of the value being provided. When you go on YouTube, your tastes are extremely idiosyncratic. I'm looking for the latest climbing send or something on general relativity or, you know, the weird stuff, right? Or the latest performance car or something.

Yeah, definitely. So, there's a lot of work in economics that states there could be an equilibrium pricing schema where you charge one side and you pay the other side, and that still contributes to a very powerful platform in the long run. So, it's an interesting state about YouTube. It's very complicated, but extremely intriguing, which is they have accumulated so much heterogeneous content that they are able to charge enough mindshare from their users. And they monetize that mindshare with advertising dollars, which they then pay their creators.

So, it's a position of power that creates enough firepower to keep maintaining and enlarging that lead in this particular user-generated video content market. So, that's why you keep seeing YouTube gets larger and larger. And that's because there's one source that provides them with that benefit that they can maintain in the long run. Yeah, it's interesting that there's no real scale that seems to have been reached with YouTube where anything starts getting compressed, any margin starts getting compressed.

There doesn't seem to be any place where a meaningful number of creators are going direct and, like, publishing to their own video platform that's not YouTube. That doesn't seem like YouTube is having to pay out a smaller and smaller percent of its profits to creators. It doesn't seem like advertisers are trying to get more bang for their buck. They're happy with the bang for the buck. The margins are not getting compressed there. It's kind of remarkable that as that business continues to grow and grow and grow, there is not margin compression in any corner of the business.

Yeah, and the cost of multi-homing on the creator side is actually not huge. You take your catalog, you upload to another platform, it shouldn't be theoretically impossible. We will multi-home this episode on both Spotify and YouTube. Exactly. So, the real question really you should focus on is what's binding all the viewers? Why do people keep coming back to YouTube? And a good thought experiment to go through is, let's say there's a copycat version of YouTube stood up tomorrow, which carried over every single video YouTube has today.

Would you move? Would viewers suddenly start to go to the other platform? And this sort of goes back to my earlier point about this is such a heterogeneous space that the cost of search is not immaterial. For me to, you know, over time accumulate the set of influencers I follow and tell YouTube which ones are the content I love to watch, then YouTube knows the watch time, which is a very important input into their algorithm of, you know, what are the high-quality content as proved by all my other users.

So, those information dramatically reduces my search cost that the other platform, even if they have the whole suite of content, cannot really match. So, there's frictions right there on the viewer end that I think is protecting YouTube's business. Jenny and I argue about this. Yes. Do you have a different opinion? For me, the reason I go back to YouTube is an expectation that exactly the content that I want is there, and it's not the efficacy of the search.

And so, that's the dispute. There seems to be some sort of buried thing in my brain that is aware of their network economy power, where at first I thought it was brand, but it's more like I have the assumption that the latest SNL skit is going to be there because everyone uploads their stuff there. Because I'm with you, Hamilton. I have the same – I'm going to go to YouTube every time because I have the absolute most confidence that the thing that I'm looking for is there because of their network economy power.

Oh, I'm so excited here. And when you say network economies, what you're referring to is that a network effect is when somebody uploads something, it makes the whole site more valuable to everybody else for all these reasons. It makes matching more efficient and more likely to result in a good thing happening in terms of value. So, yeah. Okay. I totally disagree. I'm with Chen Yi on this one. There you go. That is perfect. Two for two, right?

David's a real YouTube user. You spend hours a day. I've become a real YouTube user. I think there is a baked-in assumption, Hamilton and Ben, in the way you think about YouTube. And you represent a large class of people, but I think there's also another class of people that I'm guessing Chen Yi and I may be part of. We won't get into age demographics here. The assumption is you both think you are going to YouTube with intent to look for something.

I don't go to YouTube with intent to look for something. I go to YouTube with intent for YouTube to surface for me without me doing anything, things that I will enjoy. Interesting. Different use case. Yeah. I hate to say it, but probably history is on the side of the younger demographic. Although, I guess Ben, I hate to group you with me. That's a dangerous... No, no. I use YouTube the way that one might speculate older generations use YouTube.

And David is a little Gen Z in that respect. There you go, Dave. Now we've classified you appropriately. Oh, I'm so happy. I can retire now. But that's a good point for operators to think through. Because without being YouTube, you won't know exactly how customers split between people who search and people who just look for enjoyment on their platform. And every platform should have a clear sense of how their customers are split into those groups. Yeah.

Both of those segments exist for YouTube. This is what I loved about how specific you were, Chen Yi, in your second question of how do different customer segments perceive value? Because in analyzing the value created by YouTube to me, but also the power that YouTube as a business has for me as a specific customer is totally dependent on my use case of it. Yeah, exactly. And I would like to give some example that's kind of a more stark contrast between different modes of how customers might perceive the value.

So I think we often hear like, oh, more buyers will attract more sellers and more sellers going to attract more buyers. Like this is oftentimes true, but it's so general that it's oftentimes missing the actual granularity that matters to you. So the example I give is think about an Amazon seller and how does he optimize his economic value from a platform? So these sellers typically have a built-out supply chain. They more or less sell around the market price.

So what they're trying to do is I want to sell more units at that single price. And whether or not I'm going to go to a different platform, the question he's going to ask is, how many more units am I going to sell? And is that going to cover my cost of going on to the other platform? Now, this rationale, if you take it onto an eBay seller who's trying to auction his antique watch, is going to be very different.

He has only one watch to sell, but the equation going in his mind is, how can I sell this watch for the highest price possible? And again, this just depends on the better matching thing that we talked about before, because the more participants you have on this platform, you're going to have better matching. It's more likely for you to have a buyer who values this watch much more than others. So the question he's going to ask when thinking about this alternative platform is, what is the likelihood that a buyer with more willingness to pay is going to appear on the other platform and by how much?

And how do I compare that with the cost I have to manage this other platform that I may want to hop on? So you have to understand what's the equation going on in your customer's mind very carefully, because sometimes they could be really different. And that determines the equilibrium of the competition. Yeah, the Amazon seller doesn't care about finding people who will pay a higher price for their product. They want to find as many people as possible who will pay the market price for their product, right?

Exactly. Assuming their margin is roughly comparable and also putting aside the question of, is it strategically a good idea for my business to sell through Amazon versus selling direct? Or I think intentionally not diving into that because it's a whole bag of worms. Right, right. Okay, so one question I want to ask you all, maybe YouTube is the right one to go back to. I think we can all agree that YouTube has a large power opportunity.

When and how is the right time for a company that finds themselves with a large power opportunity as a platform to start capturing that value? Famously, everybody thought that Google was nuts with YouTube because it lost billions of dollars for more than a decade. And in retrospect now, perhaps that was a brilliant strategic decision by Google. How do you all think an operator should think about that? For me, the key thing here is to remember that the product market fit and power questions are different questions.

And one doesn't necessarily answer the other. In fact, often doesn't. And that it may be that when you have a business model that gets you to product market fit, there may be a power opportunity embedded in that and there may not be. And so those are two very different problems. One is the problem of capitalizing on an inherent potential for power. And the other is trying to figure out what are you going to do that will get you power in something that currently doesn't have it.

And the second is a very hard problem, right? It's a second invent that's every bit as hard as the product market fit invent. So think of Steve Jobs trying to figure out where to take Apple when the PC business turned out to not have any power, right? Here's the most brilliant innovator of our generation. And yet he couldn't solve the problem and he ended up losing his job. It's actually worth drilling in on that Steve Jobs comment for a minute.

And then, of course, you like tempt me with the Apple history catnip. I have to jump on it. It's interesting that he tried. The power computing, I don't know if the Mac OS running on other hardware devices was an initiative that started while Steve was still there. But it's interesting how obsessed Steve was with we control the whole stack. You know, we're the software and the hardware and our software only runs on our hardware and our hardware can only run our software.

And with power computing and all these sort of Mac clones that Apple authorized and said, we're going to enable our OS to run on these other PCs. I actually owned one and it was a cheaper Mac. And it did not save the company because they had no ability to take the profit dollars that they earned from that and build something defensible with it. And it's fascinating that Hamilton, exactly to your point, they did try things. And also, there was no power opportunity that they or anyone really found to be a very profitable PC manufacturer.

Yeah, I mean, I think if they hadn't completely flubbed the Mac 3, I think they might have ended up in a very attractive power position because they did own the stack and they did own the operating system. And they didn't yet own the processor, but they could eventually. And if you think of the PC business, the two nodes of power were the OS and the CPU. But you needed to have superior scale and get everybody signed up to have that work.

And the Mac 3 was such a complete flub that it made it possible for the IBM PC to just take over the world, basically. Right. Well, of course, the IBM PC didn't have power either. And so that ended up being a long-term, not that attractive a business. Well, it's funny because the Mac, just the desktop line, has a lot of power today. It is a remarkable amount that people will pay in dollars they wouldn't pay to a different manufacturer with a different operating system for an Apple computer.

And now they bundle in proprietary chips that are the best on the market. So they have dramatically lower cost structure. So it's just margins everywhere. Right. As you might guess, I'm a huge fan of Steve Jobs. And I think his impulse to control the stack was not based on sort of power, but aesthetics almost. He wanted to control the experience and had this sense of the aesthetics of the experience. He was a genius at that kind of stuff, right?

And it could have aligned with power had their execution been better. But it was a failed opportunity, unfortunately. This will really date me. But I remember when IBM entered the PC business, Apple took out a full-page ad in, what was it, the Wall Street Journal or something? Welcome. Welcoming them to the business. And that's sort of a common trap that you sometimes see, which is people that are just amazing at innovation, which I have a huge admiration for.

That is to say, getting product market fit, think that they can just out-invent the competition forever. And that story usually doesn't end well, I don't think. And the things you're in, you have to solve the power equation or else you end up competing in a commodity way on your base business. So obviously, the answer is different for every company, but what are some ways to think about when you can feel comfortable enough in your power position to start dialing up your profitability, which I guess would equate to dialing back your subsidies on the platform?

In the early stage of a platform with product market fit, what that means is there are a lot of people want this stuff, right? And so you scale like crazy and you're rewarded for that. Your B round and your C round, all of a sudden the numbers start to look pretty darn good. But then later on, you face the power question, which is, is that a profitable proposition or have you just acquired a lot of customers?

So I think it's an idiosyncratic tactical question that as a business progresses, you have to make a decision about when you start increasing prices and it's not that you would eliminate the subsidy. And I think like all decisions like that, you have to look at sort of the immediate effects and the long-term effects. But I have to say that in general, and you see this in my book, there's this difference in businesses between the takeoff phase and the later phase.

Takeoff phase when there's enormously rapid growth. And in that phase, the acquisition price of a customer, it's not priced properly. And so it's a good time to get customers. And later on, you sort of can tighten the screws a little bit. Well, the YouTube example is interesting, right? Like in their case, dialing down the subsidy and dialing up profitability means increasing ad loads to users. And for years and years and years, ad loads were very low.

It was comical. Users were getting so much value for very little ads on YouTube. And then in recent years, they've been dialing it up quite a bit. And they don't seem to be bleeding customers in doing so. I'm going to supplement Hamilton with his own book so Hamilton don't feel awkward. So I think there are two parts to this. So number one is, if they're readers of Seven Powers, and you're patient enough to flip to the appendix after each chapter, there is this concept that we call surplus leader margin, which is the maximum price you can charge more than a competitor while still maintaining your competitive position.

So this is essentially what we're talking about here is, how much can I charge while maintaining the leadership I have today? And that number is not dynamic. That number is dependent on the differential scale you have against the other platform. So that's one of the high levels, what's the overall thing you want to achieve? But at the same time, we recognize the difficulty. And power is both market share and differential margin. And Hamilton always says, no, it's an active trade-off between both.

Entrepreneurs do it because when you see such a large greenfield, you can penetrate. You should grab that and sacrifice short-term margin for larger market share. And that's still power because you can realize those profits in the future. So it's hard to tell people the one size fits all, this is the exact point. But understanding what is your surplus leader margin, how much is the maximum you can charge given the best alternative out there, and dial up the tune when is the right time.

All right, listeners. Now is a great time to tell you about a longtime friend of the show, Vanta. AI has scrambled the whole security picture. It used to be that you proved that you were secure once a year on audit or a static PDF. Then everyone would nod and you're done. But in an AI-first world, that doesn't hold up anymore. Yep. Your risk surface changes every week now. A vendor turns on an AI feature or someone writes in a new model without telling IT.

And your posture is different than it was last week, let alone at your last audit. Vanta's own research found that around 70% of companies have this, quote-unquote, shadow AI running with no security review at all. Right. And that's where Vanta comes in. They're the leading agentic trust platform, meaning they've built the thing that closes the gap. And the way that they close that gap is Vanta agent. Think of it as a GRC engineer. That's governance, risk, and compliance.

Except that it's software and it doesn't sleep. It finds the issues, drafts the fixes, and cuts the time that you'd spend on vendor assessments in half. In half! Which is exactly why more than 16,000 companies today run on Vanta. Companies like Ramp, Cursor, and Snowflake all stay audit-ready and catch the risks that crop up between audits across every vendor, every AI tool, the whole environment. And that's the real value. Trust has to be continuous now, which is why Vanta automates your security, your compliance, and the work to earn and prove trust.

We're huge fans of Vanta over here, and literally hundreds of acquired listeners have become Vanta customers at their companies over the years. So you can get $1,000 off Vanta at vanta.com slash acquired. That's V-A-N-T-A dot com slash acquired for $1,000 off. And just tell them that Ben and David sent you. There's something that I've been thinking about that I want to ask you. So we had, months ago, after our TSMC episode, a couple of the investors from NZS Capital on.

And one of the points that they made is, we actually don't like to hold Apple in our portfolio right now because our view is that they're being overly extractive to their customers or over the whole ecosystem. You know, it's the 30% stuff with the App Store. They are realizing their market leadership, and they are squeezing as much as they possibly can. And you contrast that against a TSMC, which does not charge the very most they possibly can to the customers to manufacture their chips.

And it's a very intentional strategy, and they believe that that's sort of a long-term view that they have in order to do that. I'm curious, as investors, how you think about, from the perspective of maximizing enterprise value for a firm, what should a company do? Should they be maximally extractive to their ecosystem, or should they leave some surplus on the table? Yeah, it's a great question. Before commenting specifically on TSMC, one of their primary competitive interfaces, obviously, in terms of fab technology and getting ahead in that is Intel.

And just a caveat about both those businesses are amazing and well-managed and successful. And the fact that TSMC seems to be gaining ground is also a reminder to everybody that power is not forever. Because I use Intel in my book. And that's the way life is. You know, technology is changing, competition changes. And for me, you know, one of the longest-term power things I've ever observed is that of elite universities and being able to maintain, which is ironic that it's not even in the private sector.

So I think on a pricing question like that, pricing may well be tied to a strategic goal, but it's tactically available to anyone. So its justification has to be tied to underlying fundamentals. And essentially, what you're doing is, in this case of TSMC, what you've cited is they're sort of giving up current profits for something in the future, right? And presumably, it's sort of future revenues. So they either get retention through customer loyalty or acquiring new customers.

So that leads you to ask, well, more customers, more revenue in the future give you more differential returns. And that gets you down to ask about scale economies. Do they have it? And I would argue, yes, but it's a pretty unusual type of scale economy. So there are some very strange industry characteristics here. So really large, lumpy capital. I don't know. What's the new fab now? Is it $10 billion or something? Oh, I think it's $10 billion plus, yeah, if I remember right from our episode.

Yeah, huge, huge number. There are kind of quote-unquote predictable material performance advances, i.e. Moore's Law, so that life isn't usually like that. You don't think that, well, I've got a fairly high assurance that if I make the right technology choices that I'll have a 10x improvement in 18 months or something. You know, life isn't usually like that. And so this is very unusual. And it's driven by the correct technology choices. And the other thing is that the tech advantages are driven by upstream suppliers.

Right. ASML and others in this case. And that's a hard choice, actually. I don't know if you go back a ways. There is a long discussion in semiconductor companies of whether to go through the x-ray lithography. And it turned out there's a lot of money invested in thinking about that. It was very contrarian. In the early 2000s, even Intel, I think, funded ASML's development, a big industry consortium, including Intel, of the EUV. And then thought the technology was going to die.

And so I think three or four years in, divested. Right. Whereas, obviously, now that's the dominant way that we have these three and five nanometer processes and TSMCs to benefit. Right. So under those sort of really odd industry conditions, it means that if you're quite comfortable with forward guarantees of customers, you can make a capital commitment, this huge capital commitment. It wouldn't matter much if there weren't those big material performance advances. And it wouldn't matter much if it was not all that lumpy.

You could do it in small increments. And it wouldn't matter if it was relatively small amounts. And it wouldn't matter if it was all vertically integrated. So that allows a scale company to get to the newer technology first. And that has profound implications for sort of continuing that cycle. But it means that they're lower cost per transistor. And they also can do higher performance products, which probably are higher margin. Before TSMC, the companies that crossed that threshold of enough comfort with forward guarantees were the companies that sold the products.

So back in the day when you guys are too young to remember this, but there was a time when IBM controlled the computer industry. There's nothing like that today. People talk about tech dominance. But IBM just was a force that was just so far and above everybody else. And they backward integrated into fabs. Right. And they were the fab leader because they could do this stuff. And then Intel. Right. So we've moved from a vertical to a horizontal organization of this.

So I may be too reductive about this. But to me, their pricing makes perfect sense, even from a purely Machiavellian shareholder value kind of perspective, is that that allows them to get a customer lock-in is too strong a word, but a comfort with future customer. Comfort that NVIDIA is going to stay with them for generations to come and be paying them billions of dollars to allow them to. Right. And remember, for the really tricky upstream suppliers, what's the, is it the Dutch company that does, you guys know who does?

Yeah, ASML. Yeah, ASML. You have to make a long forward commitment. It's not just the amount. And so it allows them to do that. And then, of course, if they had that advantage and didn't make the right choices about technology, and I don't know if that's luck or skill, then it also doesn't work. But they've made the right technology choices, and they have enough guarantee of future business that they can now be the leader in technology.

And this is a business where that performance frontier is moving in sort of a kind of predictable way. And so being a leader in technology means that you are a cost and performance leader in the business. And so it makes a lot of sense to me to kind of give up current profits to guarantee that ability. Does that make sense to you guys? Yeah. I had not previously thought of the notion that, you know, of course I thought, well, maybe it's not actually benevolent that they're not maximally extractive.

But I couldn't quite put my finger on why. Morris seems like a nice guy, but, you know. Yeah. And the why is so interesting that, well, to the extent that they can massively increase the probability that their customers stick around for a while, they can spend this $10 to $50 billion of capex to build these new fabs. They can be one of the very few customers in the world that's guaranteed to, over the next three to five years, get these machines from ASML and other suppliers, like Drumpf, the laser company, etc.

And so because those are scarce resources to be the customers of those companies and because it's so expensive to build these leading edge fabs, to the extent you make the right technology choices, there is this sort of self-fulfilling prophecy of guaranteeing all of those future profits if you have that magic ingredient of being certain that your customers are going to stick around. Yeah, it's going to be very interesting. TSMC is just such an amazing story. So they've been able to take what in all prior generations had been a vertically organized business and made it horizontal.

And the thing that makes horizontal organization work is typically a scale economy so that you pick up more scale. The strange industry characteristics means that predictability of future customers is profoundly important in terms of creating company value. So when we contrast that kind of back to Ben's original question with Apple's situation today, I guess it is quite different. I mean, I think a lot of people feel that Apple's 30% rate that they charge developers to be in the app stores is maximally extractive.

It's maximally extractive. Yes, that's a good way to put it. But they're not in a situation like that. They're a vertically integrated company. They control the whole stack and they're able to fund their also quite large capital expenditures, but they fund that through hardware sales in a very different manner than TSMC. Yeah, it's not clear to me that sort of giving that money back to customers or suppliers would benefit them a lot. I have an iPhone and I really like it.

I realize that every time I turn around, it's trying to get me to buy, you know, the iCloud or something, you know, and they're trying to take advantage of me. But I really love the product. And there are, of course, very high switching costs. I was on a plane and unfortunately ruined my iPhone and landed in Hawaii and had to get another one. And it was a very easy choice to get another iPhone as opposed to another product.

So I don't think people are terribly dissatisfied with the situation. And so they're not risking a lot. So it doesn't seem insensible to me. I mean, there are regulatory issues they're going to face and they'll be constrained in certain ways and all that. But it makes sense to me. So maybe I'm too cynical in my old age. No, I think it makes sense. They clearly they haven't changed the policy. They continue to succeed enormously. Well, Apple just keeps looking at the batten of each of the parties related to them.

You know, they keep looking at developers and they keep saying, OK, are these developers going to stop making iPhone apps? Like that's where all of the most affluent customers in the world do their computing, like spend their time so that they're probably not. And like, are the consumers going to switch? Well, people kind of only switch in the direction from Android to iPhone. They tend not to switch in the opposite direction based on the last 10 years of data.

And consumers don't feel that pain as much as the developers. Exactly. Most consumers don't know or care about Apple's take rate. Right. They're sitting there going, why would we change what we're doing? Yeah. I mean, I have to confess a bias here. As you well know, economics roots were in the Scottish Enlightenment. And so if you think of Adam Smith's ideas and David Heumann and so on, it's the idea of self-interest. So behavioral takes a little bit of a detour from this, but it basically says if you look at people's self-interest, you can understand a lot.

And so I see both of these as being quite self-interested. And to phrase this in a much more tongue-in-cheek way, it's almost as if these companies that have us over a barrel as consumers are saying, well, look, I mean, we're married at this point. And as long as you're not going to divorce me, I will become a worse and worse partner to you over time. That feels like the concern to me with companies gaining more and more power over consumers in the long run.

If you lead a company that has very high switching costs, and I'd argue for most technical people, non-technical people, an iPhone has very high switching costs. You have to realize there's a conflict there between how much lock-in you have and how you treat your customer and manage it very, very carefully. I often encounter companies where they have a wonderful product market fit, but they don't yet have power. And one of the things is, could you develop switching costs?

And my advice always is, don't think, how am I going to lock in my customers? But actually think of it a different way. What is it that I can do for my customers that create value for them? And then if that proposition has with it a way that they're tied to me, then so be it. But think about the value creation part first, or else you tend to go down these paths of win-lose propositions, which don't end up very well, typically.

What the ride-sharing companies did with their membership programs, would that be a bad example, maybe? That that was trying to create switching costs, but didn't really help anybody? Yeah. The one I always think of here is potentially adding insurance to a transaction. So like Airbnb's million-dollar insurance policy, I won't just wire someone money directly. I will pay Airbnb's ridiculous fees because it's kind of nice to have that insurance policy and them as a trusted facilitator of the transaction.

Ben, I think you're touching on a good point, which is platforms are naturally also competing with their own customers. Because what if your customers go direct to each other? Buyers and sellers just transact off you. So as a platform, you have to first prove that you provide enough value that they should pay you a cut and stay on the platform. Now, that's one front that every platform has to think about, and we call it table stakes.

And then the second issue that the platform has to think about is what you offer is that differential. So the insurance program of Airbnb, could a competitor also offer a similar insurance program and achieve similar value add to your customers? So there's two layers of question that competitively we also have to think about. Right. Yeah. Great observation that the insurance example makes it better than wiring money into a void, but it doesn't directly to the person that's hosting the Airbnb, but not necessarily better than VRBO or someone that could very easily go negotiate a very similar insurance policy.

Why should you be interested in power? That may sound like an odd question for you, but I was doing a class recently with some earlier stage founders talking about power, and I mentioned some companies that I thought probably might well not have it. And yet some of them were public and had really high market caps. And some of the people in the audience response was, throw me in that briar patch, you know? So what I just want to say is that it is important.

We talked about the earlier example of Steve Jobs. And what he found when PCs started to go south was it was not fun, right? A lot of the best and brightest would leave the company. And if you get into the stage where you don't have power, this isn't going to be a fun company to manage, and it won't become iconic. And I think that most of the founders I talk to, they're not in it just for the money.

There's this sense of this is kind of me, and it matters to me the durability of this thing and its success. And so if you want your company to be durable, attractive, a great place to work, a great example for people, something you're proud of, you need to answer that second step change question is where is their power? So I know you guys know that, but I should mention that because I think people wonder why should I bother if I've solved the first problem?

Yeah. I've just climbed this great mountain of achieving product market fit. Isn't that enough? Right, which is an amazing accomplishment, right, when you think about it, and really hard. And isn't that enough? And so I totally get where people are coming from. You both, but especially you over your career, Hamilton, worked with so many founders, managers. Do you have any advice for founders at that point psychologically when they get there? I bet a lot of people are just exhausted, right?

Like they've just climbed this mountain. Isn't that enough? I don't think there's anything really easy that solves things for them. I think as founders, they are used to hard challenges and approaching those positively. And I think that's the best that you can do is just say that you've got another hill to climb. I love that. That's such great framing. We teased earlier that we were going to talk about the difference between network effects and network economies.

And this is something that David and I have flubbed on a few episodes where I think I've conflated them in our power section. And I'm curious, what are some telltale signs of a company that has network effects but did not develop network economies power? So I think network effects in the types of things that we've been talking about in this episode are common. You know, it's when a driver joins Uber, he makes the platform more valuable to passengers because more efficient route structures are now enabled, right?

And that's a network effect. So the things that happen there are somebody joins the network. That's the network part. That's the new driver joining. And something happens to somebody else in it that has a value implication. That's the effect. That's a network effect. So the question is, what would you like to call network economies? That sounds like an odd thing to say, but that's really the question. And you could say anything that their network effects and there's power, you could call that might be one choice.

And Chen, you and I are currently debating this. And so there's another choice, which is the one that I'm currently going down, which is that it's when there's power from direct network effects. And a direct network effect is where your joining has an immediate value impact on somebody that's sort of on the same side. So I joined Facebook because I'm your friend. And those effects are strong because they're additive. So another friend's joined, it doesn't substitute for the one that just joins, it adds to it.

So it's those kind of effects that do more lead to win or take all kinds of situations. So my naming choice right now is if there's power as a result of direct network effects, then that's a network economy. And those aren't very common. I'd say indirect network effects like the one in Uber are much more common. It's just a value impact. That's the benefit side, right? But as you two well know. Is it arbitrageable? You need both the benefit and a barrier side.

So is it arbitrageable? So then you have to get to the much rarer set of cases where there's a benefit. It's material, right? And also it's barrier. And then that's a much, much narrower set of things. So do you want to add something? I'm hesitant to drag David and Ben into our whole debate. But please. I think the short version of things here is network effects describes only the value creation. And it's a statement without consideration about competition, which the latter is all power is about.

So regardless of how we end up defining network economies, it is the type of platform that we believe has power. And that's differentiating from network effects on its own. Gee, I wish I'd said that. You can edit out my comments and put you in. All right, listeners. Now is a great time to thank our longtime friend of the show, ServiceNow. If you are running a large enterprise, AI agents are likely spread across every team. And deploying them is no longer the hard part.

Yeah, the hard part is knowing what permissions they have, what employees are using them for, or what decisions AI is making. AI security for an enterprise at scale is not a small concern. Like, the risks are real. Exactly. And the challenge with AI is governing it, securing it, measuring it, and making sure that it actually delivers value. That is why ServiceNow built the AI control tower. Yep, AI control tower gives enterprises a single place to see, manage, govern, and optimize AI across the entire business.

And it works with any AI, not just theirs. Every device on your network, every permission across every system, every AI agent, visible and secure in one place. And ServiceNow can do this because they've spent more than 20 years building the operational backbone of the enterprise. The workflows, governance, approvals, security controls, and institutional knowledge that power how work actually gets done across IT, HR, customer service, finance, and security. ServiceNow already runs more than 100 billion workflows annually and trillions of transactions for more than 85% of the Fortune 500.

So when companies need a place to govern AI at enterprise scale, they're building on a platform at the center of how their business already operates. And in a future that isn't going to be one AI, it's going to be thousands of AI agents working across every function of the company. But the question is, who's managing them all? So if you're trying to turn AI ambition into real business outcomes and make it work safely, securely, at scale, go check out ServiceNow.com slash acquired and tell them that Ben and David sent you.

So when we were emailing before this, Hamilton, you wrote something that I know is going to be provocative. You wrote, when you see a flywheel, run for the hills. Right, right, right, right. I did say that. The number of decks that I see that have flywheels these days is about 100%. So this is going to be a new way of thinking for folks. What is the concern about flywheels? So flywheels are a sign of product market fit and tell you absolutely nothing about power.

And it's not easy to get a flywheel. And that is how these platforms do find life. And so you get into issues of critical mass and ignition and all that stuff. But that's product market fit. And it doesn't say that if one company has benefited from this, that another, as I mentioned before, often the very technology that enables that happening quickly enables multi-homing as well. And so the benefits are usually from a differential size on one or both sides of the platform.

And so you have to somehow prevent that. And anecdotally, there's one exercise we did, which is really simple, but surprisingly interesting, which is we took a flywheel of a company that's really popular. I won't name which one. But we swapped that logo with their competitor's logo in the middle of that flywheel. And it still works. So that's exactly what we're looking for. That is a brilliant thought exercise. Assuming that they have more scale than their nearest competitor, is there some way that you could draw the XY axis where they're growing in a way that's n cubed?

And their competitor is either further down the n cubed because it's a time series axis, or perhaps they're only growing in a way that's like n squared, so there's no way that they could ever catch up. That seems to be the thing that you're trying to tease out. Yeah. I mean, so typically these things do depend on differential scale on one or more or both of the sides, right? And the way you've asked that question, Ben, you've just sort of assumed the critical thing, which is you've said it can scale three times as fast.

So then the question is, why? Because it's a flywheel. You know, there's a lot of stuff feeding into each other. Because I read the everything store. Right, right. Exactly. And so then you get into the multi-homing question. So if somebody's looking at that, and like most things in power, the barrier question is usually the hard question. And so that's the one you'd have to tease apart is to say, okay, why do you really think they can scale faster?

And once the market is kind of saturated, why can they keep that difference? So there's a dynamic antistatic question involved in that. So the character of multi-homing is critical in all of this stuff. I do have one sort of summarizing question on this whole concept of power as it pertains to platforms. Is it right to think about it like if I was operating a general store in 1850, I could apply the seven powers framework. But today, if I'm operating a platform business, as many, many tech businesses are, especially the most valuable ones, as you pointed out, I kind of have to perform a seven powers analysis on each of the transactors on my platform, whether they're the supply side or the demand side.

I need to understand what gives my business power given that specific segment. Is that right? I actually echo with Ben's observation. Interestingly, we've been trying to mathematize all we talked about today for a long time. And every two weeks, we come across one of those edge cases, be like, ouch, that one is not included in this framework. And the exact reason of that is, Ben, what you said, the participants in this transaction is going to have a different economic structure.

They have a different industry structure that alternates how they behave and what they optimize for. So this is exactly what makes platforms so complex, but so interesting. It is you have to look very carefully at those, and every platform is going to come out a little bit different. You know, it's working progress for us. We hope we'll one day have a very abstracted and generalized framework that can describe all of those. But for every operator, you have to focus on the one that you own and how that's different from others.

I thought it might be useful for me to do a quick take of kind of a summary of some of the points that we've talked about in platforms. It's such an important topic. They're sort of like markets that are a way of exchanging, and that's what business is about in a certain way. And so there's a lot of value there. And the tech frontier that's been advancing has made many, many more platform plays viable. If you think of processors and displays and mobile distribution and all that, basically we all have this little computer that we carry around with us that makes a lot of stuff possible that just wasn't possible before.

And we see that in that many of the largest market cap companies the world are platform plays. So the first reason is lots of value there. The second reason it's an important topic is it's really complicated. And so it kind of strays into multi-sided market territory, which economists have spent a couple of decades trying to sort out. And I find personally to understand really what's going on is very intellectually challenging. So lots of value. It's very complicated.

And then the third reason is that we talked earlier in this podcast about there are two step changes in value in a company, product market fit, and then power. And in the case of platforms, one of the things that's really odd but very important is that those signals go in opposite directions often, that you can have tremendous product market fit traction. Something can scale really fast and be a big market. But the very thing that allows it to scale also lowers barriers, allows easy competition, multi-homing.

And so you might see something where there's huge value created. But if you ask the question of is there power, a completely different issue and may be at odds with the very thing that's driving all the value. So those are the three reasons. So then getting into the nature of the platform, value facilitates an exchange between heterogeneous buyers and heterogeneous sellers. And so it's exchange is what creates value matching typically. And to your point on the first component here, technology is really such a huge lever because technology lowers friction.

And so it makes possible much, much, much more efficient making of these transactions and making of these matches. And so that's why it seems to play such a huge role. Yeah, utterly. Think of matching. There has to be discovery. And then there has to be the mechanics of the transaction itself. And think of discovery without mobile. Right. Let's say I wanted a ride in the San Francisco area. Would I just call a lot of people to see, do you have a car and time to give me a ride somewhere?

I mean, that's really what we're talking about here, right? Having a computer that sits with you all the time that you pay a lot of attention to is a tremendous advance in availability. And then on the transaction side, all of a sudden you can push a button and you don't have to negotiate price or payment information is already in there. I mean, we all take this stuff for granted, but this just enables all kinds of essentially markets that didn't exist before.

So the value comes from matching a buyer and a seller. But power in these things is really different. It comes from one platform doing this materially better than another platform. And that the thing that drives that difference in performance is sustainable. Typically, that the difference in performance is driven by differences in the size of at least one side of the platform. So the New York Stock Exchange is more attractive than the London Stock Exchange because the liquidity is better.

There are more buyers and sellers and you got lower bid-ask spreads, right? But if all the parties involved had easy multi-homing, that would go away. But in fact, it's maintained by contractual arrangements. And it's really a pain to switch from one to the other. Yeah, the easy, obvious example of this that has been discussed much in the past decade is Airbnb versus Uber, right? Like Airbnb has so much unique supply that you're not going to find on Booking.com or HomeAway or any other platform.

Whereas, you know, Uber, it's like, yeah, I can get a car on Uber or Lyft or DD or whatever. So the question often is, first, so you have to meet two conditions. The value that one platform delivers has to be materially better than the other. And remember, this is a matching situation. And I won't get into the technical aspects of it, but basically it's equivalent to sort of a distance formulation. Even in multidimensional space is a square root function, which means the second derivative is negative, which means that it flattens out after a point.

And when it flattens out and whether there are two competitors that are both in the flat region is a critical element there. And then even if that is true, that it isn't flat enough between the two of them and there is a real material difference in deliverables, then the thing that maintains that difference, which is different scale on at least one side, has to be maintained. And so the degree to which multihoming is frictionless is critical.

Well, and we just see that play out so frequently in the startup and technology and venture investing space. I have on my mind because we're researching right now at NVIDIA. And they were the first computer graphics chip company, the first in Silicon Valley. And they thought that was going to be so great. They had such a great team and so much funding from Sequoia and Sutter Hill. And it was going to be great. Sutter Hill. Wow.

That's going back a ways. Yeah. And then, you know, they wake up 6, 12 months later and there's 90 other companies that have all been funded doing the same thing. You know, there's no power there. And NVIDIA, of course, had to develop power in other ways. Their segue of their business from one thing to the other is that'll be a wonderful thing for you to explore. You got to see what really does drive value. And is there a real difference between competitors?

And is there some lock-in of some sort of one side volume or count so that you maintain that superiority? I love it. So anyway, this has really been enjoyable as always. Thank you both. Thank you so much, Hamilton. Yeah. Thank you both. This is a real treat and a treat to have both of you. Welcome to NVIDIA to acquire it on the first time. I'm sure not the last time. And yeah, we can't wait to do this again.

All right, listeners. Now is a great time to talk about one of our favorite companies, Statsig. Yes. Long-time acquired partner. There is a reason why the best product teams at companies like OpenAI and Notion, Atlassian, Figma, Rippling, Brex, and more rely on Statsig, whether they are iterating on their core product features or shipping AI-powered experiences at scale. Yep. In the crazy speed of today's AI world, shipping fast is just table stakes now. It's basically trivial to build and deploy your app constantly.

The real advantage is how quickly you learn what changes actually created value for customers and how fast you can use that signal to guide what you ship next. Whether it's a feature tweak, a pricing change, a performance improvement, or an AI update like a model change or prompt adjustment, they're not relying on instinct. They're measuring what actually moved engagement, retention, and ultimately revenue. And as more teams build with AI, that learning loop becomes even more important.

Building with LLMs introduces non-determinism into your product experience. The same input doesn't always produce the same output. And behavior can shift in subtle ways in real-world use. So doing offline evals will give you part of the picture, but you can really only understand the impact once your product is live with real users, and then you can measure how their behavior actually changes. It's very different than the way that you would ship features in a pre-AI world where you knew exactly what the software was going to do in production.

Yeah, exactly. So this is where Statsig comes in. It brings experimentation, feature flags, and product analytics into one unified system so teams can ship safely, test rigorously, and directly link what they changed to how users actually behaved. The result is a tighter feedback loop and learning that compounds over time so you don't just ship more, you ship better. So if you want to make learning your competitive advantage, whether you're building new AI experiences or just evolving your existing core product, go to statsig.com slash acquired to get started.

All right, listeners, thank you for joining us on this journey with Hamilton and Chen Yi. I learned some stuff. David, I know you learned some stuff. We want to hang out with you, so come join us in the Slack. We've got over 11,000 members. We'll be discussing news of the day, this episode, dropping some hints for next episode. We've got a limited partner show that you should go check out. We've got three, I think David recorded, that we've released only for LPs or will release soon.

So if you're a paid limited partner at acquired.fm slash LP, you'll get access to those. Otherwise, after two weeks, all the episodes become public, and you can search acquired LP show in the podcast player of your choice to get access to those public episodes like discussing what's next in fintech with a partner from Bain Capital, Christina Melas Curiazi. Or what else have we released recently? Probably the composer episode will become public here soon. Not Boring Portfolio Company number one.

That's right. With Ben Rollert. Oh, the other thing on the LP show now with it being public, literally the number one most requested thing that we got for anything ever at acquired was, can I get the LP show in Spotify? And now thanks to doing this, you can. We've got a job board. These are jobs that David and I and head of special projects at Acquired Sani research and curate personally. So if you're looking for that next great career move, head to acquired.fm slash jobs and share this episode with a friend.

We love social media, of course, but we love that strong one-to-one communication even more. So thank you so much for listening. Later, listeners. We'll see you next time. Who got the truth? Is it you? Is it you? Is it you? Who got the truth now? Huh.