 All right, Martin, it's so exciting to have you here and talk about cloud economics. Everybody's talking about the different choices in cloud and you and Sarah Wong just did this really in-depth analysis about the economics and looking at public cloud versus alternative. So tell us a little bit like why did you even like study this in the first place? So the Genesis was actually very simple, which is I sit on a bunch of boards and started noticing that like the number one item by far was cloud spin. And it was one of these things where we're like, oh, like over time it's gonna go away but it like tended to just get bigger. And when I talk about like number one item, I'm talking about like maybe like, you know, 50% of COGS. And so based on that private market experience and you know, all public companies were once private companies, Sarah and I asked a question like is, you know, like in the last six years we've had something like 126 companies IPO and all of them are cloud native. And this is a pretty new thing. So we just asked the basic question, like is this, you know, just the companies that I know about or is this everybody? And so we just asked the very simple question, how is the cloud impacting the economics of cloud native companies? But that was the beginning. That's, yeah, and that it seems like something that all these companies are wrestling with. And so it seems like it generated a lot of controversy. There were certainly tweets flying everywhere, people arguing and, you know, maybe you can tell us a little bit about some of the big takeaways from what you found when you actually dug into the data points. Sure. Okay, so the first thing to, I think that's important to stress is that this is a very new phenomenon. So like the cloud has been around for a long time. Public cloud companies have not. I mean, like that are actually built only on the cloud. I mean, literally maybe five years new. And in like, you know, one of history's greatest bull markets ever. So what we did is we looked at public software companies that were cloud native, we looked at 50 of them. And we asked two questions. The first question we asked was, if you look at the COGS, what percent of the COGS is cloud? Now remember prior to cloud native companies, cloud was some back office thing. It wasn't a part of COGS. So it impacted the economics differently, right? So this is just software companies. This isn't banks. This isn't old IT. This is just, you know, public software companies that are cloud only. So we asked what percentage is COGS? And then we said, and then we asked the second question is, you know, if you could reduce that, how does that impact share price? So here are the results. The first results, both results to me are just totally staggering. So the first result is, if you draw, across these 50 companies, on average, 50% of their cards is cloud. 50% of it, across 50 companies. And by the way, the way that we did this is we kind of looked at the rest ones and stuff at the committed cloud spend. I mean, we actually spent a lot of time doing this. And then the second result was, if you reduce that by half, however you decide to do that, right? Like let's say that you optimize your code or you know, you repatriate some set of workflows or whatever you do, it increases the share price across those 50 companies by about $100 billion. So if you extrapolate those results, yeah, these numbers are just so huge. So if you extrapolate those results to the industry, it's a $500 billion to a trillion dollar problem. And this is like market cap that's being captured by the cloud, assuming you can reduce your crowd spend by 50%. Now, you've talked a couple of times about COGS, which is cost of goods sold. So this is the raw cost that these companies account for across everything that goes into taking their product and service, basically out to delivering it to the customer. Now, one of the things that I noticed in the conversation that happened around this, I saw that people were pushing back and they were saying that the only thing to prioritize and optimize in terms of, getting a product or service to the customer is basically time and speed. And that by even though this is a very high cost, 50% on these public clouds, it was helping them with speed to market, with time to market. And that seemed to be like a pretty common sort of response and counterpoint to some of your findings. How do you respond to that? What's your feeling on that? Yeah, so, I mean, for those of you who are listening that don't know my background, like I was part of Cloud Wars 1.0. So Cloud Wars 1.0 was cloud versus internal IT, right? And so let's say, you're at a fortune, whatever, 1,000, and you need to get some desktop, you work with internal IT like this is kind of like the framing of the first Cloud Wars. And it's a very common one. And here it totally makes sense to kind of consolidate and have professionals do it, et cetera. But again, the study that we're doing are all these public cloud companies that are these very, very large companies. And to reduce a discussion to something as simple as well, you should just be focusing on growth. Would be the same thing as saying you should just take on technical debt, right? Somehow it doesn't like make much sense, right? And what's interesting is because the economics changed so much actually reducing cloud is an argument for growth. So let me explain that. So for example, if you reduce cloud costs by 50%, for most of these companies, you free up one to $2 billion of market cap that will flow through to cash, like cash money, which you can use to hire and invest in growth, right? And so the actual economics are so acute, you can make an even stronger argument that the way to grow is actually to cut your cloud costs so you can have the spend to do that. And the reason that I brought up Cloud 1.arguments is because like I think most of the discussion around the piece was like the old discussion around cloud versus internal IT, which had nothing to do with the piece. The piece is like Cloud Wars 2.0, which is cloud versus SaaS. And I think that's where the interesting discussion is. Yeah, so it seems like your analysis is pretty compelling at these really large public sort of cloud-native companies that it is, as you said, a very new phenomenon. And perhaps when you're really small and getting started out, yeah, public cloud, swipe your credit card, iterate, like it's not a big economic issue, but is there a crossover point in your mind? It seems like you found that sort of the clear case at these 50 largest and then perhaps you would agree if you're really small, iterate, it's not a big cost, but is there some sort of crossover point in your mind? So to me, the interesting setup for the discussion is if you're a large public cloud company, on average your costs are 50% of the cloud, right? It's impacting your market cap by a few billion dollars. And that market is controlled by an oligopoly of three companies whose margins have increased from 18% to 30% over the last 10 years, right? And so it's quite unlikely that like, you'll get much margin extraction out of that. I mean, oligopolies just don't work that way. So then the question is to your point is what do you do about it? Is it something that you worry about early? Is it something that you just let the market take care of? Like what do you actually do about it? And I don't think there's actually a good answer for that. However, we are starting to see signs in the industry of people recognizing it and thinking about it quite early. So for example, Suheel, who is the founder of Mixpanel has created a new company called Mighty for exactly this reason of building the entire hardware stack. If you've watched what Cloudflare is doing, Cloudflare is taking direct aim at AWS and announcing things like R2, which is kind of an S3 competitor by doing egress pricing. And so again, I think the question is less, if you are doing a startup, should you or should you not use cloud? The answer of course is yes, you must use cloud, for sure. Like that's the only way to do modern development. I would never recommend anybody do not use clouds unless you're a core infrastructure component. Then maybe you should think earlier about that. And then once you start getting at scale, you may want to start looking at these alternatives. There are now these kind of niche clouds that are focusing on cost optimization in certain areas that you may want to consider using. And so I don't think, the purpose of this piece was not to tell SaaS vendors to repatriate, it's to say that you've got a trillion dollars that are locked into this unsteady truce with an oligopoly and something's gonna happen. And it's gonna be big and a lot of spend is gonna be moving around and I'm all for it. So one last question for you. You know, this piece was awesome. Thank you to you and Sarah for all of the time and analysis that you put into it and publishing it and starting this awesome discussion. I think one of the reason why it got so much reaction from so many people was that in some ways it sort of defied conventional wisdom that you were talking about where it's just accepted that you go and you go to a hyperscaler and that's it and that's all you ever have to think about again. So I think that was great to put that out there. My question for you now is, what other kind of pieces of conventional wisdom do you think there are, especially in the tech industry that are changing or that are wrong or that, you know, what's the next firestorm you're gonna start, Martin? Well, I mean, I could go through the firestorms I have been starting, right? I mean, like the one previous that I'll actually work back for is I'm not quite sure what the next one is. I kind of talked about them as they come up. The one previous to this was I made the statement that kind of irritated a lot of people which open source doesn't really matter anymore. And the reason I said that is because I think the primary way of monetizing infrastructure as a service these days, and as the service, the hard part about that is less code and more kind of operational stuff. And so, you know, a lot of the, you know, top infrastructure companies are strictly closed source companies, we're just talking about the cloud. The cloud is almost strictly closed source, right? We're looking at AWS or GCP. I mean, like, you know, if you ran like a piece of infrastructure, you know, 10 years ago, like the amount of open source that you touched relatively is much more than now. Like if you run a workload on GCP, how much of that is actually open source? Way less than if you're running on Red Hat. So that was one. And then before that, and I believe this very strongly, you know, like I just think that we think of data as this magic thing and we say things like, oh, data network effects and data is the new oil. And I just think it's the wrong analogy. I don't think data has network effects. I think that companies that rely on data actually have diseconomies of scale. Like the more data you have, like the worse your economics are. And so I think that you need to trade like unit economics is the first class thing of your dealing with data. So those are three contrary positions I've held in the last two years. And then the next one, listen, you'll be the first to know. All right. Well, thank you so much. You know, we've had you on the weekly show before. We'll definitely have you back on to talk about data, debate, the open source imperative or not and all these other great topics. I mean, people are gonna love to get it and mix it up and debate with you. We'll have you back on for that show. But thank you so much for joining the keynote today. Oh, it was a pleasure. Thank you so much.