London vs. SF: A VC Story — Using Count AI to Compare Two Transatlantic Tech Hubs - Fine Cut

He's GQ's remote worker of the year and one of Fortune's ten thousand under ten thousand. Welcome back, Ollie. Thank you. Thanks for having me, Josh. As you'll recall, we called this working title because it was supposed to be a working title, and now we've been referring to it as working title. However, should it actually be called working title? I would like to workshop some new names with you. Hidden Gem or the stuff that counts is cool. I think the stuff that counts probably wonder if it's Hidden Gem. You like it? Hidden Gem. I think so. I think so. Okay. Ladies and gentlemen, welcome to episode two of Hidden Gem. Cue the music. Okay. I'll I'll put music in here. So today's task, we're gonna talk about venture capital, namely in these two cities, San Francisco and London. I think that it's interesting on many fronts, but some questions that we're gonna explore. A short history lesson, and then we'll get into the real numbers and what Count AI and I found. Some of those things include average company size by funding stage, what can we learn from that? Average funding amount by stage, again, always broken down between the cities. How industry impacts this, like how industry correlates with the number of employees or funding round in each given city. And then let's close with AI because I feel as though if we don't, the robot overlords will come for us at a later point. Tell me what you think is happening here. I love this test. I did not know I was gonna get a history lesson. This is not my strength. I think this is I think I have a a suspicion this is a coffee house somewhere in Europe where a lot of business used to have I feel like I know something about coffee houses being a big part of I'm gonna assume something along the lines of financier stuff. Wow. That is exactly what you're looking at. This is a rendering, an artist's rendering from, like, the sixteen eighties. Basically, was right next to a port. Ships would come in and out. And at some point when they were in here drinking coffee or doing whatever they're doing, someone's like, hey, fifty bucks, that ship doesn't come back. And someone's like, I'll take that action. Basically, embedded in in their DNA, in your DNA, Ollie Hughes. You all are risk takers. It's just in your blood. What does what does this second image look like? It reminds me of the western mining history. Nailed it. These people are mining for gold. There were lots of people who looked out west, specifically Northern California, where San Francisco and they're like, there's gotta be gold out there, you know, f it. I don't care what I'm leaving behind. I gotta go do this thing. And I think it does mesh nicely with some of the like misfit, rebellious, Jerry Garcia, Steve Jobs types that we would ultimately get out of San Francisco. We're all just looking for gold. What is what is venture capital if not just hoping to strike gold? There you go. Who are these eight guys? I wanna say there's some sort of amazing Marvel comic team, but I don't think that's true. What you're looking at are eight guys that are known as the traitorous eight. These guys are working in semiconductors for this guy named Shockley. They don't like him. He sucks. He like represents the stodgy corporate side of the emerging tech industry in the valley. And these eight guys decide to leave altogether. And they create this new company, which would become Fairchild Semiconductor, and they would birth fair children. The employees of this company as well as their technologies would give rise to Intel and to AMD and ultimately the foundations of Silicon Valley. Like the original PayPal PayPal Mafia, basically. Exactly. They spawned a whole range of businesses very successfully. Exactly. Alright. Last one. Similarly modern, but now we're we're going to your side of the pond. It looks like they're in where are they where are they standing? Is that is that King's College Cambridge? It is. Where'd you go to school? Went went to Cambridge. I thought so. This man is your fellow alum. His name is Herman Hauser. This is in the seventies. But he would go on to start a computer company. He's an alum of Cambridge. He got his PhD there. And he would go on to start a computer company two years after Apple actually called Acorn Computers. Something about people in this time were like stuff that falls from trees. I think we have to rename count to something that falls out of a tree. I think count is surprisingly sort of playing back to the original and just one simple word rather than this kind of make up a new word thing. So Sure. But I think we should rename Count to Pinecone. Pinecone. He, at Acorn, realized it was also becoming pretty corporate and stodgy. He leaves because he's got this vision. An a four paper sized tablet. Well before we would ever start using iPads. Right? He did impress some people at AT and T, sold it to them, and of course flopped. The first Palm Pilot, which I'm obviously, Palm became the prototype of very cool. The thing is though, to call this a Palm Pilot wouldn't be fair because it's very big. It's more of like a two hands pilot. Like imagine a sheet of paper, but that sheet of paper weighs probably like eight pounds. It's it's five hundred sheets of paper. Exactly. But in any case, even though he sold this, it didn't work out. But then he went on to start ARM. For those that don't know, ARM is a holding company originally, known for its CPUs. CPUs that would become embedded systems in literally every mobile phone in the world. Do you feel like you have, sufficiently come up to speed on why a startup philosophy is one that fits nicely into places like San Francisco and London, and therefore are worth comparing. Yes. Let's get to the data. Down here, you've got a table. This is five thousand eight hundred and thirteen companies. That is every company that is labeled on Crunchbase as active that has raised at least one round since nineteen eighty six. It is worth noting that San Francisco is over represented. Right? Four thousand five hundred ninety eight companies. That's roughly eighty percent of this data is San Francisco. Companies, I account for that. Think out loud. What what questions come to mind given what we're gonna look at and what you've been thinking about? The thing I'm thinking about here is like we've got a lot of history. The founding dates go back quite a long way. Like, Some of these are very old. It'd be interesting to see if we can pick up on like waves of VC investment by industry. I. E. There was crypto wave, there was the fintech wave, now there's the AI wave. Can we see funding kind of those waves coming through? Right? Where you go, oh, look. We had a bubble here that the VCs all doubled down on this industry for a period of time, and you can sort of see industry booms and busts as it were. Should I kick that off? See if we can do it? Yeah. Go for it. Let's see if I can get this question right. Yeah. You got it. Hey, Count AI. I'm interested in looking at this dataset, the dataset of companies and their funding amounts. I'm trying to understand if we can see waves over time of different industries getting different amounts of funding. So my expectation is you might see that a particular industry gets a lot of funding for a few years or maybe a decade, and then it sort of peters off as another industry's funding amount starts taking over. I think you can do a bit of that using the last funded date in this dataset. It will give us a bit of a glimpse of when were the when were the industries and companies getting funded. Can you take a look at that and see if we can see that as a plot over time maybe? It's a punchy question. Gonna open your agent on my end so people can see. Very good. There's my monologue. Let's talk about funding versus maturity. Basically, wanna understand the average number of employees versus the funding. So do companies in San Francisco get to have fewer people and significantly more capital as maybe is your hunch. It certainly is mine. Yeah. It's interesting. You can see that series a London companies have average more employees, but then at later stages, the the generally the San Francisco based companies have more employees as they grow. That would fit. I I think if that's people often talking a little bit about how there's a lack of growth capital, and so maybe that fits with this this picture. Yeah. And it's probably easier to see if you if you go to the chart next to it. I probably should have led you there. Starting to realize Oh, did there's a real table for, five minutes and, use my brain where I could just visually see it. If we're talking about medians, there isn't a whole lot of skew. With averages, it's worth noting series d. What series d has this massive leap for London employees and that is because of Just Eat. I guess I look at this and think, well, Just Eat is a series d. It's also enormous, but that was the last funding round I got was series d. So but it's employee size of they'd obviously just stopped taking funding and then just grew organically from that point. I mean, I I guess the the the thing to take away here is that really it is by funding stage, like, relatively equal. Although, it starts to indicate that there are that there is more investment for fewer people in San Francisco than London, which we'll talk more about in in a moment. Funding by stage. So for every stage, it's kinda sensible again to believe that San Francisco companies are making more on average. This actually turns out to be true. If we're looking at, I believe averages, series a, San Francisco makes thirty eight percent more than London at series a, and then twenty one percent more at series b. But at series c, there is suddenly a shift. At series c, when we're talking about median funding, suddenly London is London companies are able to make more money. Do can you make sense of that for me? I think the problem with this dataset is is is that potentially you're only looking at the funding the last funding stage the company has. So you're gonna have I could imagine, for example, that more companies top out at series c in London. They don't raise again, but they still might survive and grow. A lot of those series c companies might have folded or they might have not raised again. They might just stayed at series C. You've you've got some people who are still actively funding or still growing and some people like being around for a long time. So if you filtered this by, you know, take out people who've or the last funding date was five years ago, it'd be interesting to see things. They've been looking at all the most recently funded companies. It's huge. It's I'm wondering actually the range is really narrow as you go up the range the stages. Right? You have some stupid series a's happening over here. Yeah. And then actually the bat the bounds narrow. Like, it gets way more as you go up the funding rounds, you're looking more at like company spreadsheet. Like, what actually is the business performance looking like? How the metrics? Whereas in series c, you're more speculative saying this company can be massive. Let's fund it. And you can therefore maybe make a case for a bigger valuation at that point. So the I think the narrowing the narrowing comes down to the fact that the companies are being judged more on their metrics rather than speculative future performance. It's the train track to to venture success. Yeah. Me too. Let's talk about funding per employee, which the which the agent dubbed funding efficiency, which makes sense. So I guess I would start not at the table, but at the chart this time. Ollie, I won't I won't do you so dirty this time. We start to see this trend emerge that I think those that are tapped in at least a little bit to each of these markets could get. Where the number, the amount of funding per employee is heavily skewed towards finance or financial startups in London. However, what was surprising was that the top category is finance and that at roughly four point four million dollars per employee, it is the highest funding efficiency metric for any industry city pairing. So compared to any other industry in San Francisco, the amount of cash per employee that a company is raising is highest for financial companies in London. I think this is down to the fact that in London, the most successful startups are fintech. So I think that means that they they just they're gonna be going higher rounds, going for more money potentially. They're also more leveraged for sure. It's interesting that it's that that I mean, the fact that the average in San Francisco is under a million, but then in it's like, you know, four and a half million per employee in London. That is wow. It's crazy. My theory with this analysis in in funding per employee is that it would be still heavily skewed toward AI, like that brand of like AI infrastructure technology. And at the top of the list was finance. And then after that was therapeutics and biopharma. Right? Now, I think that it's because we're looking at historical data and there's there's still a lot of that. But in this table over here, I kinda dug into therapeutics a bit more. This is actually I'm very I'm very much not an investor, as I'm sure you know. But I I believe that in biotech particularly, there's a very different funding profile. Like, there's a lot of upfront costs to get a drug to a medical trial. So it's actually and obviously the trial works. It's like absolutely a money making machine to then bring it to the that product to the market. But there's a lot of upfront risk. Whereas in SaaS, I guess you can fund small and validate quicker. It's much much smaller start up costs. Definitely still interesting to see though that a lot of money going in, but that doesn't mean a lot of employees. These companies are still very, very small at the time that they're raising so much money. Yeah. That's that's interesting. Yeah. Yeah. Five point six average five point five average employees is cool. Let's look at this scatter over here before we get to our our very last and very brief section is looking at funding versus employees. So average number of funding, average employees. There's like a very clear kinda like u shape happening here. A big part of that is because there's one company in particular, that being Waymo, that is sort of the reason why these industries are overrepresented in having not so many employees and having a ton of money. At the bottom right, it kinda starts to make sense. I I can't quite explain agentic AI being down there. But with things like customer service, photography, I believe social media somewhere in here, that these are companies probably pre twenty fifteen that were, like, given a lot of money at the time where those were raising a lot of capital. And they have since been able to hire a lot of people. Right? Like Instagram is in is in this bucket down here. I kinda wanna see real quick if we oh my god. I cannot believe I'm gonna do sequel while being recorded. So let's see how this goes. Where oh god. I'm so scared. Where what's the name of the category? Oh, no. Stop. Close your eyes. Don't look. I'm gonna I'm gonna dig into agentic AI. For the sake of time, we're gonna continue. Alright. Speaking of AI, the latest gold rush ain't gold. No surprises here. San Francisco is giving a lot of money to a lot of AI companies. So for AI companies in San Francisco, their average funding amount is a hundred and two million with a twenty five million dollar median, whereas non AI companies are sitting at thirty three million average and fifteen million median. Worth noting that AI companies are have been raising in the last three years. There is inflation, there is cost of living, there are things to consider there. However, those margins are still the gold rush. London, basically even between AI companies and non AI companies, despite those caveats I just mentioned. So you're in London. You have built a product where AI is sort of at the at the center of it. What do you what do you make of that? I'm not surprised the amounts of funding are different. I think there is way more capital in the San Francisco ecosystem than London generally, but not less so now, less so maybe, but yeah, I don't know why it's different for non AI companies. I would say that I think everyone who's using software is using AI now. I don't think it's maybe just whether they label themselves as a true AI versus being AI enabled fintech or whatever it is. So the next chart over distribution where left column is AI companies, right column is non AI companies. Once again, you can see our wonderful hedge fund up here at the top. And then like Anthropic x AI, Waymo representing these AI companies. So Waymo maybe historically wouldn't mark itself as artificial intelligence in twenty fourteen, but now does. There's one large AI company out of the UK called Isomorphic Labs. Have you heard of them? Well, I think I have. Yeah. Yeah. Do know what their deal is? No. Let me just do a little thing for us here. I'm just gonna do something. Take it taking taking your chart, and I've just filtered down for series a, and I think this might be interesting because this is gonna give us the most recent series a's, first of Jan founding date. So series series a's last funding round and then founding date after January first twenty twenty two. Okay. It's not well, you can see that there is still a lot of companies getting funded, non AI companies here who are getting decent funding. It's not the distribution is very different than it is in AI where there's much more clustering. Well, it certainly does say that there are non AI companies that can still raise quite a bit of money in the year of our lord twenty twenty six. Maybe it's those biotech ones. Biotech ones that are certainly claiming that AI is driving their drug discovery. Okay, Ollie. Thus thus concludes, I'm gonna steal your chart and add it to mine and make it seem like I made it and thought of that. So thank you very much. You're welcome. I'm gonna follow your cursor now so we can check out your findings and yeah. Let's just let's let's talk about it. Yeah. Well, I'm just having a little look at this what the agent's done. Let's see what this is at the top. Software and IT dominate through the twenty tens. Healthcare and biotech is the moment around twenty eleven to twenty twenty. AL and ML defining way mid twenty twenties, which we would agree with. This all sounds very sensible, but if sort of sudden spike in twenty twenty six, maybe that is a, oh, small under of very large deals. That must be automotive driving. It's cool that it pulled out all these trends that we would we would look at and go, that is true. We would definitely say software had its time when it when software was massive and then a l AIML has taken over in the mid twenty twenties. It's done a kind of a total funding group by different industry groups. This is the the hedge fund we were talking about there, which is complete outlier of the top twenty London companies by last funding amount. It probably is almost certainly skewing things. When it comes to total funding by industry, I guess, taking into account even if you would take out the that kind of outlier hedge fund, you would still see finance being incredibly high or financial services being incredibly high, I think. Ollie, I'm gonna interject here to tell you that Isomorphic Labs' slogan is reimagining drug discovery process with AI. There we go. There it is. Yeah. Clarendon Arbitrage is definitely definitely causing let's get rid of it. Is not. It's too big. He can sequel on on camera so smoothly. Just don't think about it. My suggestion. If you think about it Yeah. It's actually if imagine you're not there, it's fine. Picture the sequel in its underwear. FinTech is still high. But I gotta say, that was definitely Yeah. Skewing things. So FinTech is still high for the UK software, financial services, that's still up there, but it was definitely pulling it up. I'm not sure we're gonna necessarily with this particular dataset blow open the VC world's funding his funding patterns, but we dug in. It was cool. Yeah. And I think this will be the first of several VC slash startup world adjacent analyses we do, but hopefully, you all learn something at home. For for Ollie, I'm Josh, and, yeah, thanks for joining our episode two of hidden gem as it is now known. Well done, Josh.