The Bessemer Cloud 100 Benchmarks Report 2025

Live from Schenectady, New York, it's SAS Talk with the Metrics Brothers, Growth and CAC.

And I'm Growth, better known as Ray Reich, founder and CEO of Benchmarkit.

And I'm CAC, better known as Dave Kellogg, EIR at Balderton Capital, independent consultant, And together we are the Metrics Brothers. And we go together like dark clouds and silver linings.

I'm not sure I know how to respond to that, CAC, but 2025 sure does seem very far away from the doom and gloom everyone was talking about the SaaS industry over the last 2 years.

Well, there are a lot of silver and even gold linings in today's software industry, Ray, especially if you're an AI company like OpenAI, Anthropic, or Lovable.

Well, that sounds like some of the key takeaways that I got from reading the Bessemer Venture Partners Cloud 100 Benchmark Report that they recently published.

Right you are, Ray. And in fact, I guess I should remind everyone, you are my metrics brother and we are now the Metrics Brothers. I think this is our first episode being recorded since that decision was made.

It is the first time it's been Metrics Brothers, but it's coming back to a familiar topic because we covered the State of AI 2025 report from Bessemer a few weeks ago, and today we're going to be covering their Cloud 100 2025 report.

Awesome. Well, I look forward to discussing that report, Ray, to see how both the more traditional SaaS and cloud companies are doing, as well as the AI newbies who are grabbing a lot of the attention.

Okay, but before we do that, let's hear a word from our sponsor. The Metrics Brothers is proudly sponsored by Benchmarkit, the The B2B AI and software industry's largest, broadest, and most contextualized benchmarking index. Dave Kellogg and Ray Groth Wright have put the Metrics Brothers podcast on hiatus for the remainder of 2026. All episodes we have ever recorded are still available on your favorite podcasting app. You can reach Dave at Kellogg and Ray on LinkedIn at Ray Reich. Thank you so much for listening to the Metrics Brothers podcast. Okay, Dave, a little background on the Cloud 100 Benchmarks Report. So this has traditionally been done as a partnership between Forbes, Bessemer, and Salesforce Ventures used to be involved. But today in 2025, it appears that Forbes is doing the public ranking article, and then Bessemer is doing the deep dive report. In fact, we'll just highlight this right up front. We couldn't find the top 100 companies in the Bessemer report, and we had to go sleuthing to find it on the Forbes website.

Yeah. So, so Ray, as the marketer in the two of us, I have a feeling this is Bessemer's media relations strategy at work, because as you may recall, when they published the Rule of X, that was also similarly weird. that the post was on TechCrunch, or the master post, or the gold record post was on TechCrunch for at least 2 months before it was on Bessemer. And when you see that, at least as a marketing person, I think, oh, they've done an agreement with TechCrunch. Whereas if you promote this and highlight this, we won't cross-post it for a given period of time. Smells like that is what might be happening here.

Well, by the way, you told me go do some sleuthing and we found it, but let's get a high-level overview.

I knew it was out there, Ray. We just had to find it. And I'm glad you sleuthed and I'm glad you found it. But readers beware, if you want the list of the top 100, it is currently not on bessemer.com/atlist today, at least.

No, it's actually at forbes.com/lists/cloud100.

Cool.

Okay.

Back to you.

So one of the things that was interesting that was covered initially in the report by Bessemer was the Cloud 100 achieved a record $1.12 billion in aggregate value. And that represented a 36% increase from the $820 billion value in 2024. Now, what's interesting, Dave, is the top 10 companies, and they listed those, and we can read them if we want, but they represented almost $598 billion of that $1.12 trillion. So that's 54% of the entire Cloud 100. is concentrated into 10 of 100 companies. And that's increased from 36% last year.

But so, I mean, Ray, this concentration is a phenomenon we're seeing in the public markets as well. For example, you and I were talking earlier, I think the S&P 500, the Magnificent 7 account for what, 30-something percent of the value?

Almost 34% as of yesterday.

Yeah. So, so this concentration is a real kind of industry-wide thing, public or private. The thing that makes me a little skittery on doing it for privates is private valuations are not, my mind, not simply not as reliable as public valuations. You're selling a relatively small slice of a company. You're selling preferred stock. Who knows what strings are attached to that stock? Like the Microsoft OpenAI deal. I don't know if you had time to research it, but there's hair all over that thing. So these valuations are not necessarily that comparable.

Yeah, in fact, OpenAI, which is the number 1 ranked company on this year's report, they reported it was $300 billion of valuation, which is almost 30% of the total Cloud 100. But as I tried to do my research, I wasn't good enough to be able to figure out exactly how they got that $300 billion valuation, Dave.

Yeah, and look, in between when this got published and today, OpenAI was recently valued at $500 billion in a secondary sale. Right. So these private valuations first move quickly. And second, to me, and I know it's true for some of these companies, there's what I would call hair on the valuation, meaning special terms whereby you can't just say if I sold 10% of the company for $30 billion, that it's worth $300 billion. That's the way these things normally work. I was going to say are supposed to, but I won't say that. I'll say that's the way they normally work. But if you've got blue shares and I've got gold shares, And the blue shares have a liquidation preference, or the blue shares, uh, have, well, exclusive voting rights, or if the views— the blue shares have warrants. Uh, there's a lot of things you could put on the blue shares to make them worth a lot more than the gold shares. So you can't take $10 billion worth of blue shares and multiply by 30 to get the overall valuation.

So a couple other things. So of that $1.12 billion in aggregate value, it was 42% of that value is coming from AI companies. That doubled year over year. It was 21% last year. And 22 of the 100 companies are AI companies. So AI is really starting to— I hate to use the word dominate, but become the majority. In fact, the only other category that's even close is fintech. And fintech only represented about 16% of the total valuation.

Got it.

So I think what I'd like to do is kind of move on, but anything else just at a snapshot of the number of companies, the valuation, and how AI is dominating it?

No, look, I would just say overall, let's remember what this thing is. Bessemer is a tier 1 venture capital firm, in my opinion. They're a venture capital firm and a tier 1 one, in my opinion. And they're publishing a list of the top 100 companies. And whenever a VC does that, there can be some concerns about who they picked. They don't really, I don't think, Ray, I didn't find any picking criteria listed in this report. So let's just say it's a rather subjective call from Bessemer on who the top 100 companies are. I didn't find anything I thought was outrageous in who they picked, for what it's worth. I think we'd look to see what percent of them were Bessemer companies. You did that and it's only around 10%. But what we're basically reading here, just so readers can be, or listeners can be aware, is we're reading basically a piece of marketing for a top 100 list. Um, and the reason people do top 100 lists is to be able to kind of promote the category they care about or promote the companies within it. So, and there's nothing wrong with these lists as long as you know what you're looking at.

Yeah, well, they also— the other thing they brought up was the Centaur status, and we talked about that in one of our episodes, which is the— was the new metric, months to $100 million ARR, because Centaur equals $100 million ARR. And what it showed here is if you look at the entire Cloud 100, it's about 7.5 years. But when they looked at AI companies specifically, Dave, those 22 that are in the index, it was only 5.7 years. So for fun, I thought, hey, I'm going to go and compare that to what Bessemer called the QQTTT, which is the quadruple quadruple triple triple triple. of AI startups. And when I do that math, starting at $1 million, let's say they hit $1 million by month 12, it's only about 4.7 years if they quadruple, quadruple, triple, triple, triple. So the math here in the Cloud 100 is different than what was published in the State of AI report.

Yeah, I'm not so sure about that, Ray. I mean, I have 2 clarifications here. One, the new metric Ray was talking about was not centaur status. It was months to $100 million, i.e., months to centaur status. They coined the term. I think it was they who coined the term and they coined it several years ago, at least probably 5+ years ago. And by the way, I think centaur is actually a good term because it's a revenue-focused metric, not a valuation-focused metric. But just wanted to clarify that because you said the new metric and then said centaur. And I just wanted to make sure people didn't think centaurs are new. They're not. The new metric is how many months it takes to be a centaur because that, you know, what is it? I think in our episode, we said that the mean in the top companies was 18 months or— some crazy figure for the hot AI companies.

Incredible.

The other thing, Ray, I wanted to touch on was I wouldn't say the math was— because you kind of implied it was inconsistent or wrong. All I would say is I wish these people would give more precise definitions for these rules of thumb, like T2D3, or in this case, doesn't roll off the tongue. I have to get used to it. Q2T3, quadruple, quadruple, triple, triple, triple. And because we don't know where to start, I think Ray started at $1 million. Does it start at inception of the company? Does it start at $1 million of ARR? Does it start at $2 million of ARR? Does it start at your first month of over $100K of MRR? We don't know. Um, so I would say it's broadly consistent. You came up with something 4.5, they were saying they were taking 5.7. So I think it's roughly consistent. And the real issue there is we don't know exactly what they're measuring.

I agree. Well, the next metric I wanted to talk about, I hope there's no lack of clarity, and that's year-over-year growth rate. And this is where An old metric becomes the new metric again. And that is over the last 3 years, we've seen growth rates in the Cloud 100 go from 55% in 2023 to 70% last year. And this year it's 75%. So they're really highlighting that as a good thing. But at the same time, what they're not highlighting is how capital intensive or what type of profitability These Cloud 100 companies are driving. And I would surmise that profitability for these Cloud 100 is probably secondary, if not tertiary in priority. What say you, Dave?

Yeah. I mean, look, I think, I think there's 2 things going on here, Ray. Uh, maybe you're going to hit one later on, but they're now talking about the Cloud 100. And I think in the old days, we used to use the term SaaS as a synonym for the term cloud. And I think we're going to have to stop doing that because cloud now has 2 subtypes, SaaS traditional and AI nouveau, and cloud could mean either. So my answer to your observation is I think the AI subset of the Cloud 100 is growing like a leaf without a care in the world as to burn rate, right? You don't raise $10 billion if you don't plan on spending it, right? So they don't care much about burn rate. Whereas in the traditional SaaS world, that has very much matured and it's become a world of, I was going to say rule of 40, but it's even more constrained. It's rule of 40, but you better be profitable. So it's rule of 40 for all profit, for all free cash flow greater than 0%. And in practice, it works out really to be the rule of 20/20. Like you better be 20% EBITDA, you better be growing at 20%. And all the better if you're 20/30, right? If your EBITDA is 30, we probably have them backwards. If your EBITDA is 30 and your growth rate's 20. So I think the worlds have kind of diverged here, right? It makes it very hard to do averages of cloud companies.

Ooh, you mentioned a word that really piqued my interest, and that is the word average. And I want to talk about average in context of company valuations that they highlight in the multiples that they highlighted here in this report.

Yeah.

So they actually said, okay, the average company average is valued $11.2 billion. And then they talked about the average enterprise value to revenue multiple. And they said it's 20x for the Cloud 100 this year. Then they compared that to 23x last year, 26x in 2023. And then they made a huge statement, but the average multiple is down 41% from the 2021 peak. And then they started segmenting. They say, hey, AI companies have a 24x average multiple and a 19x average multiple for SaaS companies. And the word average for multiples, knowing that companies like OpenAI and Anthropic are getting really high multiples, I just don't know how meaningful the average multiple is, Dave.

So, you know the old joke, Ray. Larry Ellison walks into a bar. Do you know the rest of it?

I don't. I don't know this one.

The average customer is now a billionaire.

Oh, okay. Got it.

Right? He's worth $300 billion or something. Bars don't hold 300 people. So literally, if Larry Ellison walks in the room, everybody on average is a billionaire. And the purpose of that joke is to touch on the point you're hitting, which is when you have these outliers, averages get meaningless real fast. I wish they'd use medians. You know, God made a simple solution to this problem. They don't use it much in the report. Like, I'm going to Ctrl+F the report for the median, and I'm not sure it's ever mentioned. They say average a lot. Yeah, median. Oh, it's there 4 times apparently. I can't find it, but I wish they'd say median more and average less. So then we get a better feel for it. But it's also hard when you take averages of segments, right? So they're saying amongst the top companies, the average is this.

Right.

It's not really average. Right? So I think they're trying to do it to show trends in the market to defend them. They're saying, hey, the multiple amongst the top 10, the multiple has been going up, I think slow and steadily until— well, I guess there was an inter-quarter peak. Well, okay, I see down. So that's moving up from 20 to 34 at the 2021 peak, and now it's at 24. So down from 34, still 19x for their non-AI peers. I don't know. I don't know what to make of this, Ray. So you already talked about it. Yeah.

Hey, you know, Dave, you— I, I remember before I really knew you very well, I was listening to a SaaStr presentation. You said be careful being a smooth operator. Well, I definitely don't want to be an average operator.

There you go, Ray. Don't be average. So, okay, in any case, look, multiples have gone up. I— my personal take on this issue is that for regular SaaS, they've come down quite a bit, which is kind of a fact. And that for AI, they're still in the stratosphere. One of the reasons they're in the stratosphere is because of the enormous growth rates, which the other thing, Ray, if we're going to talk about multiples, we need in this environment, we need to know if they're forward-looking or backward-looking. Because when a company's growing at 100%, right, that it, you know, it just cut the multiple in half to make it forward-looking. It's 24x. No, it's 12 because it's 12 based on forward, right? So I don't believe they're talking about which direction. And with high growth rates, you have to. Yeah.

Yeah. In fact, the other thing that they don't highlight, and it's not in the Forbes list either, is when that valuation was established. Was it from 9 months ago? Was it 1 month ago? It'd be really interesting to know when that valuation was established.

Excellent point, Ray. The other peril in private company valuations, which I neglected to mention, they're always as of last valuations. And at least I think recently a lot of companies are raising money, so they're probably pretty good in that dimension here for the AI folks. But a few years ago in SaaS, it might've been 2 or 3 years since the last round, right? Because the world blew up and everybody was hunkered down trying to use their last funding round to kind of grow into their last valuation. So I think it's an excellent point, Rick.

Yeah. And then the last thing I kind of, you know, carved out of this was they have 19 new companies in the Cloud 100 this year versus last year. And the number 1 new company is AnySphere. And they're the creator of Cursor. And they shared the kind of a little bit of the revenue trends. I mean, they're about $500 million today. They have 150 people. So that's $3.4 million per FTE. By the way, that wasn't in the report. I went and did my secondary research. But even more amazing, they were $100 million in January '25 and they hit $500 million in June. 5x growth in 6 months. Just incredible.

Yeah, it is absolutely incredible. And, uh, the only question I have, and I, I'm very— I love their success. I just want to know how recurring it is because I've worked with some of these companies where they have extraordinary growth, but they can't renew it all. Right. And you're, and, or you're taking monthly subscription multiplying by 12 and, or you're taking cancelable subscriptions. And I really personally have no idea what these folks are doing. I don't know. I know we're trying to get Carrie Liu on to talk with us about Cursor, right? Or no, Lovable. Where's she at? She's at Lovable.

Lovable.

Yeah, yeah. We're trying to get her on, and I doubt she'll share some of those secrets with us, but she could at least give us an insider view into the metrics behind these things. Yeah.

Now, the other thing, I want to already bookmark this for next year, a year from now, because of the 19 new companies, I think I mentioned 10 are AI. And I want to see how many of those 19 new ones are still in the Cloud 100 a year from now. Do they stop growing as fast? Do they get surpassed? So everyone go to forbes.com to see the 100 list. You can see all the newbies there.

Yeah. Okay.

So, and then the last thing, and Dave, you pointed this out, and maybe you want to talk about a little bit. The very first chart that Bessemer includes is the 10 years of the Cloud 100. And it does it back for 10 years. So kind of a longitudinal perspective. And they show exactly which companies were number 1. And back in the day, it was Slack, and then it was Stripe for 5 out of 6 years. And the last 3 years, it's been OpenAI, OpenAI, OpenAI.

Yeah, I mean, if you look at the last 10 years of the Cloud 100, and bear in mind, Ray, just as you were talking a minute ago about you want to see who stays on and who falls off, Because the list is somewhat subjective, either quite subjective and/or at least totally unknown what the picking criteria are, they can manage that number, right, if they want to. So you got to be— these things are basically PR stunts. I don't want to be negative, but they're interesting PR stunts. They provide good data that's fun to read and analyze. But if you get over-analytical on them, you get in trouble fast because we could look at that number, but with undefined criteria.

Yeah.

Who knows? They could decide if 10 people fall off or 20 people fall off, right? I mean, unless you're out of business and, you know, you're acquired or something. But anyway, that aside, assuming they try to do a fair job, which I think they do on such matters, if you look at the last 10 years, Stripe was there. Stripe was number 1 for 5 of them, OpenAI for 3 of them. So, 2 companies have done 8 of the last 10 years. Snowflake snuck in there in 2020. Right, and then Slack in 2016. So it's also fun. I think this is fun. They also put number 100. Now, how you end up 100 versus 99, who knows? But just look at some of the names. This may be a clue for you, Ray. Sky High, don't know where they are, you know, where are they now? Sky High, I don't know them. Canva obviously did really well. Pendo, number 100 in 2018. That company, did they make it public? I know one of the product analytics companies did. No, they did not. Than Amplitude did. Did one of them make it out? I thought one of the product analytics companies made it public, but it's not Pendo. But they've done reasonably well, I think. Dashlane, where my boss at Balderton is a co-founder. I don't know if you know that, but Bernard was a co-founder of that company. They've done quite well. They were number 100. LaunchDarkly, interesting company, number 100, right? Helping you do feature flagging. Axonius, I don't know who they are, Ray. This could be my ignorance.

No.

Front, I think it's Mathilde Colin's company. Uh, which was a kind of email management front-end company. Haven't heard much from them lately, to be honest. DeepL, I've heard of them, don't know what they're doing. Abridge and EvenUp. I should know who EvenUp is, Ray. They're in the news, but I can't remember what they do.

Yeah, EvenUp is a native AI company, and, you know, Abridge is healthcare. I, I don't want to say EvenUp because I'm not sure, Dave.

Yeah, then, you know, interesting, uh, this progression chart. One thing that really catches your eye And as usual, exits—

Number of exits?

Yeah, number of exits. So I'm saying it's not quite defined, but maybe that's good enough, uh, meaning either an IPO or a sale. I think that's fair. Probably be nitpicking, but, you know, I'll just read the progression: 82, 77, 65, 51, 37, 24, 9, 8, 7, right? So if there's one clear pattern on this, even if the list itself is somewhat manually picked, it's that the number of exits are going down. Um, whereas the exit value, the pattern's a little bit less clear, although it's down. In 2016, it was $286. 2017, 327, uh, 329 in 2018, 291. Yeah, then it starts dropping— 261, 194, 120, 75, then back up to 134. So, uh, in any case, Ray, I think it's a good chart. I like looking at this kind of longitudinal view on the data.

Well, I invite everyone to go look at the report. I think it's very interesting. It tells you a little bit about where the market's going, and of course If you can't say AI, you can't say the future of the Cloud 100.

Awesome. Well, thanks, Ray. Great episode. Talk to you next time.

Thank you.

The Metrics Brothers AI, SaaS, and More is a production of The Metrics Brothers Growth and CAC and a member of the Benchmarket Podcast Network. By accessing this podcast, you acknowledge that The Metrics Brothers make no warranty, guarantee, or representation as to the accuracy or sufficiency of the information provided or the humor content of the jokes told. Ray? Information, opinions, and recommendations presented are, according to our spouses, probably wrong and provided for information purposes only. This podcast should not be considered professional or, for that matter, unprofessional advice. We disclaim any and all liability for any direct, indirect, indirect, misdirect, incidental, special, ordinary, consequential, inconsequential, or other damages arising out of any use of or God help you reliance upon the information presented here. Ray Grothreich is based in New York City and available on X at Ray Reich. Dave Kellogg is based in Silicon Valley and available on X at Kelblog. Schenectady, which is French for unspellable, is not our actual production location. Thanks for listening.