Subscribe
Share
Share
Embed
Talking through the big questions defining private capital.
Brought to you by Affinity Presents. Public conversations in private capital.
The Cap Table Ep. 1 - Repeat or Reset? Is AI Venture the New 2021? | Jason Saltzman, CB Insights
English (US)
00:00:00.160 — 00:00:06.240 · Jason Saltzman
My name is Sam and this is The Cap Table. My name is Jason and this is still The Cap Table.
00:00:10.000 — 00:01:02.850 · Jason Saltzman
Well, let's dive in. Repeat or reset. Does the AI deal market of today feel like 2021 right now? What are the parallels and what feels tangibly different? Parallels are the massive inflation of deal size. Right. Sort of across the board, whether it's at seed or pre-seed all the way out to sort of the D, E, F+ deals.
Uh, the thing that feels measurably different is the concentration of those dollars and where all of that capital is being deployed. Right? I mean, you look at the average or the deal count is dropping dramatically and, you know, deal size is saying very high. We saw sort of the two happen in parallel in 2021.
And so I think there's a lot more picking and choosing and true diligence on what will be a durable company and what is worth those? You know, high price or high ticket price bets?
00:01:02.890 — 00:01:10.330 · Sam Andersen
Yeah, it feels like I mean, when we talk about macroeconomics, you hear a lot about the K shaped economy. It feels like we almost have a K shaped venture economy.
00:01:10.370 — 00:01:11.090 · Jason Saltzman
100%.
00:01:11.090 — 00:01:36.850 · Sam Andersen
Where you've got founders who are out there and barely have to breathe, and billion dollar valuations are thrown their way. And simultaneously, you have founders who maybe aren't building in the sexiest of spaces or don't have a domain that ends in AI who can't seem to pull together around despite the fact that their business is growing on what traditionally would have been considered a respectable track.
And now, because it's not in this one category, it's kind of dead.
00:01:36.890 — 00:02:38.460 · Jason Saltzman
Yeah, and I mean, I also think it's right. It's not at the frontier of this one category. Yeah. In particular. Right. Whether that's horizontal, vertical, you know, sort of like frontier itself. So the labs, uh, and what will be sort of the future labs. Right. And you look at the hot exits from the labs and, you know, the Yann LeCuns, the Mira Muratis, the SSIs, right?
All these companies that are like not only unicorns, but decacorns on birth, right? You know, we're seeing that in a way that a has never happened before. And B, you know, really rewards this long shot bet on what the next transformer moment is almost in a way that like you look at 2021 and there wasn't a whole lot that was quite so revolutionary, right?
It was sort of this like tail end of the cloud and SaaS era where people had just gotten really good at understanding the sort of growth mechanisms, almost as opposed to the, like, foundational innovation mechanisms.
00:02:38.500 — 00:03:31.000 · Sam Andersen
Yeah, I remember CB insights, I think used to produce this. You probably still make it the the unicorn chart by itself, vertical or vertical. And there were years there where it felt like I mean, it just would have kept expanding and expanding and expanding. And I'm assuming today we're still seeing new names added, but frankly, the unicorn chart probably isn't as interesting as the decacorn chart or even the hectocorn chart.
In in the idea that some of these businesses are now getting 100 plus billion dollar valuations. Obviously, still in the early stages of private markets, uh, at least what we would call the early stages of private markets. Years away from any consideration of of public trading and that bifurcation, I think.
I don't know, it's a very interesting trend. It makes you wonder if the cash, frankly, exists out there to ever take these companies public. At least in the near term.
00:03:31.240 — 00:04:30.730 · Jason Saltzman
Well, I mean, I actually think it's almost the opposite question of it's not. Does the cash exist to take them public? It's the cat. It's almost a statement of the cash exists to keep them in, fuel their growth privately. And so it's the opposite side of the coin where, you know, even when you look at a lot of the deals from 2025 early this year already right there far down the alphabet in terms of series, um, you know, KS fs whatevers and they're for these companies that are decacorns or hectocorns that are admittedly you have so, you know, so many companies that are out past the largest ever IPO valuations and still private.
Right. I mean, if you look at the top 100 IPOs of all time versus the top 100 private company valuations of all time, and add up the valuations for both. Right now, the top 100 private companies are worth two x. One of the top 100 IPOs of all time are. And so we have this weird market.
00:04:30.730 — 00:04:52.740 · Sam Andersen
I think I saw a Wall Street Journal article that talked about how I'm pretty sure SpaceX was trying to chat with a lot of, uh, the Vanguards and the State Streets of the world to say, hey, how can we be included in the S&P 500 upon IPO? And we would like to make sure that that cash is ready to help fuel the initial listing, which is obviously something we've never seen before.
00:04:52.780 — 00:05:33.180 · Jason Saltzman
Something we've never seen before, right? I mean, I was on a panel at South by Southwest this past weekend talking with the CEO of Nasdaq Private Market, the founder of Manhattan Venture Partners, which is one of the largest secondaries funds out there. And the guy who's the head of secondaries at Citi.
And we were one of the questions that came up was like, look, our is going public a bad product for companies. Now, when you look at the pace of innovation and sort of the horizons that companies have to judge innovation on now, if you're trying to make a quarter and deal with sort of the regulatory processes and the filing processes that are required by being public, and especially if you're not within the indices, it's probably not a great product.
00:05:33.180 — 00:05:33.580 · Sam Andersen
For, you.
00:05:33.620 — 00:05:33.820 · Jason Saltzman
Know.
00:05:33.820 — 00:05:34.180 · Sam Andersen
I mean, it's.
00:05:34.180 — 00:05:52.910 · Jason Saltzman
A lot of sense looking at space, looking at the likelihood of maybe it's this year, maybe it's not then OpenAI or Anthropic or Stripe or Databricks going out as well. Right? I think they all have to go out and create. You know, maybe SpaceX will be the prototype for this, but create a mechanism by which they are included in the indices immediately.
00:05:52.950 — 00:06:44.710 · Sam Andersen
Yeah. I also think a core difference between 2021 and today is we all have our eyes pretty firmly planted on this next revolutionary shift. That is AI the last. You know, it's being compared to the creation of the internet. It's being compared to to the mobile introduction of the mobile platform with the App Store.
And where before you would look at some of these software companies. And I think the underwriting of those software companies was really based on the fact that money was free. And so prices exploded. Today, it feels like the underwriting of these companies is based on the fact that people legitimately believe we're looking down the barrel of the next Salesforce or the next Apple computers, because this new technological innovation has arrived.
And so no longer is unicorn the goal? It's we now have trillion dollar companies sitting out there that we can stare at and underwrite to, so to speak.
00:06:44.750 — 00:07:53.580 · Jason Saltzman
Yeah, I mean, I don't I don't even know that. Right? Like, yes. You say Salesforce right. I'm going to treat Salesforce and Apple is very different companies sure in this. But you know, if you look at Salesforce or this, you know, most of the SaaS products, right? It's not that AI becomes this new version of SaaS.
It's what else does AI displace or disintermediation further down the line? Right. And we've all heard the AI replaces labor. Yeah. But like that is really what that trillion dollar plus valuation or promise of these AI companies comes from is you're not just replacing software spend or tool spend, you're replacing labor spend.
You're changing the shape of how the economy works. You're changing the shape of how people interact with everything from right humanoid robots to AI for drug discovery, for all of these things that SaaS really never could touch, mobile, never really could touch cloud will enable. Right. Like, you have to think of a lot of what has come before is very foundational.
But the reality is that, like there's entire new surface areas opened up at a pace of technology cycles. That is only accelerating.
00:07:53.620 — 00:08:36.380 · Sam Andersen
Yeah, it's interesting, honestly, to see the founders that come in with pitches because you still see founders that that come and they have the old business enabled with some new AI track. It's software for this, or it's a spend card for that. Or and then you see founders who come in and, you know, maybe it sounds a little crazy or maybe it's a little further afield, but they've come up with new ideas.
And those are the ideas that I think are really capturing a lot of venture attention right now are these individuals that show up and say, hey, we're not just trying to build a new CRM, but we're fundamentally rethinking the sales process and exactly what sales could look like if it was empowered by this fleet of agents or this new technology.
00:08:36.419 — 00:09:09.870 · Jason Saltzman
And so, I mean, maybe from your seed as an investor, right? What are like what's the differentiation for you? Is it like or where's where are you willing to place bigger bets or differentiated bets? Is it on or like? Are there still valuable businesses to be made on? Like AI reinvents or recharges an existing system?
Or are you really focused on, hey, like this person came up with something net new and is a big bet. And if it works, that's going to like fundamentally reshape how that type of work gets done. Yeah, I think.
00:09:09.910 — 00:10:19.760 · Sam Andersen
Every investor will probably have a different opinion on this, but the way I've thought about it a lot is it's reentering the vision of the founder and that skill of being able to look into the future and live in the future and create something that the rest of us can't see as the primary value that you're investing behind.
I think there was a time where it felt like, for whatever reason, technology was largely stagnating, and the founders that were being rewarded were the operators who could show up, who could get a team to execute with pace, who could manage cost effectively and drive growth metrics to completion. And today we're back to hey, who's got the vision?
You know, where does the the idea really lead us and how does it change the future if it becomes realized? Those, frankly, are harder to find. And the line between what's a great vision and what's crazy is pretty thin. And so you're always trying to weigh, hey, is this actually like if if this feature were real, is it valuable?
Yeah. Would would people want that. And I think sometimes even some great ideas frankly are not viable. Yeah.
00:10:19.760 — 00:10:45.290 · Jason Saltzman
So and maybe maybe that's the best possible answer we could have for the question we just kicked off with, which is what's different between 2021 and now, right, 2021 rewarded the the operator founder. And people were more willing to bet on something that was maybe a little bit less of a rare breed. Right now, the bets have shifted.
The dollars in deployed capital have shifted towards founders that are these visionary, you know, steps ahead. Futuristic thinkers again.
00:10:45.330 — 00:10:46.170 · Sam Andersen
Yeah, the reward.
00:10:46.170 — 00:10:58.850 · Jason Saltzman
Is big and there's fewer of them. There are fewer deals, but with bigger checks. Yeah. Are there largest firms behaving fundamentally differently on ideals than everyone else, or is this whole market moving at the same speed?
00:10:59.570 — 00:11:15.729 · Sam Andersen
Uh, I think they are behaving fundamentally different. Uh, there's a lot of voices out in venture capital, but one that I read and like consistently is, uh, venture a general partner of Contrary Capital. His name's Kyle Harrison, and he he wrote about how
00:11:16.730 — 00:11:53.350 · Sam Andersen
what some of these funds are doing with these giant rounds, uh, be it the Sequoias or the a16zs, etc. of the world and what your run of the mill venture capital fund is doing are different asset classes. It's not even the same game anymore. When you can step up and deploy 150 million or 200 million into what is, by any other stretch, a seed company, uh, you're not playing the same game, and I assume LPs know this and think about this and and operate differently when they're considering these different investments.
But I think they are fundamentally. I would assume the data would say the same. CB insights.
00:11:53.390 — 00:12:59.400 · Jason Saltzman
Absolutely. The data certainly backs up the fact that, you know, the largest firms sort of, at every stage, are behaving fundamentally differently than sort of their peers that are either stage specific or sort of a little bit smaller in scale and have to be a little bit pickier and chooser. Right. I mean, everyone's picking and choosing to the conversation we were just having on what's different between 2021 and now.
But when you look at the data and start to see, just like just the volume of activity, whether it's in deals done or dollars deployed or both by the largest firms, the entries into the sequoias of the world, right. You know, they have everything that's starting. You know, speedrun is starting to rival YC, right?
Some of the Sequoia early programs are starting to rival YC. And then you have the same firm doing deals that are deploying billion dollar checks into single companies and or single rounds. And again, there are LPs and different funds within those firms. Right. You almost have to treat like how each of these large firms is operating fund by fund as opposed to firm wide.
00:12:59.440 — 00:13:30.640 · Sam Andersen
Yeah. Do you do you find that is it fun size specific in the sense that the larger the funds get, the more they behave like this? Or is it more tier or venture? Respect. And let me rephrase the question. Are they behaving these large traditional venture capital funds beginning to behave more like growth funds, like we have seen almost private equity style mid-market funds that focus on technology.
Are they mirroring each other more and more?
00:13:30.840 — 00:13:53.410 · Jason Saltzman
Yeah, I mean certainly. Right. Again, I'm going to say with like for those fund for the growth funds for the funds that are like have this mandate within these large firms? Yes. Right. I mean, you look at, you know, the canonical example of this right now is General Catalyst. And they're literally acquiring companies doing roll ups and trying to treat this as a PE style play, all while still being a VC.
00:13:53.450 — 00:13:53.730 · Sam Andersen
Yeah.
00:13:53.770 — 00:14:02.450 · Jason Saltzman
And so, I mean, yeah, the blending of asset classes and the blending of skill sets that are happening at these large firms is certainly sort of
00:14:03.890 — 00:14:10.090 · Jason Saltzman
just a different game than you're like run of the mill even like run of the mill top tier,
00:14:11.170 — 00:14:13.370 · Jason Saltzman
uh, VC firms are operating at.
00:14:13.410 — 00:14:43.660 · Sam Andersen
Yeah. I mean, General Catalyst is a prime example. They bought Summa Health. Yeah. Out of Ohio, if I'm not mistaken, is a literal hospital chain. And the whole goal from what I read was just to deploy portfolio companies and products and figure out exactly what works to rebuild our health care system. Obviously, that's a pretty grand vision that goes beyond just selecting a single product or a single technology company to invest some money in hopes for an outsized return.
They're reshaping a space.
00:14:43.700 — 00:15:24.340 · Jason Saltzman
Yeah, well, I mean, even you see the news yesterday or today about what Jeff Bezos is doing with Prometheus, his new manufacturing venture. Right. Same same principle applied theirs. Right. Look, where can I take all of this learning, this AI forward approach, this right future vision of what manufacturing could look like, not just at Amazon's scale, but at like if we had took Amazon scale and rolled it out to everything.
Yeah. And yeah, you're playing a different game. You're making is more like you make the future that you want to see and create new surface areas for your portfolio companies to operate in, as opposed to just, hey, we're going to back great companies that are going to themselves change this space.
00:15:24.380 — 00:16:13.310 · Sam Andersen
Yeah. It's interesting. As a young venture capitalist, I often heard about fund returning companies. You invest in one company, and the goal is that every fund needs to have one that returns the whole fund at least once, if not many times over. That's how the power law dynamic was taught to me. When you get to these multi-billion dollar funds, I don't know how they're teaching young investors to think about the underwriting model for individual companies inside of that fund, because obviously, not every deal you do has the capacity to grow that large.
Um, obviously some may. There's a once in a generation OpenAI or anthropic that comes along that truly could scale to that size, but those are rare, and you can't build a model on any consistency believing that that size of bets are possible.
00:16:13.350 — 00:17:23.610 · Jason Saltzman
No, but I think a lot of what we're seeing is this increased sort of like, hey, because right to what we were talking about in the first question. Right? Companies are staying private longer than ever. Private markets are growing, and firms want to be able to either support or capture return at more and more stages over that duration of a company staying private.
So you think about Stripe for an example, right. If you did the A yeah, it's been a long time since that. A there's been a lot of value creation for that company. And there's been a lot of opportunities at BCD tenders. Yeah, etc., to capture more and more of that at different risk profiles. Yeah. Right. So, you know, yes.
You underwrite the A at a certain risk profile that look, we want this to be a fun returning company for a early stage fund. Yeah. Then you underwrite the D or the E or the F or the tender. Yeah. At a far different risk profile and a far different. Hey, how much does this company have to grow for it to be worth our investment at this stage?
But now you can sort of support the entire trajectory of a company, if you like. These largest firms are doing operate different funds all the way along the way.
00:17:23.650 — 00:17:24.850 · Sam Andersen
They're almost passing.
00:17:24.850 — 00:17:25.370 · Jason Saltzman
Shares.
00:17:25.730 — 00:17:36.260 · Sam Andersen
Are they passing shares between the funds? Are our growth funds buying secondary from seed funds within the same kind of corporate umbrella.
00:17:36.860 — 00:17:39.100 · Jason Saltzman
It's a good question, I don't know. Uh,
00:17:40.740 — 00:17:46.260 · Jason Saltzman
we can we can maybe ask, uh, the affinity folk if they have some data on that, and I can maybe see what I can pull out.
00:17:46.300 — 00:17:46.740 · Sam Andersen
Yeah, that'd.
00:17:46.780 — 00:17:55.979 · Jason Saltzman
Be interesting insights. I mean, I know that, you know, there are certainly funds that land that are landing in primary and then expanding through secondaries. Yeah. But the
00:17:56.980 — 00:18:04.420 · Jason Saltzman
you know, when you look at the behemoths, uh, in the room, I don't know how it's operating within Sequoia or within a16z.
00:18:04.460 — 00:18:16.260 · Sam Andersen
Yeah. You know, a lot of venture funds in this last fundraising cycle. It's raising fall on funds or continuation vehicles or, you know, there's lots of names for for what they wanted to.
00:18:16.300 — 00:18:18.820 · Jason Saltzman
But put more put some money in. Put more money in later.
00:18:18.860 — 00:19:32.960 · Sam Andersen
Yeah, exactly. And the idea being, hey, we've got the same LPs. We now have this waning batch, um, of companies that we wish we could continue investing in and just don't have the same capital amount. And I think obviously the vintage was rough for them. I think those fallen funds, largely as a whole, will be seen as a failed experiment because they invested strong capital beyond their stage of expertise.
These are often seed in series A style investors who are now investing in B and C rounds as follow on. And they're not re underwriting a broad universe of company options. They're restricted only to the companies that they invested in at the seed or series around. And I think it'll be interesting to see how this all shakes out, because to your point, they can follow a company through its trajectory, but they're also underwriting a much broader suite of companies alongside them.
So they kind of get this advantage when they want to use it, but there's no requirement. I don't know. Yeah. It's interesting. Okay. Question number three. Our founders pitching differently knowing firms are moving faster is speed to fund changing what actually gets funded. Um, I'll take this one.
00:19:33.000 — 00:19:35.200 · Jason Saltzman
Yeah. This is this is a far better question for you.
00:19:35.200 — 00:21:23.470 · Sam Andersen
I see a lot of pitches. Um, I think that founders are pitching differently, you know, to, um, to say that salesmanship is not a strong part of what gets funded and what doesn't get funded, and that decision would be lying. Of course it does. And I think founders capitalize on what they know VCs want and what they know VCs are looking for.
And a lot of what we see today in the market is that speed matters. It always has, but it certainly does now. And so a founder that can effectively sell a vision quickly can express why they are at a unique time to capitalize on an idea and simultaneously make the investor feel this FOMO of you got to commit now or else you're going to miss this train.
The deals get done fast and we're seeing them done very, very quickly. I think the similar venture mentality, the similar herd of once one term sheet is in, the rest will throw theirs in as well remains true. You'll see, uh, you know, NEA, throw in a term sheet and before you know it, there's 12 term sheets on the table.
And, you know, we're moving down the road. But, um, but yes, I think absolutely. And I think founders who have the product and the ability to create that FOMO successfully are rewarded simultaneously. I think founders who try and don't quite have either the vision or the product or the skill to truly instill that feeling, can almost be punished because it feels, to a certain extent, like the founder is over pressuring, overplaying their hand.
At which point then at least I start to ask a lot more questions and in more than happy, usually to hang around the hoop and see what happens. Yeah, Yeah.
00:21:23.510 — 00:22:51.520 · Jason Saltzman
No, I mean, I think with some, like, anecdotes, uh, even just from around in an investor seat, but I certainly talked to a lot of founders. Yeah. Um, sitting in this overpowered spectator seat with all of the data. Uh, and within the ecosystem, it's pretty clear when you talk to a founder right now, like, yes, everyone's trying to move fast, but the different, like, the juxtaposition between the two.
Like, I'm going to call it two archetypes, right? It's like founders who really are dialed right and are going to get that sort of like every term sheet all at once. Yeah. Uh, because they're moving fast, because they have the differentiated product and vision, because they understand how to communicate that in a way that like, seems at least seems defensible.
Yeah. Rather than I mean, you meet you meet a new founder every day who says they're doing AI for financial modeling. And you realize that, well, there are three companies that started in on this problem in 2022. Uh, you know, around the first releases of the GPT and have deep domain expertise. Yeah. Already have embedded relationships, and they're moving at the same speed, just with a much stronger foundation than someone who was a quant and thought, uh, now is the right time range.
Like, there's a difference between right time and right person. Right time. Right market. Right velocity.
00:22:51.560 — 00:22:56.840 · Sam Andersen
Yeah. And I mean, the founder really does make a lot of that difference. Yeah. Yeah.
00:22:57.400 — 00:23:25.050 · Jason Saltzman
All right. Uh. Question four. Um, and maybe this is a question about, uh, how you get in touch with those founders. Sure. Uh, affinity. These data shows that the least efficient firms send five times more outreach per person and get two and a half times fewer results. Is category crowding part of the explanation?
Are more firms chasing the same AI deals in the same spaces? whether it's PE or VC. Uh, and is nobody converting?
00:23:25.130 — 00:23:26.209 · Sam Andersen
Yeah. I think
00:23:27.370 — 00:25:00.180 · Sam Andersen
it's an interesting question. Is category crowding part of the problem? I think there will always be a scarcity of truly excellent founders with truly excellent ideas. And so obviously, if you get into those types of ideas at the ground floor, the seed round the A round, the returns are enormous. And when it feels like almost a done deal before the checks are even written, of course the herd crowds and all wants to get in.
Um, I think when it comes to how we're finding ideas and how VCs pitch themselves and think about their own value position to a founder, the product, so to speak, that they are selling to the founder. I think every firm is going to have their own philosophy. I come out of consulting. I spent years in Bain's private equity group, and one of the things that we would talk about with private equity clients is that your strategy is no less important than the strategy of, say, Walmart as it positions itself to customers.
Everyone knows Walmart exists to provide every day, every day low prices to customers. So when you go to Walmart, you expect a wide selection at a very low price. I think equally, venture capital firms are realizing, hey, we need to have a strategy and a brand that is consistent in the product that we're offering.
So you see different firms positioning themselves in a way that capitalizes on what they think founders want to buy. And, you know, whether that's, you know, founder loyalty. Hey, we're you know,
00:25:01.380 — 00:25:28.079 · Sam Andersen
we'll let you do whatever you want. Effectively, we'll have your back no matter what or whether it's we provide a suite of company resources. We'll do your recruiting, we'll do your back office. We'll, you know, hold your hand the whole way through. Those are different value propositions meant for different founders, and we hear them coming from different funds today.
Um, equally, I think across the country, even when you're not talking about mega funds, you're seeing regional funds step up and play a very important role because they are
00:25:29.880 — 00:26:26.650 · Sam Andersen
with Founders Day in and day out physically, long before those founders can find their way to Sand Hill Road, they probably know them as they're coming out of universities. They know them as they're in incubator programs or kind of progressing. Sure, that's not the same type of deal that's going to get $100 million seed round.
But it's where a lot of these young ideas start before they get picked up by a Y Combinator, before they get picked up in any type of mainstream press. And so those regional funds, I think, also are trying to recognize, hey, our regional strategy needs to work because we're certainly not going to be the seed round for the next anthropic if it's starting in the Bay area, like, you know, whatever name your small town venture firm just isn't going to get it.
So everyone is trying to think through, hey, how do we zero in on our strategy? Make sure that we're executing against our thesis in a really crisp way so we don't miss a deal that does come out of our niche, because you can't afford to do that either.
00:26:26.690 — 00:27:53.220 · Jason Saltzman
Yeah. I mean, I think I'd almost answer this question in the sense of like, look, at the end of the day. VC is a version of sales, right? You know, you're running a company that is trying to sell a product, right? And that is a a company or a startup taking your capital and then working with you over time. And the reality to answer the question on, like sending more emails and doing more pitches or overcrowding in a space, is that it's the same problem you see in go to market systems for almost any sector.
Right. You know, you have to do good marketing and say good emails and have good response rate. And guess what? That's harder than ever. But the firms that get it right are the companies that get it right. Get it really right. The same goes for a good sales motion, right? You know your GPS have to be out there really creating this ecosystem of, hey, you know, we want to have this community, we want to have these relationships, and we want to know that when someone is right for us and when we're right for them, we're going to win that deal, right.
You know, if you think about the full cycle of investing in a company, right, it's very, very similar or mirrors very closely, a go to market process for a company selling a normal product, you know, and so the thesis and the fit and the market that you're operating in, you know, to the, you know, the regional point you were making earlier is like, look, that's what avoids or that's what helps the best firms avoid this overcrowding or sort of like rush to this knife fight in a field.
00:27:53.260 — 00:28:11.830 · Sam Andersen
Yeah, I'm a VC that's based in Salt Lake City, Utah. I think if I spent my whole day sending emails to small tech companies based out of Silicon Valley, you know, Seattle, Washington or Austin, Texas, I think I would find my hit rate to be a lot lower than if I spend my day meeting, networking and emailing with founders that are based in Salt Lake City, Utah.
00:28:11.870 — 00:28:57.710 · Jason Saltzman
But I mean, it's the same. Again, it's really everyone for a while is like, oh, there's the death of the SDR, right? It's or, you know, it's the death of enterprise sales. Guess what? The death of taking someone to lunch isn't happening. Yeah. And, you know, that's going to be the thing that like again, avoids this overcrowding gives that differentiation.
And, you know, like whether it's the analogy you want to make is the the like good enterprise AI that's taking people to lunch or having a Ford deployed engineer or Ford deployed implementation person. Right. Like those are the differentiators and the ability to like, figure out and build a machine or build an operating system around that is what's separating, you know, sort of again, at like every size and scope of fund, the the winners from the rest.
00:28:57.710 — 00:29:09.710 · Sam Andersen
And I would assume that in the age of AI, the relationship that you build in person obviously becomes much more important and that humanity that you build.
00:29:10.720 — 00:29:46.440 · Sam Andersen
matters to a founder. You know, I used to joke that the average American. The average marriage in the United States of America lasts something like eight years. And the average seed deal in the US is a 12 year fund. And so this your seed investors can be with you longer than your spouse on average. And being able to build that relationship in a world where if you're a founder, you're probably getting who knows how many automated emails from whatever outreach, being able to build a sincere relationship over time, to know who is investing in your company, who you're inviting on to the cap table, and what kind of partner they're going to be over the long run, matters more than it ever has before.
00:29:46.480 — 00:29:47.600 · Jason Saltzman
100%.
00:29:49.400 — 00:29:50.120 · Sam Andersen
Okay.
00:29:51.520 — 00:30:06.480 · Sam Andersen
Question five. Are we on question five? Version five. Okay. What actually broke beyond high valuations. What specifically broke in 2021. What were the consequences for the firms and the companies? Was speed the primary killer?
00:30:09.380 — 00:31:20.670 · Jason Saltzman
I think it was just I mean, the easiest answer to this is probably it was the end of it was the end of a tech cycle. Money was free. Uh, right. Which sort of gets to the valuation problem. But, you know, like there was this over indexing on VC subsidized growth at all costs and whether that looked like high valuations, whether that looked like companies burning money faster than ever.
So, yes, speed, but like the wrong type of speed. Right. You know, it wasn't speed to go and win deployments. It was speed to tweak a few things and try to spend a bunch of money on paid, or try to spend a bunch of money on hiring people. Right? You know that it was like where the speed was happening. That was the killer.
Yeah. And I mean, you might have a different lens, um, to that than I do, but the I think the reality is, yeah, I mean, speed without discipline or speed without this idea that, like, what are we really trying to move fast at? And this goes back to the founder conversation or a question about the founders earlier.
Right. Like what's the difference between the really good founder and the like or like the good founder and the visionary founder?
00:31:20.710 — 00:33:37.010 · Sam Andersen
Yeah, I think yeah, I completely agree. And I don't think you can understate the fact that money was free. Yeah. And I think coming out of the pandemic, everybody imagined a world that was fundamentally different than before the pandemic. And I think if we look around today looks relatively the same. We went back to normal, things returned and reverted to kind of a pre-pandemic level.
And I think unfortunately or fortunately, depending on what bet you placed in 2021, there were there was a lot of money riding on that. Yeah. I also think that historically I was taught in school and I learned as a young investor that a company's value is a discount against all of their future cash flows.
Pulled to today, and I think when you take that discount rate and you drop it to zero, what happens is any company that's going to produce cash flows into perpetuity is infinite in value. You look at an Amazon, you look at a Google, you look at any of these companies that none of us can truly foresee the end of.
And you say they're worth infinite amounts of money. Rationally, if money is free, they are rationally worth, you know, trillions upon trillions upon trillions of dollars. And I think people began to see that in the venture space as well, to say, hey, if we can pull these terminal values of these DCF models up to, you know, $1 trillion, what can't you underwrite?
Yeah. And truthfully, that was never long lasting. I think a lot of people knew it at the time. We heard a lot of phrases from people who knew that it was crazy, like, you have to play the game on the field, and I understand why we said that. I understand why people thought these are the prices that are being paid.
This is what we have to do. But simultaneously, we see the consequences of that in the vintage data. I mean, frankly, these vintages are going to struggle relative to what we saw in the the teens simply because the entry points were way too high and everybody knew it at the time. If you go back, I'm sure we could find podcasts from people not too different than us sitting in a room talking about how crazy the prices were.
Um, yet you invest across cycles and sometimes you win, sometimes you lose.
00:33:37.050 — 00:34:07.180 · Jason Saltzman
Yeah. And I mean, I think one of the other things is, you know, when you look at what broke, right, I think you saw a lot of wedge products that had the promise of being platforms. Right. You look at all of the fintech companies that wanted to say, hey, we're going to start with a corporate card, and we're going to roll this out into how finance gets rewritten for everyone.
Yeah. Or pick pick your other example of this. We're going to start with a number of companies. We're going to start in mental telehealth. And then we're going to roll out to everything telehealth.
00:34:07.220 — 00:34:12.060 · Sam Andersen
The number of companies that had, you know, payments as their monetization strategy was wide.
00:34:12.100 — 00:34:43.260 · Jason Saltzman
And I mean, look, at the end of the day, you're still competing with like, the normal card networks, with the normal financial institutions to capture all of that GMV. And, you know, like all of the yield on any given payment coming through. And if you're only capturing a small sliver of that, and right, especially when the spigot turns off on money, not being free to subsidize that growth and subsidize the capture, then you sort of like, get left, right.
You know, it's the age old cliche, like the water goes out and everyone has their pants down.
00:34:43.300 — 00:35:36.960 · Sam Andersen
Yeah. And it does feel like some founders were able to see past the cycle and say, okay, money right now is really easy to get. I'm gonna go bring in a bag and I'm going to lock it down, and I'm going to run a responsible business building on solid unit economics over time because I may never raise again, at least at a price north of this.
So I want to make this money last. And then there are probably others who didn't. Who thought money will always be free. My experience raising that series aid took about four days. I'll do that again when I need more money and found themselves in a world of hurt, either laying off huge swaths of their employees or, you know, shutting down.
Have you guys seen at CBN sites? I don't know if we want to call them like zombie unicorns. I think that's the phrase that I'm seeing a bit. But are you guys seeing a lot of those unicorns that were on your charts actually close their doors?
00:35:36.960 — 00:35:37.399 · Jason Saltzman
Or
00:35:39.240 — 00:37:19.100 · Jason Saltzman
is this is this slow crawl, right? Yes. Some of them have closed their doors, but not as many as will over time. I think there's a lot of I think the thing that we're less likely or more likely to see than, you know, like going bankrupt or closing doors is just this, like gradual decline until He gets to use all the dry powder they're sitting on to roll up a bunch of these.
Right. To my point or to what I said earlier, the wedge, uh, products into the platforms they all wanted to go and become. Yeah. Or to reimagine them with AI. Right. As we're hearing the promise of for almost everything right now, the and then I think the other piece of it is like, yeah, I ran some data earlier at the start of 2026 to look at, look, from the end of 2024 to the end of 2025.
If you look at all 1300 unicorns or 1300 plus unicorns, how many of them were better off over that year period versus worse off over that year period? 53% were worse off, and most of those were of the vintage of companies that raised in 2020 to early 2022. So we'll call them the 2021 vintage for simplicity.
And yeah, like again, they're they're crawling towards their grave or digging their own hole or whatever phrasing you want to use. But I think a lot of it is just like that money that they. Set up as the sort of like to to burn is dwindling. And there, you know, a lot of them thought they were going to get to profitability or get to.
Like revenue generating or revenue funded, uh, growth stages. And they just haven't and they won't.
00:37:19.580 — 00:37:20.180 · Sam Andersen
Yeah. You know.
00:37:20.180 — 00:37:31.940 · Jason Saltzman
Unfortunately, yeah. This is this is this is I'd rather be a techno optimist, uh, in this. But there is the reality that we're facing sort of this, like, inflated and like what, what goes up must come down.
00:37:31.940 — 00:38:36.910 · Sam Andersen
Yeah. It's interesting. I used a calendaring tool for a long time called Clockwise, and I really like Clockwise. If the Clockwise team ever hears this I love the product. Um, and it raised money. I don't know how long Clockwise has been around, but it's longer than the AI revolution has been around. So probably early 2020s, You know, five, six, seven years or more.
And they just got bought by Salesforce. Yesterday they announced that they were being acquired by Salesforce. And the Clockwise product is shutting down next week. And Salesforce has no intention of keeping the Clockwise product. This is an aqua hire. It probably got some multiple you know some fraction point six back for their investors.
I hope the Clockwise team got a nice little bonus from John Somorjai and the Salesforce M&A team. And away goes that product. Now it was never a unicorn product but but the idea that we need 40 calendar apps or whatever out there in the world I think is just quickly being, uh,
00:38:38.030 — 00:39:02.480 · Sam Andersen
disproved. Yeah. You know, and this technology is being wrapped into these larger companies, or these people are being hired by these larger companies, and we're seeing more and more of this. Hey, well, we'll buy your company so that your investors get some money back. But really, we just want to hire you to join our engineering team.
We don't need anything that you've built as part of this company to last.
00:39:02.520 — 00:39:57.700 · Jason Saltzman
So it's funny, I was listening to a podcast yesterday. I think it was one of the a16z podcast episodes that also mentioned a Clockwise. I think that episode was filmed or recorded before the acquisition yesterday. But, you know, we were talking they were talking about the fact that, like, look like at the end of the day, I could probably go spend a weekend in vibe code my own clockwise for for me.
Right. And that doesn't, you know, they wouldn't necessarily operate at the scale that Clockwise was. But really to your point, like if, if it's easier than ever to recreate or reimagine one piece of functionality, uh, or like single functionality software, then really what you're hiring these people for is like, hey, how do I go build the native or forward looking version of this into, um, something like Salesforce, right?
It's like, how do how do I go do automated meeting setting. Yeah, as part of Agentforce or something crazy. And even more forward looking than that.
00:39:57.740 — 00:40:04.220 · Sam Andersen
Yeah, I'll be interested to see what those particular individuals actually end up working on inside of Salesforce 100%. Yeah.
00:40:04.260 — 00:40:15.780 · Jason Saltzman
To what we were just talking about and AI acquisitions, uh, what is actually differentiated versus what's notable? Uh, is there a specific category within AI that looks exactly like a 2021 mistake?
00:40:15.820 — 00:41:12.510 · Sam Andersen
Yeah, I think, you know, we talk a lot about like, wrappers. Yeah. Uh, you know, there's some cautionary tales and not wanting to mention names, but there is a whole fleet of businesses that were destroyed when ChatGPT was released because everybody recognized that for a $20 a month subscription, they could do what those businesses were charging thousands of dollars to do and more themselves.
Yeah. And I think today we see a lot of pitches that come in that say, hey, we're using AI for this vertical or AI for that vertical. And the question is, why wouldn't someone just use anthropic for this? Like, why don't they just go use Claude and co work to solve this problem? Um, at the same time. Well, for generations we've used different companies as kind of the, the bad guy of.
Will they just do this? It was Microsoft. It was certainly Google for a long time. Why won't Google build this? Couldn't Google just hackathon this in a weekend?
00:41:12.550 — 00:41:14.710 · Jason Saltzman
It's probably still, uh. Google. Yeah.
00:41:15.150 — 00:41:26.150 · Sam Andersen
Yeah. Well, now it feels like anthropic. Why won't anthropic just go build this themselves? Yeah, we've seen what they've done in legal. We see what they do in finance, you see. I mean they've they've taken some security, right.
00:41:26.190 — 00:41:45.870 · Jason Saltzman
You know, you want to watch all of the what I'm going to call hottest sectors, whether they're vertical or now more horizontal. And anthropic does start like their one press release away from really, really messing up, uh, a bunch of companies at least, you know, for a 72 hour window until everything recalibrate.
00:41:45.910 — 00:43:09.250 · Sam Andersen
Yeah. No, absolutely. And I think what we see today is that there are truly companies who I think, see the ease of building with AI as an excuse to not have a technologically rich company or a technology rich idea. And they're saying, hey, I can now vibe code this. Maybe they themselves are not technical in their background.
And so rather than thinking, hey, I need a technical co-founder who truly understands the underpinnings of what I'm building, they just build it themselves on lovable or any other platform. And ultimately those are not defensible for lots of reasons. They're not defensible, but they invite way too much competition.
Obviously, over time, it can all get found out. What I will say this is doing, though, is is putting a lot of new pressure on product managers to really understand the workflow and the needs that their users, their customers have. And I think it's making them ask sincere questions of the customer and of themselves.
What would this workflow look like in some idealized future state where AI can take the burden away? What does the practitioner actually want? I use my dad's orthodontic practice as an example. A lot to think that my dad is ever going to sit down with with Claude and Vibe Code himself, a charting software is laughable.
He's never going to do it.
00:43:09.290 — 00:43:16.690 · Jason Saltzman
Well, I mean, you start with all of the regulatory issues there. The HIPAA, the things that deal with PII,
00:43:18.530 — 00:43:20.650 · Jason Saltzman
all the rest of the workflows that plugs into.
00:43:20.970 — 00:43:22.170 · Sam Andersen
He's just not going to do it.
00:43:22.170 — 00:43:22.490 · Jason Saltzman
Yeah.
00:43:22.530 — 00:44:00.390 · Sam Andersen
And yet what this is doing is it's going to invite a whole new generation of founders who can understand the orthodontic workflow better than we've seen before, to go out and create at a much lower cost. And they can invest the time and the effort and the money into building software. That's just better for an orthodontist, and my dad will continue to pay somebody else a hefty sum of money every year to produce that kind of leverage for his practice.
Whether it makes him his staff, his, you know, real estate, his other costs, more efficient to produce more profits for him over time.
00:44:00.390 — 00:45:03.480 · Jason Saltzman
Yeah. And I mean, I think this gets back, you know, this tees off two questions, right? One is like the moat question. Right. Like what remains defensible, what remains defensible over time. And the other is the question of like what's the true difference between internal building, whether that's a single person or at the enterprise level, and paying for good tools that someone else, right.
Is there day job to make sure this thing works? And I always go to the look like I'm sure there are plenty of enterprises that we got. This 95% of AI pilots fail study from MIT out of, because they just spent 5200 K a year engineers, so $1 million trying to go and build and maintain this internal piece of software, as opposed to going and paying some AI company that's going to give you forward deployed engineers, that's going to give you support, that's going to really dial in on trying to make sure that you're getting value in ROI out of this and are going to keep pushing features faster and faster and are going to get learnings from you and all of the people around you about what.
00:45:03.480 — 00:45:05.160 · Sam Andersen
Would be subsidized by venture capitalists.
00:45:05.160 — 00:45:13.720 · Jason Saltzman
And be subsidized by venture capitalists to keep price a little bit lower. Like, that's a way better spend of $1 million in my mind for most use cases.
00:45:13.720 — 00:45:14.200 · Sam Andersen
Totally.
00:45:14.240 — 00:45:54.890 · Jason Saltzman
Now you can get to the places where that fails, right? You know, you go and look at like the big investment firms and yeah, a lot of them are going to want to build internal tooling because they don't want anything else touching. Yeah. Their systems because they, you know, that's where the alpha is. But you, you know, you think about most of these like your hospital system is going to go by whatever AI, uh, EHR system.
Totally. Instead of hiring all their engineers, all these engineers do like build, maintain, grow, adapt all of these systems over time. You're like. It's much easier to get pulled in, to pay, to get pulled into the future as opposed to like, reimagine a lot of your business, at least around pushing yourself into the future.
00:45:54.930 — 00:47:01.700 · Sam Andersen
It's interesting. Probably not a perfect analogy, but I was at the the MoMA in New York City with my mom years ago, and she joked about the Jackson Pollock that was on display. That she could do that. She's like, I can drip paint from a can onto a canvas. And I said to her, I said, but you didn't. You know, like at the end of the day, you didn't do it.
You didn't come up with the idea. And there is something really powerful about the individual that understands and can see that future and has the idea. It's why we see the power of founders as they stay in and or leave their businesses, and the impact that it has on the company that they're building there.
The idea is, in and of itself, valuable and the person that truly captures that idea. Can communicate that idea to other people, can sell that idea to customers, to investors, to team, etc. is valuable. I mean, it's why I think we as venture capitalists are perfectly comfortable that the founders that succeed walk away with billions because we know that they they created the value at the end of the day.
They should, in our opinion, make that kind of money.
00:47:01.740 — 00:47:02.420 · Jason Saltzman
100%.
00:47:02.420 — 00:47:03.020 · Sam Andersen
Yeah.
00:47:04.260 — 00:47:30.060 · Sam Andersen
Okay. Half the industry is losing ground. Question number seven. Half of firms worked harder in 2025 and got worse results. The other half work harder and got better results. Same effort, opposite outcomes. The data that we're showing on the screen from affinity is from private equity. But this pattern applies to anyone chasing deals right now.
What separates the two halves?
00:47:31.180 — 00:49:12.370 · Jason Saltzman
Uh, that's a hard question. Thinking about it in terms of what separates the two halves, I think it's going to be differentiated. Access and understanding. Uh, and when you think about the access side. Right. You know, that's really on the sourcing. That's on back to the relationship, uh, conversation that we were having earlier.
And on the understanding side, that's access to either sort of like proprietary intelligence in one way, shape or form, right? Whether that's from data providers, whether that's from communities and access to sort of like the qualitative or Annick data that you get from talking to people. Right? You know, I think the the leverage right now, the alpha right now is to be found in sort of who you know and what you can know.
Uh, and we're seeing sort of this differentiated shake out where it's maybe not just a financial modeling game. Yeah. Right now, uh, although we'll see how that shakes out in when we get to, you know, fund lifecycles, uh, a handful of years from now. But I think, you know, when we think about what is like same effort, opposite outcomes, the Outcomes or the differentiated outcomes are coming from sort of whatever is that unique perspective?
Uh, whether it's the people, you know, the data you have access to, or the community you've built and operated around, or even like internal fund or firm understanding of a given problem space, right? You see a ton of firms that are these outperforming firms, uh, hiring specific people with domain expertise in AI, in finance, in healthcare, in pick your different sector that they really want to go and target that fund or that firm around.
00:49:12.410 — 00:49:26.050 · Sam Andersen
Yeah. Affinity just publishes data. I think if memory serves, they chatted with 291 different private equity funds and saw the data. And ultimately what I think sits behind this is
00:49:27.290 — 00:51:08.550 · Sam Andersen
that focus on strategy. I think we would find that the people who are sending fewer emails, but getting the better results at the end of the day knew who they were sending the emails to. It was it was more targeted. It was more direct. It was more thoughtful. It was a function of a relationship as opposed to just a shotgun approach in M&A.
I've been involved in a few successful M&A deals over my life, and we often talk that, you know, the Destiny buyer before you even put yourself on the market, you know, the 2 or 3 companies that are very likely going to be the ones that actually step up and bid on the deal. That said, you're going to go pitch to 20 or 25 because you don't want to miss somebody who might just come out of left field.
And you certainly don't want to be accused of laziness and the deal not going through because you weren't willing to work for it. That said, 90% of the time it's one of the ones you thought. And the reason for that is because you know where your company is going to fit best, and you know which of the other industry players is likely going to receive the most value from what you're doing.
It's probably a company that you have known and worked with. It's probably a company whose leadership already knows each other. There's an interpersonal relationship there that really matters, and it's a company that you can go to without 50 emails trying to break through the spam filter. It's someone you can probably shoot a text message to because you've already got them in your contact book.
And so when I think of these firms, what separates those that work and those that don't, it's probably those that are taking a much more strategic approach, a much more relationship based attack. Yeah. And I think.
00:51:08.790 — 00:51:50.680 · Jason Saltzman
We're seeing that from the data that we have at CB, insights on how our customers and clients are using our data to do do that initial strategic filtering, right. You know, the hey, who's going to be the best strategic fit, right, either, as you know, on either the buy side or the sell side of those deals, whether it's M&A, investment or partnerships.
Right. It's the companies that take the time to do that homework to understand the playing field, to do everything from sourcing to competitive intelligence or diligence and sort of understanding of the market landscape, your in-depth before they even go out. Right. You know, that's that sort of strategic view of your next move is a lot of what we're seeing, uh, make the difference between that 5149 split.
00:51:50.720 — 00:51:51.640 · Sam Andersen
Yeah, that makes sense.
00:51:51.680 — 00:52:06.960 · Jason Saltzman
Two years ago the firm's now declining. Were producing two times more deal access per person. By early 2025, they'd been overtaken by firms that converted more efficiently. If you're a GP at any firm PE or VC, how would you even know if this was happening to you?
00:52:06.960 — 00:52:25.080 · Sam Andersen
It may. It makes sense to me simultaneously. Venture or private equity is not necessarily a game of volume. Like at the end of the day, the goal is not how many deals can I do this here? At least at the firm level. Maybe some individual investors feel that way, but well.
00:52:25.440 — 00:52:31.850 · Jason Saltzman
It's. How many good deals can I do this year under the right circumstances. Yeah. With the right, uh, right price structure?
00:52:31.890 — 00:52:32.410 · Sam Andersen
Yeah, totally.
00:52:32.930 — 00:52:49.049 · Jason Saltzman
And so, I mean, I think a lot of this goes back to the conversation we were just having on like, hey, how strategic are you being about that sourcing process? And I think that's probably where the flip happened with this affinity data is, look, the firms that are now taking
00:52:50.090 — 00:53:01.090 · Jason Saltzman
charge of the top of the, you know, whatever this metric is on deal access, are the ones that just have better data and sourcing criteria or better relationships to filter through.
00:53:01.130 — 00:53:44.030 · Sam Andersen
Yeah. And I think it is relationships. I mean, you see the addition of Jack Altman to the Benchmark team and why it's relationships. You know, like they believe that he brings a proprietary set of relationships, either from his Alt Capital time or because of some connections that he might have, uh, to AI startups that bring value to the firm.
I don't think it has anything to do with his ability to fundraise. I don't think benchmark has any problem with that. I don't think it has anything to do with his ability to get a call back. I think, you know, Chetan and the others can get their phone calls answered just fine. I think it has to do with the fact that he comes with a pre-existing set of relationships that are valuable to the firm.
00:53:44.070 — 00:54:31.280 · Jason Saltzman
Yeah, and I mean, I think you see this for almost all of the firms now, especially as you see personnel changes or swaps across these firms is right. Most of them are people that are have some tie to the current like sort of swath or community of operators and whether that's, you know, ties to the large hyperscalers and their VC startup communities or whether it's ties to, you know, they look like they were running teams at these mega companies, whether they were the mega public or mega private companies.
And they've now left. Right. They know the people that they were working with coming behind them. And is that game that really sort of has become the difference. It's like this is again, I'm going to use the analogy of this is like go to market and anything else. Yeah, right.
00:54:32.880 — 00:55:11.320 · Jason Saltzman
Cold email spam cannon. Yeah, right. You know, the ability to like, hey, shoot a text to what you were saying earlier is the differentiator, the ability to rate. You know, one of the things we see in CB insights data is there's more and more deals being done. Like if you start everything with like there's a incubation incubator or accelerator, there's a partnership, there's an investment.
And then maybe down the line there's M&A, the sort of like funnel rate across those four stages. Right. And it's not always linear. It's not always clean. Sometimes you jump between a few of them. But that funnel is getting tighter. Yeah. And so you're seeing that these relationships are being built
00:55:12.920 — 00:55:23.360 · Jason Saltzman
longer with more involvement across the process of company building and more sort of like joint access to the innovation and the growth trajectory of these companies.
00:55:23.360 — 00:55:47.249 · Sam Andersen
And, we're talking about this as if the majority of deals for these funds come through outbound like, yeah, you know, search. That's not true. The vast majority of deals for these companies are coming through inbound you know, interest. And so it's really who are the founders reaching out to. Who do they want on their boards.
Who do they want writing the checks. And so
00:55:48.930 — 00:56:09.210 · Sam Andersen
even for the funds that are out there shooting emails, that might actually be a fairly bad sign. You know, there's a famous phrase, I think it was Groucho Marx. We should fact check that. But who said, I don't want to be a part of any club that wants me? And I think to a certain extent, VCs are kind of the same way.
It's the founders want to choose their VCs before the VCs are choosing the founder.
00:56:09.210 — 00:56:40.540 · Jason Saltzman
And I think maybe right, if we take this question to the part of the affinity data that you'll know better, or you know the part of this data that you'll know the world better than I do, which is the PE world. That's probably where you get more of this outbound that's happening and you start to look at, you know, sort of, hey, if we're going and trying to find these, you know, sort of financial acquisitions, uh, rather than sort of strategic acquisitions or investments on the VC side, you know, sort of what does that look like?
And I think that probably has more of an outbound component.
00:56:40.580 — 00:57:53.600 · Sam Andersen
Yeah, I think you're probably right. But at that stage of business, you get many more players involved than a single founder or set of founders. You now have boards. You have large teams of investors who have backed this company, who are then doing the outreach themselves. So, you know, you maybe have had a position in a company for ten years, you'd now like it sold and you'd think, hey, this would fit great inside of workday.
And it just so happens, you know, a bunch of people who work for workday. So now your investor team, your board of directors goes to work to find those investors and their relationships play into the relationships that are required to get a deal done. And so I think as a business matures, so does the web of relationships that get spun out around a company and the importance of tracking that, managing it, ensuring that it's healthy and growing, and that it can be called upon when in need for a deal or for an investment or whatever else.
Yeah. 100% conviction versus FOMO. CB insights has mortality data. Affinity has velocity data. Our firm's moving faster on AI because of conviction. Or is this a repeat of the 2021 FOMO cycle?
00:57:54.840 — 00:58:46.690 · Jason Saltzman
I think when you look at the affinity data that shows that firms maintained this, like higher outreach volume for the last two years, but saw conversion drop. They weren't slowing down, they just weren't converting as much. And when we look at really, I think it's a little bit early to talk about the mortality data for this wave of AI companies.
Everyone can have their bets. We can all call it a bubble. Power laws maintain true even in a bubble. right? We'll have the winners. We'll have the losers. I think the this goes back to something we talked about earlier, which is just like what's generally the difference between 2021 and now, and what's the promise of AI and the promise of like moving from this operator centric version of venture inversion of startups to this founder visionary centric, uh, or like if everything comes in waves and comes in cycles, we're back to that.
Uh,
00:58:47.970 — 00:59:14.250 · Jason Saltzman
I think that's probably the big difference in terms of really? Yeah, we're going to see a lot of that, like dreams die all the time. Uh, unfortunately not to be, uh, negative about, you know, or pessimistic about that view, but that will always be the case, right? Startups will always fail, and we'll always find winners.
I just think the winners and will look different now than they did previously.
00:59:14.290 — 01:00:07.300 · Sam Andersen
Yeah, I completely agree. I think firms are moving quicker, not because they have, frankly, more conviction. I think they're moving quicker simply because the velocity of the industry is moving quicker. Founders need to move quicker. There's pace is a natural advantage and I think they want to capitalize on that.
In a world where they feel uncertain about all the other advantages, pace is one that they know they can get. And so they're they're pushing. You know, we hear a lot about what these AI companies are pushing their teams to accomplish. You know, the nine, nine, six famous for what these AI companies expect of their employees.
I think similarly, VC funds are now saying, hey, we got to push our founders faster. We got to encourage them to continue to move quicker because everybody feels the acceleration.
01:00:07.340 — 01:01:13.800 · Jason Saltzman
I also think right when you look at the data on sort of like deal progression and step ups or progression between rounds, right? You see, for the hottest companies, that is also accelerating. And so the question is sort of like what's different and what's accelerating those deals versus what was accelerating deals in 2021?
And it's interesting because I think you get a few things that are similar, right. Like the push on go to market is similar, right. You know, the the winning distribution and winning sort of market share is similar if not more competitive and faster. Now, I think the ways that people are doing it are slightly different.
And then the other one is just like talent compute. And I think the combination of those three are what generates speed, right? You can shift faster if you have more compute, you can ship faster. If you have better talent, you can go to market faster. If you have a stronger go to market motion. And all of that requires fuel, right?
Like capital fuel into the engine to go and spend money on. Yeah. And so that's probably what's going to end up in this like what's different conversation.
01:01:13.840 — 01:01:15.040 · Sam Andersen
Yeah I agree.
01:01:15.520 — 01:01:26.360 · Jason Saltzman
Uh, okay. If that's what's different, uh, how do you underwrite AI without 2021 logic? Yeah. And does the process look different at all?
01:01:26.400 — 01:02:55.460 · Sam Andersen
Yeah. I think, you know, what is lucky about a pre-seed or a seed stage investor is that I don't have to know where it ends up, as long as it ends up somewhere big. Um, and so all I have to think about is, is this an idea that I think is going to add value to somebodies life or work or company in such a way that they'll pay money for it?
Um, when I look at what these growth investors are having to wade through today, and whether that business is generating sufficient IRR to turn into future profitable cash flows at X marks, I don't envy them. I don't know how anybody really knows. And they're placing much larger bets. I think what we're seeing in seed and Pre-seed stage is that people are placing a lot of bets and I think they believe, hey, the power law will still apply if I'm picking good founders and we're placing strong bets, 10% of them will turn into great massive businesses and that will make my fund, you know, we'll see if it turns out to be true.
To your point, like the data doesn't exist yet, but I think we'll see a lot of winners emerge. And I think we'll look back on this and some great funds will come because of that uncertainty feels like there was an era where everyone kind of was in the middle, because there really wasn't a lot of differentiation between a great company and a fine company, whereas now I feel like we'll see more separation.
01:02:55.460 — 01:03:08.980 · Jason Saltzman
Yeah, I mean, I think even if you look at it right, like 2021 was five years ago. Yeah, right. Like the reality is for most of the funds that are 2020 or 2021 vintage funds, we're not going to know the outcome like the true outcomes in.
01:03:08.980 — 01:03:10.470 · Sam Andersen
6 or 7 more years? Yes.
01:03:10.510 — 01:03:22.709 · Jason Saltzman
For. For a long time still. Right. What we're seeing now as the outcomes are 2015 vintage. Yeah. Or even earlier still. And there are
01:03:24.070 — 01:03:52.710 · Jason Saltzman
like it's almost worth talking about like the exit market and liquidity in a sense of like that's what we're seeing. Exit and provide returns now. I just don't think that like the market for those companies is currently great, if you're right. And so like you're trying to project ten plus years into the future of yeah, what's the like growth trajectory and then the exit pathways for all of these companies you're backing at seed or a now.
Yeah. If
01:03:53.790 — 01:03:55.190 · Jason Saltzman
it's ten years until we get there.
01:03:55.190 — 01:04:05.990 · Sam Andersen
Totally. It does beg the question what vintage did OpenAI started. When was it true like pre-seed or seed funding.
01:04:06.710 — 01:04:08.960 · Jason Saltzman
I mean, do we call it coming out OIC.
01:04:09.040 — 01:04:09.560 · Sam Andersen
Yeah.
01:04:09.920 — 01:04:11.360 · Jason Saltzman
Like do we call it 2016.
01:04:11.360 — 01:04:23.520 · Sam Andersen
And that's what I'm saying is like that vintage. I mean I think the 2016 vintage has looked good for a while. Yeah. But I think we'll obviously look a lot better when you tack $1 trillion onto the hopeful DPI at the end of the day.
01:04:23.560 — 01:04:48.280 · Jason Saltzman
Well, and I mean, it's not just open AI. Yeah, right. You know, you look at the other companies at the top of the private market valuation charts right now, and, you know, you have stripes and data bricks that are, you know, maybe they're not 2016 specifically, but they're like adjacent. Yeah. In what I'll call that.
Like I'm gonna call it pre era right. You know the era of 19 or 20 through 2022.
01:04:49.600 — 01:04:52.000 · Jason Saltzman
Yeah. Maybe like who knows. Uh.
01:04:52.720 — 01:05:03.520 · Sam Andersen
But I mean that's a lot of value that I think, you know, none of us frankly, saw until the ChatGPT revolution came along.
01:05:03.560 — 01:05:05.480 · Jason Saltzman
Until late 20, 22 or 3, 20, 20.
01:05:05.520 — 01:05:09.250 · Sam Andersen
Five years, six years into the vintage. Yeah, that's the point.
01:05:09.290 — 01:05:49.810 · Jason Saltzman
And I mean, I think I have this, like, you know, if I was trying to decide this in a single line, right, you could stop all model development right now. And we have ten plus years of figuring out what to do with it. Yeah. And that's the phases of which of company building and growth that will determine the returns from the vintages of companies that were, you know, like this AI vintage.
Yeah. And the I mean, almost the problem is that we're not gonna start model development. We're not going to stop the agent wave. And what comes next after the Atlantic wave. We're not going to stop the development of world models and physical AI and all of these things. And that's just going to keep making the playing field.
01:05:49.850 — 01:05:50.450 · Sam Andersen
Yeah.
01:05:50.930 — 01:05:56.170 · Jason Saltzman
Like the play like the playing field is moving faster than most companies even on the field are.
01:05:56.210 — 01:06:24.860 · Sam Andersen
And I think if we were to chat with some of those CEOs, they probably would wish that the development would just stop. Yeah. It would produce the stability they feel they need to build a business. Yeah. Yeah. Um, the most efficient firms scaled their outreach 12 x over two years while maintaining steady conversion.
They tested whether output held at each level before scaling further. Are the best AI investors doing the same, or is everyone just taking more meetings?
01:06:26.300 — 01:06:26.740 · Jason Saltzman
Uh,
01:06:28.540 — 01:06:58.300 · Jason Saltzman
I'll answer that with data and then turn it around to you from the investing seat, because you can tell me how full your calendar is. Uh, no, that's not run by Clockwise. Yeah. Uh, the yeah, the anecdotal, I would say. Yes. I mean, like, I, the investors I know are busier than ever. Yeah, right. They're talking to more people.
They're talking to more people more quickly. I know people had heard a story last week where someone took a meeting, uh, at 7 a.m. and had another meeting by noon. Uh, same founder. Same team like that. That was the pace at which this investor was.
01:06:58.340 — 01:07:00.060 · Sam Andersen
Yeah. I need to speed up. Yeah.
01:07:00.100 — 01:07:10.680 · Jason Saltzman
Uh, so maybe I'll ask you how fast, like, how how how much faster do you feel like you're moving? Or how much faster do you feel like? Or how many more founders do you feel like you have to talk to in any given period of time?
01:07:10.720 — 01:07:56.240 · Sam Andersen
Yeah, I mean, we are taking a lot of meetings, and I think that the energy in the startup ecosystem is really palpable. I think in order for someone to take the risk to become a founder, they need to believe that there's a path and an outcome for them to build a great business. I think that's step number one, and I think AI has opened that door in that vision for a lot of people who want to step in and say, hey, let me take a crack at this.
We're seeing a lot of businesses. I think simultaneously we're saying no quicker than we've said before, um, to a lot more businesses simply because we have to we have to be like, discriminate in a very quick way, or else we'll get bogged down in diligence for deals that we're never going to do.
01:07:56.360 — 01:08:02.970 · Jason Saltzman
I think the volume is higher, uh, and like it's the same split of yes no's. Or do you think it's just.
01:08:02.970 — 01:09:04.620 · Sam Andersen
Like we're getting pickier? All right. Yeah, I think we've we increased the bar. Now, part of that is because we have a deployment pacing that we're going to stick to, and we're going to deploy the amount of capital that we plan to deploy this year. Um, we're not going to deploy our funds, you know, quicker.
I think if VCs maybe learned a lesson from 2021, it would be that pacing schedules do matter to LPs, and they should try and stick to them. Um, but no, we're we're just saying no more. And that's not a bad thing. Honestly, I think, uh, another era of 2021 is that a lot of companies that never had the unit economics, that never had the viable plan, that didn't have the capacity to build a large team, got a lot of money to go and do that, and it never materialized.
And you took the life and careers of some very talented team members on a 3 or 4 year detour, until they figured out that those companies weren't actually viable. And if someone had spent a little more time in the model early on looking at the unit economics, that could have prevented a lot of that.
01:09:04.819 — 01:09:10.140 · Jason Saltzman
Um, uh, how do you think about AI unit economics or AI company unit economics now?
01:09:10.180 — 01:09:59.149 · Sam Andersen
I think a lot of investors, myself included, uh, are believing that there is an experience curve to this that will bring the unit economics into alignment over time. That said, I think everyone is still very wary of burn ratios. Uh, because we learned again in 2021 that the market does value a quick pivot to profitability.
Every once in a while. And so you need to be in a position where that can happen. I still think we hear founders. We still hear founders who are going for series A and B rounds. Talk to us about how quickly they could be profitable. That is still very much a talking point of founders and something we coach our founders to talk about as well, because we know funds don't want to feel like
01:10:00.510 — 01:10:45.550 · Sam Andersen
their investment is 12 months away from death. Assuming some other investor isn't willing to pick up the back. Yeah. Um, yeah. It is interesting to see which models are allowing AI companies to pass the cost of the AI, like the inference itself on to their customers versus those that the business models require them to subsidize that.
And I don't know. We'll see. You know, there's there's a lot of talk about the SaaS model and like, is it going to be dead in this world of agents? Can we not charge by seat anymore? And I don't know, I'm waiting for some of the behemoths to come out and tell me they're not going to start charging by seat. But I think in the near term that's going to stay the same.
01:10:45.550 — 01:10:48.710 · Jason Saltzman
And so I pay for a seat to ChatGPT and Claude.
01:10:48.750 — 01:11:24.530 · Sam Andersen
Yeah, absolutely. And so then there's some utilization threshold that sits on top of that. And I bet you we see some combination of that in the future. We are a record multiples traditional return drivers of increasing multiples or adding more leverage are pretty much gone in the private equity space.
So whether you're a PE firm paying almost 12 times EBITDA, or VC fund writing AI checks at inflated revenue based valuations, the only remaining edge is sourcing discipline. Does this conversion efficiency gap of 17 times that we've discussed become a return predictor?
01:11:25.210 — 01:12:15.540 · Jason Saltzman
No. Uh, and it's no. And I think this goes back to everything we've been talking about earlier. Right. It's the sourcing that matters. It's the ability to take those bets. Understand that power laws are going to be the thing that ultimately determines the success of any given fund. And do you have the right companies in that fund?
Uh, to have the 10% of them hit it out of the park? Yeah. And maybe within that, the 17% gap is still the question of like, look how what's within that 70%? Are you seeing a lot of companies and not winning a lot, or are you seeing a lot of companies and being very choosy? Yeah. And what's the difference between those two?
Because I think at the end of the day, you can't win a deal that you don't see. Yeah. Uh, and so.
01:12:15.580 — 01:12:17.900 · Sam Andersen
And some of the deals, you see, you frankly don't want to win.
01:12:17.940 — 01:12:18.180 · Jason Saltzman
Yeah.
01:12:18.220 — 01:13:49.549 · Sam Andersen
Of course, you know. Yeah. It's interesting when we think of of this. It's why all these metrics are really tricky. A lot of VC funds internally have these guardrails to protect the fund from the junior investors who are just looking to get deals done, looking to make a mark, looking to build a resume as a junior investor.
Very rarely are you punished, quote unquote, for a bad deal, but you're pretty heavily rewarded for a really good deal. And so the incentive structure is all around doing lots of deals. I think what we'll see as this vintage plays out is that the funds that were selective, the funds that had the options and were deliberate, in which options they chose to invest in.
Well, obviously you see the best returns. I think what you'll find is there will be funds with the spray and frame model who took all the options that they had. We'll still have the same flat returns they've had in a long time, and there will be funds who bet the farm and lost. Yeah, you know, they they played a concentrated game, which is the key to success in getting large multiple funds.
And they place the wrong bets. And then of course you'll have the decile of funds that outperform. I think it's pretty normal. Um, yeah. It's it's interesting to think about how these funds are actually going to shake out at the end of the day, because we were talking earlier about these massive funds, the insides of the world, and
01:13:51.280 — 01:13:58.520 · Sam Andersen
Think how much it takes to fundamentally move the needle on like a $4 billion fund. You know, like it just.
01:13:58.560 — 01:14:10.760 · Jason Saltzman
Well, especially when you think about ownership stake in any given company and the fact that, you know, even if you go and look at it, we'll take Wiz as the example. Right. A bunch of a bunch of firms made a ton of money off of Wiz. Yeah.
01:14:10.880 — 01:14:15.480 · Sam Andersen
Shout out to Arsham who was at Lightspeed is now at Meritech. Yeah.
01:14:16.040 — 01:14:45.920 · Jason Saltzman
But there's also a bunch of funds that, like, didn't have a fund return out of ways. Right. In the same. You'll see it's like a lot of this is going to become position dependent in any given company of did that return the fund? Yeah. Or did the liquidity event return the fund? And I think when you think about these firms having a few percent in these, you know, deck of corn plus outcomes, if the fund is big enough, it's just not it doesn't cover it.
01:14:45.960 — 01:14:46.800 · Sam Andersen
Yeah. Yeah.
01:14:46.840 — 01:15:20.570 · Jason Saltzman
And I mean, this maybe actually gets to the earlier conversation we were having on sort of continuation funds or funds that pull certain companies that are the winners or that are being king made or whatever you want to call it, across enough funds to have different risk profiles that are underwritten is that, you know, maybe that's the model for this.
And that's the difference is like once you once you source the winner, once you have conviction, once you've then sort of decided that, like, hey, that's a company we're all in on. You have to pull it all the way forward.
01:15:20.610 — 01:15:36.210 · Sam Andersen
Yeah, yeah, that makes sense. Um, you know, some of these mega deals, I mean, what, $32 billion for Wiz to think that it doesn't return a vehicle? Absurd is crazy. Yeah. You know. Yeah. Um,
01:15:37.330 — 01:16:04.910 · Sam Andersen
so, just as a side conversation, um, I was on a board with Arsham at Spiff, and I remember when he did that deal, we. This was prior to the acquisition by Salesforce. If I remember correctly, he was at Lightspeed and I remember chatting with him about it because I want to say he did it at $1 billion, something like that.
And I remember saying for Wiz and Arsham said, you wait, so awesome. You were right. You were right. Yeah. Um, okay.
01:16:04.990 — 01:16:10.550 · Jason Saltzman
If I've learned almost anything during their entire funds, uh, dedicated to this thesis,
01:16:12.070 — 01:16:17.550 · Jason Saltzman
you probably don't want to bet against people coming out of Unit 8200 or Talpiot. Yeah. Um.
01:16:18.710 — 01:16:29.710 · Sam Andersen
Yeah. No. And Arsham is, uh, phenomenal investor. So we are lucky to have on our board. I'm lucky to call him a friend. So incredible. Um, okay.
01:16:29.710 — 01:16:38.510 · Jason Saltzman
Okay. One final question to wrap this up. Yeah. Uh, Sam, five years from now, does the 2026 AI vintage of companies look like 2021?
01:16:38.550 — 01:16:52.720 · Sam Andersen
No, I think 2026 will be a different vintage than 2021 from a results perspective. The technology innovation of AI is going to fundamentally unlock better results than 2021 saw.
01:16:52.760 — 01:17:15.440 · Jason Saltzman
Yeah, I mean, the way I look at it is right? You know, the what independent of when we see the companies from 2021 finally reach whatever exit or death that they will. The 2026 vintage just has a far greater promise of economic impact, both for the companies themselves and for what they end up doing across how we live our lives.
01:17:15.480 — 01:17:39.760 · Sam Andersen
Thank you for listening to The Cap Table. Thank you to Affinity for the questions and for sponsoring the podcast. We would invite you to encourage you to read the report that we referenced today called The Invisible Edge, now published on Affinity's website, and also check out their YouTube channel at at Affinity Underscore Presents.
They also have a variety of other podcasts and publications across YouTube, Spotify, and wherever else you get your podcasts.