Episode #19: Venture 3.0, Deciphering Early Stage Traction, 90's Tech Bubble vs. AI Valuations

The functional difference between 500k and ARR and a million is nothing relative to the end goal. But I feel bad because founders read all of these blog posts that say you need a million in ARR, you need two and a half million in ARR, you need to double every. Those things will get you attention, but they won't necessarily get you money it's.

The week of February 10th. Welcome to the Learning Corner, a Precursor Ventures podcast.

Welcome to the Learning Corner by Precursor. I'm Mia Farnam and I'm joined by our GP and Managing Partner Charles Hudson, where we will be talking through the articles and Goodreads that caught our team's attention over the past week. Topic number one was written by James from NFX outlining how venture capital3.0 impacts founders in the AI age. The essay argued that venture is undergoing a significant transformation driven by AI software automation and changing dynamics between founders and investors. Amongst a variety of insights, the essay suggests that the version of venture we live In Today, version 3.0, or the version of Ubiquity where venture has explicitly expanded internationally and across a multitude of sectors, which has created a whole host of new dynamics for investors and founders. James called out that VCs face more pressure than ever to compete with other VCs. This creates a system where investors will need to fund more outlier ideas and non consensus deals in order to perform. I agree with this, but I also know that a massive percentage of deals that we see getting done are in AI, which feels at the moment like a consensus deal archetype. So I wanted to hear your perspective on the article.

I enjoyed reading the article. It's always fun to read James's writing. He's always so positive and optimistic. I just thought there were so many things in there that I hadn't thought through the implications of. You know. For example, he talks a lot about how there'll be many more venture capitalists and probably my favorite line in the article is there are low barriers to entry in this business. Anybody with money can become a venture capitalist and that is true. But I think, you know, this idea that we're going to have many more VCs is kind of at odds with a lot of the articles about firm contraction, particularly among emerging managers and many of those firms going away. But I think we have so many one and two person venture firms being started and spin outs from big platforms. Will there be tens of thousands of more VCs in the future? I don't know. But for me the big question is does AI make the venture business easier or harder. We've certainly had companies where their data has been analyzed by some AI part of the process and spit out a number and in some case that number or conclusion's been illuminating and in some cases it's been gobbledygook and it's been contradictory. The AI has said like the company is both great and flawed at the same time. I have so many questions about where we're headed to James's article It does feel like we're on the cusp of a new era in venture. Whether that's super big mega funds, whether that's AI's impact on venture, whether that's just more dollars coming into the asset class in general. But for me the other question I had is he talks about a bunch of new LPs trying to get into the ass class and I think there are things that are probably true for venture that might not be true for firms of all size. And I haven't met quite as many LPs who are both clamoring to get into venture and whose principal entry point is going to be small to medium sized firms. I've certainly met people who are like hey, I haven't been in venture. I'm going to start with the big mega funds. I'm going to go get Sequoia Andreessen founders, one of those as the bedrock of my portfolio and maybe then I'll start to experiment with the stuff that feels more exotic. But you know, I tend to think of this through the lens of not just what does it mean for our industry but what does it mean for firms of different size.

And going back to piece of we're not sure where AI tools will actually end up affecting how people do the day to day job in venture. I do think that there is great opportunities there for founders. Maybe if we apply AI it might cut down on some low hanging fruit questions. But then the tough side about exactly what you said. Maybe you have biases baked into the models misinterpreting basic information. I do think as big AI tool users ourself on the team we found different ways to tinker around to feel comfortable. From your perspective, what are some things that you feel like you really can't optimize with AI in our roles?

I think so much of this job and fairly or unfairly is your subjective view of the individual running the company. Especially at Precede and Seed, the people are such a big component. I was talking to someone this weekend who's like well what about late stage? I'm like the Late stage people are already quite analytical. It's unclear to me what the delta will be in terms of the output that those folks can create today and what AI would enable them to do. And I worry that if we get too algorithmic without the people piece will miss out on special people. And you know, as I've told a lot of our portfolio company founders, even things like traction, it's very rare that for a C stage company, the amount of traction you have is the reason that somebody invests that you go, wow, this traction's so amazing. I don't need to know anything else. No, you need to know about the market and the people. And my hope is that AI makes the computational traction analytical piece simple so that people can spend more time on the people piece. So that's my hope, is that you can get all that stuff out of the way. And if I know anything about founders, if you give them a set of rules, somebody will figure out a way to hack or optimize it. And so I'm also just very curious in this arms race, if VCs start using more AI to analyze companies, that will provoke some response, which I have not thought through, what it will be from founders who will modify what they send us. And I don't mean in a fraudulent way, they will optimize whatever it is that they send us to get a good answer out of the AI. So it'll be interesting to see how.

That all plays out at Precede. Also kind of embedding AI tools is difficult because there's typically no traction whatsoever. So ingesting any type of metrics or traction into a model that you've customized is. I don't, I don't know what the output. I mean, I played around with it, but ultimately it lands out in stuff that's probably like 25% of the way that I could get it if I just chatted with the founder directly. It does open the door for opportunities to support founders within our portfolio and I think specifically within our own team, 400 plus companies. And I know we've talked about, like, what are some ways that maybe we can bubble up insights about, you know, perception of investors when our founders want to go out to raise. Are there ways that we can do this through a model to let them know what people are thinking? And that's a lot. What the article calls out is that AI tools will be used more and more to either find you, engage with you, evaluate you, or support you, which I believe is true.

I've seen a few people who've built these pitch bots on their websites where you can submit your deck and it will tell you this pitch bot is a proxy for us, it's not a decision. And the cool thing about that it's, you know, I get people who send me pitch decks sometimes and like, what do you think? I'm like, I think I'm too busy to give you feedback on this because I'm working on either our existing portfolio, but the idea that someone could get feedback from before they submit, go through a couple of revisions, that to me is interesting. Also I have a bunch of checklists and documents that I've put together for our founders that are kind of pre fundraising things and the idea of folding those into AI is interesting to me because instead of relying on them to actually manually go through and say yes, I can just hey, give me the stuff, we'll put it in there. The bot knows what's on the checklist and the bot will determine whether you've done a good enough job addressing those issues.

Yeah, that's a great summary. I think there's a lot of low hanging fruit for us to make our founders lives easier with the portfolio instead of us kind of reaching out and asking them to fill out a specific set of forms. It's a great segue actually into our second topic. It's a medium essay written by Jared at Rebel Fund who took the time and effort to try and extract data driven insights on what early stage startup traction should mean to investors. While they're diligencing companies with a massive data lake built from investing in hundreds of YC startups. Jared wanted to see if there was any correlation between early traction that you see leading up to those coveted demo days and their latest respective valuation. So it turns out after doing the analysis, Jerry determined that there is weak to negligible correlation between any of the traction metrics that he used and the latest company valuation. Or in other words, the hype around the up into the right revenue graph traction that you see at demo days or just in pitch decks early on wasn't all that important and perhaps it was always the softer stuff. So your team, your market, et cetera, things we just talked about five minutes ago, that should have taken priority. It's hard to ignore traction though, especially when there are such few metrics to point to as early stage investors. So how do we strike this balance of what to actually place weight on?

I've been going to YC demo day since almost the beginning and I feel like to their credit, YC does an excellent job of getting companies ready to fundraise and does an excellent job of helping them distill, you know, what could otherwise be a complex narrative into something you can explain to somebody in a minute and get them excited. That should like not be trivialized. But we also have a lot of YC alums in our portfolio. Yeah. And if I look at the ones that have been incredibly successful, many of them were not the hottest companies on demo day. And many of them YC aside, across all sellers, many of them didn't hit their stride until after. And I always ask myself like what can you really get done in 12 weeks of focus work if you're really going from 0 to 1? Now YC, it's gotten more complicated because not every company that gets into yc, it used to be they were mostly cold start companies. Like they were really starting from scratch. That's not really the case anymore. Also the batches are bigger and the YC network has gotten a lot bigger. And I know for a lot of our B2B companies the ability to sell to other YC alum companies is a big benefit. I don't know. I feel like whenever someone tells me this is what we've done in three weeks we were here on day one, I'm like, but your laser focus has been on driving revenue. What would it have looked like if, if you weren't in demo day? What's not shown on this chart is did you ramp up marketing spend? Where did all that come from? And I always appreciate people who have traction. But to your point, I care a lot about who the people are, the market they're in. And after we walk away from demo day, I gotta wake up the next day and be like what did I just do? What did I just put money into? And so I go into this assuming that everybody's going to show me a chart that's up and to the right because almost everybody will show you a chart.

Why would you not?

Why you not? You can take the labels off the act. Like there are a lot of things you can do. And then every now and then you do see someone who has such extraordinary attraction at demo day that you do pay attention, go, wow, that feels like more meaningful. And I get it. It's a, it's another lens through which to view the companies in the batch. And I enjoyed this article because I've always felt like demo day heat and revenue traction is an input but shouldn't be the thing that guides your decision.

Yeah, it's a beautiful thing about precede and seed venture funds. And that, at least in the way that I feel like we think about it, we're very unapologetic and not disillusioned by the fact that there is probably to no traction at the earliest stages. And so to your earlier point to place a lot of weight on it could almost just signal the wrong things. Once you're almost at a seed Series A's fund, when there's a little bit of traction, you just want to generate as many insights as possible on it because you have something to base those judgments off of. And I find that oftentimes it just makes you swirl and maybe create conclusions that have nothing really to do with where the company is. Because even from pre C to Series A, like you still have so much left to go. And we've seen within our own portfolio, people pivot post Series A because they have to. So it's funny, I do find investors getting hung up and wanting to kind of push traction in metrics, the things that got them over the line or when they're trying to sell me on the deal.

I always tell our portfolio companies two things. One is traction, in my experience is mostly confirmation bias. Like once you've decided that you like the company, if they have no traction, then you're like, well, traction doesn't matter. They just got started. If you don't like the company, you're like, I wish they had more traction. It's just confirmation bias. And I think the best VCs I know understand what their partnership needs to see and what make sure that the traction level the company's achieved will satisfy whatever their internal process is. But people are like, well, if you get to 1 million an ARR, I'm like, well, if you get to 1 million ARR, it doesn't mean you're going to raise money. It means that a lack of traction is probably not the reason someone will say no. Those are two really different things. And the second thing I tell people is most VCs are underwriting companies that will eventually get to, let's call it, 100, 250 million in revenue. The functional difference between 500k and ARR and a million is nothing. It still rounds to basically zero relative to the end goal. But I feel bad because founders read all of these blog posts that say you need a million in ARR, you need two and a half million in ARR, you need to double every. Those things will get you attention, but they won't necessarily get you money. And so what I tell founders all the time is you are in the zone where it's unlikely that traction will be the immediate disqualifier for your business. It doesn't mean people are going to think you're a good founder. It doesn't mean they're going to like your market, doesn't mean they're going to think your company's the winner. It just eliminates that as the easy pass reason.

It's almost what I think about when folks at the earliest stages ask or work so hard on their deck and you're like, it's definitely never going to win you the investment, but it will for sure, maybe make that conversation or solidify a call. But by no means are they going to use that as a way of like, look how good this deck was. Let's push it all the way through.

IC not to make this like a YC fanboy session. I really do think they have solved the problem that I have struggled with, which is oftentimes I get first drafts of pitch decks from founders and I'm like, I understand what you want to talk about, but you don't understand what I want to hear. And I spend a lot of time with our portfolio companies trying to get them to translate the story that they're trying to tell or they want to tell into a story that will land with investors. And it's very rare that I get a deck from a YC alum or a current YC company where you have to do significant work on that. Like, they. They seem to be able to teach people, whether it's cynical or not, this is what investors want to hear. This is how they want to see that information presented. And I generally find it works. At least it works for me.

It's half the battle. It's basically just the storytelling exercise. The one last I will say is that it's great to tell founders all of this in advance, to just brace themselves for the investor perception at the earliest stages. But also I don't want to discount the excitement at the earliest stages when they have seen some traction.

I don't want to tell people, traction doesn't matter. I don't want to tell them, you shouldn't try to get it. I just think it's important to contextualize the traction you have. Like, at this point, if you're a Series A investor, a million in ARR is not impressive. It doesn't mean it's not a huge accomplishment for you as a founder. It just means relative to the things that hit that person's desk, a million in ARR does not help you stand out in the ARR our quartile, it just doesn't. And those two things can be true at the same time. And so what I'll tell people is like, it's a, it's really impressive what you've done, but relative to other people, it's not maybe as impressive as you think.

Well, our last article for the day was an essay by Paul Krugman where he compared the price bubbles days of tech stocks in the late 1990s to that of the AI frenzy we are currently living in today. It's titled have We Been partying like it's 1999. He published in the New York Times in 2000 actually a piece titled the Ponzi Paradigm, which argued that bubbles in which asset prices lose touch with reality are common and deeply rooted in human psychology. The rise in tech stocks and in our current day scenario, AI company valuations create bubbles that almost mirror Ponzi scheme behavior where, quote, those who get in early make money not from underlying asset return, but from rising prices driven by later entrants. When folks basically pile in haphazardly to join the party. For listeners, you definitely have to read the article because it's just, it's a really great read. But Paul obviously doesn't want to just dismiss the economic advancements that AI will provide altogether similar to the late 90s with the Internet, he thinks that there will certainly be a boost in productivity, but perhaps not in this enduring growth rate that folks are underwriting too. So this raises this interesting question. If AI is more about efficiency gains rather than unlocking entirely new markets, shouldn't that change how we should be thinking about its impact on startups and venture returns?

Yeah, well, I've like, I read this and I was like, I've had some of these conversations with people and I don't know, I. Part of the reason I don't make big predictions is I think predictions are scary and the world changes a lot. And you know, there was a period of time where he thought that the Internet wouldn't have much more impact than the fax machine. He since has admitted, hey, I was wrong about that. I admire anybody who can make public predictions and admit when they've gotten it wrong. I literally told somebody MIA that I was like, it felt like 1999. And that person's like, it's 1996. And I was like, okay. Arguing over windows of time. I was like. And we were both in, we were both around the Valley then. And I said, look, almost everything we thought the Internet would do, it does it just didn't do it in 99 through 2003 because people weren't ready for it. The technology wasn't up to, for a bunch of reasons. It just didn't happen on the schedule that people wanted. And I look at a lot of the AI stuff and it feels the same way to me that like a lot of this stuff is eventually going to work the way people think it will. It might not take 20 years, it might take take 5, 7, 10 years for this stuff to like fully play out. But most startups are not set up to work that way. They need to be in a two to three year technology window. But I think the thing that's interesting about this piece is what's different is in 99 and 2000, if you were a venture capitalist who bought shares in a company, there was no way to get out of that trade for the most part short of IPO or M& A. And part of the reason that cycle was so disastrous is you, it was kind of a long only trade. Like there was no way to. You couldn't go short obviously, but also you, there wasn't a robust secondary market that for the most part would allow you to, to offload some of the companies that weren't working where maybe you realized the valuations were stretched. And this environment's really different. You could get in on a buzzy AI company at seed, ride it to the sea and sell the whole thing. And so I do think the dynamic that's different here, you could be maybe a person who thinks that this stuff is overdone and there's too much enthusiasm, but you could still actually like essentially trade it. Easier said than done, but I think it changes the dynamic when you can get out because you're not playing the same game as the people who have to hold it forever.

What does that mean then for folks who come in later stage who looking into AI?

I think those folks had better believe that they're right. The two funny things to me, I told somebody at a dinner, we were talking about an AI company that's had like a really strong revenue ramp. I'm like, guys, you know revenue can go down. And everyone's like, what do you mean to go? No, no, no. Like it, it could go down. Like the thing they're doing is very important today. It's not forever. We don't know that it's forever important. And we've seen this. Like you, you look at Jasper like Jasper's reinvented itself. But that company had quite the run up and they had to reinvent themselves. And we have quite a few companies like this company's at 100 million ARR. I'm like, well let's look at what are people paying them. We think those are things people will get paid to do three or four years from now. And are those real budgets or experimental budgets? And I say this not to pick on any AI company in particular, but I've certainly seen some use cases. You know, we had HTML website builders in Web 1.0 because there was a period of time where building an HTML website was hard for people and then it just wasn't hard anymore. But we still have website builders, we have Wix and Squarespace. It's just they perform a different function and also it's a different set of companies that are winning that business today than we're winning that business in the beginning. And I don't know, I told somebody at a different dinner I've only seen in my entire venture career to consensus trades where there was almost no dissent. The Internet 1.0 had almost no dissent. Everybody was pretty much in on that trade and that eventually worked out, but there was a lot of pain along the way. And AI crypto had dissent, mobile had dissent. Even cloud SaaS had to send people like no one's going to rent software people. But this is probably the only the second time in my venture career where the resistance to the consensus trade is minimal to non existent.

I do like that this article highlighted in a very data driven way why do we need to build? Like what's the need? Yesterday night I caught up with Lauren from MGV and she said this great quote. We were talking about the same thing and she said rollerblades have been around forever but you don't see me skating around New York City every day, you know, and you're like, it's kind of the same thing. Like why? Why today?

You know me, I think AI is going to be amazing for people who use and buy software and will make the business of selling software different. I don't know if it'll be better or worse, but definitely different.

Well, we are all out of time today. Thank you Charles for stopping by and thank you to our listeners for listening. Please like and subscribe and we will talk to you all next week. Thank you by Sam.