Reshaping the Future: Insights and Strategies for Navigating the 2023-2024 Pre-Seed and Seed VC Landscape | Predict 2024

Hi, I am Jonah Madan. I'm a partner here at Forum Ventures. Uh, forum Ventures is a B2B SaaS studio where we build native AI businesses Accelerator, where we help, uh, B2B SaaS founders go from zero to one and Prese Seed Fund, where we're investing in a hundred B2B SaaS companies a year, uh, across North America.

And I'm here to talk to you today about the state of the fundraising market, as well as some predictions for next year, 2024. So first off, before I get into the state of the market, this is actually original first party data that my team compiled over the course of the fourth quarter of 2023. And they did this by looking at the underlying data around the deals, but also speaking to, uh, a huge swath of venture funds to find out not just what was happening inside their funds with deals that were not yet public, but some of the why that was happening.

And what I'm gonna walk through today is a summary of that data and a lot of the reasons that that data exists, and then try to extrapolate what that means for the broader ecosystem next year. So first off, and this isn't exactly a huge surprise, but seeing it in data is really, um, gives you a good indication just how much this has been true. But fundraising activity has fallen dramatically as we can see from this chart.

And again, as we all know, 20 21, 20 22 were huge peaks, but the amount that that has fallen off in 2023 and beyond and like for the course of this year is really quite high. So the number of rounds raised fell 64%, and the total dollars raised fell 86% from the peak. The good news is what we saw in our first party data in Q4 was that we have now at steady state, we've stopped falling.

So the fall was very real, it was very precipitous. Um, there was a dramatic drop both in the number and the amount of deals that were happening, but that, that drop seems to have leveled off and if anything picked up a little bit in the fourth quarter. So, um, it's our view that, you know, the, the, the fall has happened, it's kind of already been priced in, and we're at a new steady state, at least for the immediate near term.

The other thing, and this again, isn't, isn't a surprise, is that valuations are down significantly across all stages. And the later you go, the more the valuation drops. We're seeing Series D, there's been a 50% drop, 44% for B series C's, 30%, 17%, and 14%.

And actually, the interesting part is when you actually go beyond the data, what you're finding is actually even worse. Than's been reported, and the reason being is in 20 21, 20 20, 20 22, a lot of these deals were being written is clean deals. By that I mean, there wasn't a lot of financial structure, there wasn't a lot of incentives there, just written clean.

But what's happened now is a lot of these deals are being written as inside rounds, particularly at later stages. They have much more aggressive investor friendly terms. So even though the headline number might be you raised 20 at a hundred million, or you raised two at 10, a lot of things are now true that weren't true two years ago.

And that can be in the form of liquidation preferences that can be in the form of anti-dilution protection that can be in the form of rounds that used to get written without needing a board seat now due. Um, so both financial and non-financial incentives are being written into these deals that actually make it even more investor friendly and even worse for the founder, or even more challenging for the founder, um, than we had seen previously. And it, it's not fully baked into these numbers.

You only really see this when you start to look at the actual structure of the deals themselves and realize the amount of structure has, has gone up a huge amount. Now the other thing is, while this does seem terrifying, is in fact this, if you look at this graph about how valuations have shaped out over the last decade, what we're seeing in 2023 is actually much more like a reversion to a the mean than it is a total collapse. Another words, the bubble of 2021 and 2022 was just that a bubble.

And you can see it quite clearly in this chart where we are right now in Q4 of 2023 is roughly the historical average for the past decade if were you to remove 2021 and 2022. Um, so, and you can see that basically across all stages. So there's slightly more seed activity than there's been in the past, um, with the exception of those two years.

But everything else is, is kind of just reverted to that new normal, which is very similar to the old normal. So internally we like to think of it as, you know, it's, it, it's more of a 2016 market, which was, uh, a solid market just nowhere near the peaks that we had seen. The other thing is these valuations, and this mean have actually reflected what was true again in 20 16 20 18, which is that in order to raise these rounds, companies are needing significantly more revenue traction than they did two years ago.

So the number of deals is down, the amount going into them is down. Um, the valuations are down, there's more structure, and on top of it, you need more traction to get to an actual round, um, relative to what you did in the past. So you can see here on this graph, it's showing mean valuations, um, median valuations rather versus traction.

And you can see them, the median valuations are, you know, these are for pre-seed. Seed rounds are sitting in that like 10 million ish dollar range, um, with some exceptions. Interestingly, uh, generative ai, which drove a lot of the hype for the early part of 2023, these are Q4 2023 numbers has started to come back to earth as we've seen that first wave of the hype cycle kind of peter out.

And investors increasingly look for solid businesses or really deep technology. So you're seeing that, you know, in order to get to those big valuations, those outlier valuations are almost always a function of just having deep traction and them having taken longer to raises or being deeper technology businesses. So all of those do represent tailwinds to the founder-friendly market we saw in 2021.

Um, and we don't really foresee any of these fundamentals dramatically changing. Um, the last piece is in addition to all of the things I just said. So it's taking, you know, it rounds are getting off, there's less of them, they're going off for less money.

There's more structure, uh, there's more traction. All of those things are true. Um, but there's one last thing that's true, which is it's taking much longer to go between rounds.

Uh, so it's taking 20% longer than it was even just a year ago and about 40% longer than it was two years ago. So it's a significantly longer journey to get to that next fundraising round. And that's a function of two things.

One, there's just less capital being deployed, so less rounds are getting done. And the other thing, the bar for traction, as we just saw in the, in the last slide, has gone up. So it's taking longer to get to each step, whereas in 2021, we would tell founders to model out two years in between rounds on the assumption that hopefully if they hit metrics at 12 to 18 months, they'd be raising the next round.

Uh, now increasingly we're remodeling three years and beyond in between rounds with the assumption that, you know, it takes two years to hit those metrics and actually raise that round and it gives you a little bit of cushion. So a pretty material change there. Um, and again, much closer to the reversion to the mean, we've seen if you remove 2021 and 2022 for the last decade.

Okay, so we've just heard that the market has gotten harder for sure. Um, we've heard, you know, valuation's down time is longer, traction is up. Great.

Some of that'll be really difficult. So what I would love to talk about now is kind of what comes next. And I have three, two major, um, things I'm gonna talk through.

One of which relates to fundraising, the other of which is relating to a, a big macro trend we think is gonna drive a lot of, uh, both venture financing and a lot of change in the AI and DevOps markets. Okay, so let's talk about what comes next. So these are two big predictions that we've got for 2024.

The first one is, uh, that VCs go broke. And the second is we're gonna see AI agents take off in a very, very major way. So in a nutshell, what we're expecting 2024 to be is a year of a lot less venture capital firms.

Um, venture capital firms have already started going out of business, mostly quietly, but our headline prediction is that one in three venture capital firms that was operational in 2023 will not be a going concern moving forward. And that's a pretty significant reduction. You're talking about a, a wipe out of at least a third in the market.

Um, and we think that for the primary reason that, you know, the people that invest in venture funds have, are overweighted in venture and haven't seen returns for a bit. So there's, there's a lot less willingness to put money into venture capital funds. And venture capital funds need to raise their next fund to survive.

So in the, with the inability to do that, um, they simply go bankrupt. They or they, they don't have capital to invest. One of those two things.

And the interesting thing is, while that seems like a, a scary headline that you know is shocking, it actually might already be true that this is the case right now and we just don't know it yet. So what you're looking at here, and this is, this is courtesy of PitchBook, uh, is the number of funds that have invested in two or more venture deals for each time period. And you can see there was, you know, north of 7,000 funds at the absolute peak in 2021.

And we're actually seeing already that for 2023 and, and the Q4 data doesn't change this in any major way, we're still in Q4, so we don't have complete data, but what limited data we do have, uh, will tell us that actually we're already seeing a 38% reduction in the number of funds that have written a deal this year versus the peak. Um, we think that that is likely to continue. And you know, that number of three to 4,000 venture funds is much more likely to be the number that exists rather than 7,000, which is the number of active venture funds or venture funds that claim they're active in 2023.

A lot of those funds really aren't active. Um, and that's because of the how. So when you have a big headline number, like one third of an industry is gonna go away, it feels like you'd be getting announcements all the time.

And we just saw one of those, right? 4 billion under management. They recently announced that they were winding down.

And that's one of two ways that venture funds close. They implode, right? The partners say, we can't raise the next fund, we're gonna either return capital or we're shutting down.

This is done. But actually what's more interesting is that is actually not gonna be the way that most venture funds cease operating or cease investing. They're just gonna quietly simply stop investing.

And you're seeing that now. And they do that because they still have a portfolio. They're still trying to raise an next fund in some cases they're hoping the market turns, they're hoping they can make it work.

Um, and maybe a few of them will, they'll have some rebounds, they'll, you know, any number of things can happen. But again, our prediction is many of these funds will, will not persist. And what you'll see there is, you know, quiet layoffs, they'll move down to a skeleton staff, they'll manage their existing portfolio, maybe one to two partners will continue to see if a next fund is reasonable.

Um, and then just slowly that will go away to nothing. Um, and you'll see that over the next 12 months, um, with, with a lot more of them officially shutting doors, you know, as time passes and they get further and further away from having investable capital. So that's our first big prediction for, for next year is one in three venture capital firms is gonna cease to exist functionally.

Um, the next one is that 2024 will be the year of the agent. If we look at last year, last year was probably the year of the large language model or you know, I've heard the year AI went mainstream. There's any number of of terms, we're all talking about the same macro trend, right?

I think the launch or chat, GTP, uh, really kind of officially kickstarted the gold rush there. And we are seeing the AI ification of pretty much every vertical across software and many hardware verticals as Raul, when you think about robotics, and that was last year. So this year's gonna be the application of that technology to autonomous agents.

So our view is that, uh, foundation models will give way to fully autonomous production AI agents that will be a part of most products and will exist across basically every vertical. And there's, there's four major agent trends that we're tracking. The first is just their ubiquity.

So this will be something that will be a part of many, if not most products and workflows as they add more and more autonomous pieces to their tech stack. And that could be something like what Microsoft is rolling out, or those could be completely new product paradigms. Um, the other thing is the fragmentation of the AI agent tech stack.

Right now, the most popular and kind of ubiquitous tech stack is either within the cloud providers or open AI's assistant platform. Um, we think that as this becomes more verticalized and more ubiquitous, there's going to become more and more needs for what agents are built on, uh, across any number of dimensions, whether that's observability, whether that's, um, building, whether that's data, whether that's instruction sets. And so the, the, the tech stack is gonna be increasingly fragmented and that's a huge opportunity and also a big growth area in both DevOps and ai, where we're gonna see just a lot more happening.

Um, the next piece is kind of related to that, where ML ops has been around for a while and, and obviously it's empowered us to do a huge number of things and we're seeing some really, really big companies in that space like Databricks. Um, but agent ops is gonna be the new level of infrastructure for macho software and a lot of that infrastructure. Yes, the foundation models that are powering these agents is basically an ML ops stack that's been a machine learning operation stack that's existed for a long time.

But how these agents function in the wild is kind of unclear and what the infrastructure stack is to build, deploy, manage, observe, maintain, all of that as it fragments out across everything is still very nascent and I think a, a huge, uh, a huge opportunity set. Um, and then the last piece will be the rise of agents as a service. Uh, because there's so many people who can build agents pretty cheaply.

'cause it, it's not the same as building a foundation model. You don't necessarily need to train, you don't need to build f foundational tech. There'll be a big market for agents as a service on a task basis, um, and on a vertical basis that we'll be replacing like pieces of software that are currently used by humans.

And so whether that's in marketplaces or elsewhere, that rise of agents as a service, um, as being a pretty significant part of people's workflow and software stack is, is something we see in 2024. And so our last little kind of bit of sub predictions around the, the agent market is, um, agent marketplaces are gonna become much more prolific as people can quickly build pieces of autonomous software that can do things without human supervision. Those that work, there will be high demand for those pieces of workflow, um, that people can use in a non-proprietary way.

And that's gonna give rise to marketplaces. Uh, much in the same way. You might have gone on Fiverr to get a task done.

You will now go to an agent marketplace to get a task done. Um, agents will become increasingly more specialized and verticalized much as software did. We started with, you know, pretty generic basic software and we've increasingly become verticalized.

Uh, AI is increasingly becoming verticalized. Agents will become much more specialized in verticalized this year as well. Um, moving away from more general purpose agents to, you know, deeply specific agents that perform tasks in in highly specific verticals.

Um, the last piece and, and potentially one of the more interesting pieces is agents are gonna breed a new monitory monitoring and, and security, uh, regimen. So these are, you know, fully autonomous p pieces of software, um, how we monitor those, how we track those, what effect it has on our analytics of who's using what and who's in which system, um, what that poses a security threat. Those are all, you know, things we've dealt with in part but never on a scale and scope of, of what we have now, where basically anyone can log in and build in a completely autonomous, uh, agent in, in minutes using, you know, pre-existing technology.

Um, so that's our last piece is that, that that monitoring and security regimen will be an extremely interesting one to watch full of risks and opportunities. So those are our two big predictions, um, that, that two of the big predictions we've got for the year around, you know, a third of the venture capital ecosystem quietly shutting their doors and 2024 being the rise of the agent. And then we had one small bonus prediction is actually because of the confluence of these two trends, um, we actually think this is a great time for venture capital.

So you can see here, uh, and you know, this has been said by many others, um, but downturns and tough times typically do lead to great returns for venture. And you combine that with this brand new, uh, paradigm of operating with, with large language models and agents. Um, and we think that this 2024 period is gonna be that purple line on the graph.

Um, and just a great time to build great companies and uh, and for the people who fund them. Uh, thank you very much for your time.