It’s been a little over a year, so I wanted to share some updated thoughts. Unfortunately, these trends have continued to persist, except now, most of my brethren seem to agree and are saying so out loud. So, the million-dollar question is – what’s an early stage investor to do?

At this moment in the market, there are only three strategies available to early-stage managers. In a way, these three have always existed. What’s changed is the relative attractiveness of each.

Strategy #1: Fighting over diamonds in plain sight

The first strategy is to chase the best founders working on the most consensus ideas and fight hard for access. That fight happens either before the company even starts, or after it has shown significant traction. This is a natural outcome of the two consensuses I wrote about last year. The whole world has woken up to the power law, and the whole world agrees that AI is the mega-trend of our lifetime. Put those together and everyone arrives at the same place ready to rumble.

This segment of the market is extremely hotly contested. It’s where capital is concentrating, and it’s produced seed valuations that would have made late-stage investors blush 10 years ago.

But it’s hard to argue that this approach hasn’t also produced real results. Interim markups and pre-exit value creation have been extreme. And the exits that matter in this market have mostly come from mega-compounders that can go public or get acquired for enormous sums (eg Cursor at $60B). A16Z’s performance has been quite strong and consistent, even as their platform has scaled considerably. The same is likely true for quite a few of their peers.

I was speaking to a senior partner at a megafund recently. Not too long ago, he had backed a SaaS company at seed that went public and is worth over $5B. He lamented that investments like that just don’t matter for his fund today, because those returns are dwarfed by their recent growth stage investment in one of the leading AI labs. He really enjoyed and had success in the old model of venture, but fighting (and winning) the diamonds in plain sight was just too attractive to ignore.

Strategy #2: Mining for silver

The second strategy is to avoid the venture game altogether.

The premise of this strategy would be that the venture market is oversaturated, and that shooting for decacorn and centacorn outcomes only makes sense for a very small number of companies. So maybe there’s a strategy where founders raise a relatively small amount of capital in verticals that are less interesting to the megafunds.  You can theoretically enter at better prices, have greater access, and navigate toward large but not outrageous exits — something in the $200 million to $1 billion range. Historically, this is where a lot of venture exits happened. Also, AI gives founders significant leverage that they didn’t have before, truly opening the door for much more capital efficient company building in spaces that are less interesting to traditional venture.

For this to work, three things need to be true:

  1. You need to pick well. This would lead to some combination of a low enough cost basis and/or a low enough loss rate to make the fund math to work.
  2. The companies can get to profitability off the seed and/or find downstream capital that supports this kind of growth path.
  3. There are ample exit opportunities at attractive valuations.

Unfortunately, I think the path forward here is pretty murky.

First, entry prices are still quite high. Valuations have gone up across the board, so it’s very hard to get into these companies at a reasonable price. Even median seed prices have more than doubled in recent years and don’t support this model. Moreover, I think the founders most equipped to build on this path would be very dilution sensitive and find a way to bootstrap for as long as possible, skipping the seed round altogether unless they are offered attractive pricing for their equity.

Second, trying to win on a lower loss rate is likely a fool’s errand as well. This may work in the growth stage post product market fit, but there is just too much uncertainty and risk at the early stages to avoid high loss rates. This necessitates some level of power-law outcome to make up for losses at the portfolio level, which draws investors right back to the more traditional VC game.

Third, downstream capital isn’t there to support this path. If founders can’t get to profitability off their initial round, they will find a downstream market that’s mostly hunting for strategy #1 companies. A business building toward a $500M outcome isn’t what most Series A and growth funds are looking for today.

Fourth, the exits just aren’t there. That’s true both in terms of absolute volume of exits as well as the nature of the few exits that are happening. We tend to be in a world where exits are few and far between, but the ones that are occurring are getting much bigger, unless you are willing to sell for a steal to Bending Spoons.

Strategy #3: Searching for diamonds in the rough

That leaves the third path, which is to search for and cultivate diamonds in the rough.

This is actually the realm seed investors have always occupied. But historically, seed VCs go about this in a very haphazard way. Ask most seed VCs, and you’ll be quite disappointed at the rigor through which they search for these diamonds and how far into the rough they are actually willing to stray. Most seed VCs have small teams and go about their day taking a relatively small number of meetings with relatively well pedigreed founders that arrive on their calendars through referrals from their trusted network. That’s not exactly searching systematically for diamonds in the rough – it’s more like stumbling around the edges of the fairway. In a world where diamonds are plentiful and easy to spot, this approach might work for a while. But I think we’re now in a world where these diamonds are very, very hard to find. They lie deep in the rough. And there’s plenty of cubic zirconia sitting on the edge of the fairway that looks like the real thing, creating headfakes for anyone pursuing the strategy.

Where this leaves us

The uncomfortable reality is that, of these three, the strategy with the best recent results has been strategy #1. Because of the convergence of the two consensuses, a lot of these companies are showing private value creation that’s pretty hard to believe. Will it hold, and will these companies actually exit? Perhaps not. But perhaps the few true blockbuster exits will cover up all the sins.

Meanwhile, the folks searching the rough are having a much harder time. The downstream market is mostly looking for fairway companies, so everyone else is finding it very difficult to raise follow-on financing. At the same time, megafunds are investing in just about every early-stage company that matches the consensus profile, which has pushed valuations up across the board. Even the median seed valuation is up more than 50% over the last few years.

So most seed managers are paying meaningfully more to get in, downstream capital is scarce, and the exit market is thin. That’s a tough combination. Especially because the rough is, at least in theory, where the highest potential returns will be found.