Round names are meaningless right now.

Or what I get asked about everyday by LPs, founders, and VCs.

I spent some time talking to Dominic-Madori Davis from TechCrunch last week about seed round pricing. She’s been covering venture and the early stage startup market and has also noticed the valuation and round name fluctuation we are all seeing. She asked me to define where the market sits. Pre-seed. Seed. Series A. Which numbers are outliers, and which are the new normal? I don’t really have great answers on what defines each round name but I could give the reasoning.

The reality is that round names are sort of meaningless right now. And I’m not sure they should mean anything at all anymore.

TechCrunch published a piece today that I contributed to, and I want to go deeper on a lot of my reasoning on the quotes. It’s weird to see myself in the outlets I’ve always helped my founders get mentions and coverage in. I don’t hate it.

I’ve spent my career at the intersection of operators and investors. Twilio in the early days long before the IPO run. GitLab when we were scrappy and figuring out what enterprise open source even meant. GitHub through the Microsoft acquisition. I’ve watched markets cycle and I’ve watched founders make decisions about their companies based on what a round was called instead of what it actually needed to accomplish.

So if round names don’t matter, what is actually going on in the market?

Pre-seed is sitting around $10M post. Seed is around $25-30M post, which is where the Carta data lands for this quarter. YC Demo Day was last week and most companies were raising at $40M post for a Seed. I was there, talking to founders, watching the number climb.

Those aren’t seed rounds. Not by any definition most of us were trained on. But who cares?

A $5M check into a $40M post-money valuation for a company that is eight weeks old, with pilots that haven’t had time to churn or stick, is a different financial instrument than what we used to call a seed round. The label is the same. The risk profile is completely different. For the big funds writing scout style checks into 10-20 party rounds each cycle, totally makes sense. For YC’s LPs that get exposure to the 2-3 outliers per cycle, totally makes sense. For funds like mine? It’s getting harder to manage that risk. The founders this batch were exceptional. The why behind the companies were exceptional. It is still a hard check to write if you’re a small early stage fund.

Part of what’s driving the numbers is that a few companies have genuinely rewritten the rules on how fast traction can materialize. Cursor (holy shit what a run) hit $100M in revenue in 12 months. Lovable, Bolt, ElevenLabs, there’s a cohort of companies that showed what AI-native velocity can look like. The problem is the market is now pricing every company as if it’s Cursor but in reality, the normal growth patterns are actually what is happening in the underlying non-Cursor world. The outliers are setting the floor, not the ceiling, and that math doesn’t work for most founders.

I’ve seen things further out than that. A $300M seed closed recently but super amazing founder with a great product space. Likely a good bet. But not a seed. I pointed to it when the Dominic asked about extremes. While not a seed, it is a bet that a single company will become base infrastructure for an entire era of technology. The round name is a rounding error on the thesis.

Why this is happening

So many different winds are blowing in the same direction that it’s hard to separate them, but I’ll try.

The first wind is the mentality shift around platform bets. Investors believe a small set of winners will become the base layer of the AI era, the next AWS, the next Google. If you believe that, price almost doesn’t matter, because the returns on a genuine infrastructure monopoly dwarf anything you gave up on entry valuation. Wildly, the logic makes sense. It’s just not applicable to most companies, and the market is applying it broadly.

The second wind is large funds moving earlier. If you’re managing a $5B fund and you want meaningful ownership in the companies that define AI infrastructure, you either write a lot of small checks or a few massive ones. Most are going with option two. A $50M check into a $300M seed round is a rounding error in a multi-billion dollar fund. For that fund, it makes sense. For everyone else trying to benchmark off that number, it creates a distorted picture of what the market actually looks like.

The third wind is YC. I’ll say what investors say quietly in rooms: there is a YC tax. YC sets the benchmark. I’ve watched it happen quarter over quarter, they price their batch, the SF market follows within a quarter or two, the national market catches up after that. Right now, most companies out of Demo Day were pricing at $40M post. I heard some went higher. I was standing in a room with other investors and multiple people said out loud: this is priced too high. The network, the alumni relationships, the downstream signaling from YC are all great value adds, truly, they are an exceptional group. It doesn’t change whether a team with eight weeks of runway can grow into a $40M valuation before they need to raise again at an even higher valuation. Are we setting a lot of those teams up to fizzle out because they can’t hit the higher expectations? I think so.

There is also a talent premium baked into the most extreme numbers. Investors are paying extraordinary prices for founders with the right pedigree. Repeat founders, ex-OpenAI, the names that carry their own meaning in a room. One investor quoted in the TechCrunch piece put it plainly: it’s a war for great researchers right now. That’s true. But the risk is that the premium applied to a handful of genuinely exceptional people bleeds into the pricing logic for everyone building in adjacent spaces. The market starts conflating founder brand with company value, and those aren’t the same thing.

What this means for small funds like mine

I run a small, specialized fund. We invest at pre-seed and seed in AI infrastructure. My check sizes are $750K at pre-seed, around $1M at seed. I won’t go above a $60M post-money valuation, and even that is the high end. The math stops working for LP returns above that number if you’re a fund our size. But the math works for larger funds.

What this market has done is push us earlier and out of some of the higher priced “seed” deals that are getting A and B prices. Not because we changed our strategy, but because if the first round a company raises is already priced at $40M with no product and no traction, I can’t write a check into that. So we are backing founders before that round exists. Before the product. Sometimes before they’ve left their current job. We give them enough capital to hire a couple of engineers, get some pilot customers, and start building. Then we help them raise the larger seed from funds with bigger pools of capital.

This only works because we are highly specialized. Our network runs deep in AI infrastructure. We can be useful in ways that generalist funds can’t be at that stage. If you’re trying to be a generalist fund at our size right now, I think that’s a very hard path. The market is telling you to either be specialized and early or generalist and large. As i’ve said many times before, the middle is hard. Anyway, if you’re thinking about building in AI infra or know someone that is, send them my way. :)

What this means for founders

So this is more complicated.

The most important thing I can tell any founder right now is: the headline rounds are outliers, not the market.

When you see a $300M seed, that is not a benchmark. When you see $40M YC pricing, that is a specific set of companies with a specific set of investors making a specific kind of bet. The data for everyone else shows pre-seed sitting around $10M post and seed around $25-30M post. I pulled the 2025 data from Carta, Sapphire, and others, and it all points in the same direction.

My biggest warning: pricing your round too aggressively too early is one of the fastest ways to make your next round harder. You are not just setting a valuation. You are setting a bar you have to clear in 18 months to raise again. And Series A is at record-high valuations right now, the hardest to achieve in many, many cycles. Time between rounds is the longest it has been in a long time. If you price yourself into a corner and the market resets or growth takes longer than expected, which I see happen every day, you are stuck.

Jonathan Lehr of Work-Bench said it well in the TechCrunch piece: higher seed valuations mean less margin for error, less room for experimentation, less tolerance for pivots, and more scrutiny if progress doesn’t match the capital raised. I’d add that this dynamic hits first-time founders hardest. When capital moves this fast, access matters more than merit at the start. Repeat founders and well-networked teams get the benefit of the doubt. First-time founders have to work harder to break through, and then they get less runway to prove themselves once they do.

Raise what you need to get to the next meaningful milestone. Optimize for working with the best investors, not for hitting the highest number. Early valuation is not a win. It is a constraint you carry into every conversation that follows. And I need to write a piece on Party Rounds and why they can also set founders up to fail. Another day.

The bigger question

Dominic asked me at one point whether we should just retire round names altogether. I said yes. It might be time.

Round names exist to create shared benchmarks, a way for founders, investors, and LPs to align on what stage a company is at and what the risk profile looks like. When a seed round can mean anything from $1M on a $5M post to $300M with no traction, the benchmark is gone. You’re not benchmarking anymore. You’re just making it up as you go. Which is really sort of fine? Isn’t that what early stage venture and startup building is anyways? Anyone that says otherwise is bullshitting.

And storytelling is fine. Narrative has always been part of venture. The problem is when we start using it to price rounds instead of to describe them. When “this could be the next AWS” becomes the reason a pre-product company is valued at $150M, we’ve crossed from investing in potential to investing in hopes and dreams. Sometimes fiction turns out to be true. Most of the time it doesn’t, and the founders are the ones who pay for it but they also are the easiest teams to back on attempt 2 and 3.

The round name is dead. What actually matters is: what do you need, what will you do with it, and what does your next milestone look like when you come back to raise again?

Get those three things right. The label on the round doesn’t matter at all.

Dominic, thanks for being so thoughtful with your questions and your writing. Excited to see where round names and valuations go next.