A Valuation Is a Promise You Have to Keep

Almost every incentive in fundraising pushes toward the same answer. Raise more. Raise at a higher price. Take the term sheet with the bigger number on it. It signals momentum, it buys runway, it reads as validation, and turning it down feels like leaving value on the table for no reason.

Most of the time that instinct is right. The exception is specific and worth understanding, because a valuation is not a reward for work already done. It is a price set against work not yet done, and the company has to grow into it before the next round or the arithmetic turns against everyone holding common stock.

Where Prices Are Right Now

The current market makes this concrete. In the fourth quarter of 2025 the median post-money valuation for seed rounds on Carta reached an all-time high of 24 million dollars, up from 18 million a year earlier and 16 million two years before that. Series A medians moved further, to 78.7 million post-money, a 37 percent year-over-year increase. The rise was not confined to the top of the market: the 25th and 75th percentiles both climbed to new highs as well.¹

Some of that reflects genuine progress in what companies can do, and some of it reflects a bidding contest for a shrinking number of deals. Deal counts have been falling even as prices climb, and the concentration is steep: in 2025 the top 10 percent of US startups on Carta that closed a round raised about half of all capital, while the bottom half of companies raised 14 percent between them.¹ One investor quoted in Carta's own analysis put it plainly, saying that alongside sincere belief in these companies you also see prices that are in many cases hard to justify.

AI accounts for much of the lift. Companies positioned as AI startups have been drawing meaningfully higher valuations at seed and Series A than comparable companies without that framing. That premium is worth naming for a specific reason: if your price reflects a category rather than your traction, the growth expected of you was set by companies you may have little in common with.

The Step-Up Problem

Here is the mechanism that makes a high entry price expensive rather than free. Your next round is judged against your last one. Investors look for a step-up, a multiple of the previous valuation that reflects demonstrated progress, and the size of the step-up they expect has not risen alongside the entry prices.

The median seed-to-Series-A valuation step-up was about 2.6 times in 2025. That is a recovery from 2.4 times in 2024, and it is still well short of the 4.2 times recorded at the 2021 peak.² So companies are entering at record prices while the multiple they will be measured against has compressed. Those two facts point in opposite directions, and the gap between them is where founders get caught.

What it takes to clear a Series A has moved as well. In 2021, roughly 500,000 dollars of annual recurring revenue growing at 300 percent could attract a Series A lead. The current floor sits closer to 1 to 2 million in ARR with growth above 150 percent, and the strongest deals are doing better than that. Meanwhile the share of seed companies that graduate to a Series A at all has fallen from around 50 percent to roughly 38 percent.³

And the clock runs longer than it used to. The median gap between primary rounds reached about 696 days in the second quarter of 2025, roughly 23 months, against closer to 600 days two years earlier.⁴ A higher price therefore has to be justified by more revenue, against a higher bar, after a longer wait, with worse odds of getting a round at all. None of those four things is improved by having raised at a bigger number.

What Happened Last Time

This is not hypothetical, and the precedent is recent enough that many of the people involved are still working through it. Companies that raised during the 2021 boom, a good number at 50 to 100 times ARR, met a market where software multiples had compressed from a median around 15 times forward revenue in early 2021 to roughly 6 times by late 2023. For companies that had not grown into their price, a down round became arithmetically unavoidable rather than a judgment on their quality. Somewhere between 25 and 30 percent of late-stage venture rounds between 2023 and 2025 were priced at or below the prior round.⁵

A down round is survivable and it is not shameful, but it is not free either. Anti-dilution provisions adjust the conversion price for existing preferred holders, which transfers ownership away from common shareholders, meaning founders and employees. Staff hired at the peak find their options underwater. And the negotiation happens at the point of weakest bargaining position, because the alternative to accepting the price is running out of cash.

The other pattern from that period is still with us. Bridge rounds, which usually set no new valuation and buy time rather than solving anything, accounted for 16.6 percent of all cash raised by startups on Carta in the second quarter of 2025, up from 11.8 percent a year earlier. During 2021 that figure was typically below 10 percent.⁶ A bridge is a deferral. It is a reasonable one when there is a dated milestone on the other side of it, and a slow problem when there is not.

Why the Lesson Is Easy to Forget Right Now

Here is the part that makes this worth writing about in 2026 rather than in 2023. The hangover has cleared. The down-round rate fell to 11.4 percent in the first quarter of 2026, back in line with 2019 and 2020 levels, after peaking above 22 percent in 2023. Dilution is down and terms have moved in founders' favor.⁷

That is a material improvement and not a trick. It also means the market currently offers very little friction against overpricing, which is the same condition that existed in 2021. The correction is not a prediction anyone can make with confidence, and this is not an argument that a reset is coming. It is a narrower point: the discipline that protects a company against a stretched valuation is cheapest to apply when nothing appears to require it.

The Costs That Show Up Even Without a Reset

Suppose prices hold and no correction arrives. Raising more than a company needs still carries a bill.

Dilution is the obvious one and it compounds quietly. Seed rounds have consistently involved selling about 20 percent of the company across recent quarters, and the median founding team holds roughly 56 percent on a fully diluted basis after a seed round once the option pool is accounted for.⁸ Every additional round refreshes that pool, usually pre-money, which dilutes common holders before new investors arrive.

Then there is the operating effect, which gets less attention. Capital raised tends to get spent, and spending against a valuation rather than against a plan is how companies build cost structures they cannot sustain if growth slows. The team-size data points the other way lately, with average equity-holding headcount at seed-stage companies dropping considerably from the 2021 peak, which suggests some of this discipline has been learned. Money raised to hit a milestone behaves differently from money raised because it was available.

And a high price narrows your buyer pool at exit. A company valued at 80 million in its last round has ruled out most acquisitions below that number as anything other than a disappointment for its investors, even where such an outcome would be excellent for its founders.

When the Bigger Round Is the Right Call

This would be a dishonest piece if it argued that less is always better, so here is the other side. In a genuine land-grab market where a competitor raising more will simply outspend you on distribution, the larger round is the correct defensive move. In capital-intensive categories, hardware, biotech, anything with a long pre-revenue development path, undercapitalizing is the more common way to die. And given a median 23-month gap between rounds, runway is itself among the scarcest resources a company has; a round sized to 18 months in a market that takes 24 is a planning error dressed up as discipline.

The distinction is not size but justification. A larger round raised to fund a specific plan through a specific milestone is a different instrument from a larger round raised because the term sheet was available and the number felt validating.

The Questions That Sort It

Three, and they are answerable before you sign anything. What revenue and growth rate would justify a two to three times step-up on this price, and does the plan this money funds get you there with margin for the things that go wrong? If the answer requires everything to work, the price is borrowing from a future you have not yet earned.

Is this price reflecting my company or my category? If a comparable business without an AI framing would be raising at half the number, the premium belongs to the sector, and sector premiums compress. The expectations attached to the price will not compress with them.

And what does this round have to accomplish, stated as a milestone rather than a runway period? Founders who can name the metric that opens the next round tend to raise to that number. Founders who cannot tend to raise the maximum available and discover the milestone later.

This is also where the choice of investor matters more than the price they offer. An investor who understands the category well enough to have a view on what your next milestone should be is worth more than one paying a higher number on the strength of a trend. That is the match AngelHive is built to find, on stated thesis and assessed evidence rather than on who is willing to bid highest in a hot quarter. The highest offer and the right partner are sometimes the same term sheet, and it is worth knowing which one you are accepting.

Sources

1. Carta, Record-setting early-stage valuations. https://carta.com/data/record-setting-valuations/

2. SaaStr, The Real State of Seed Today: Top 10 Learnings from 50,000 Startups (summarizing Carta State of Seed). https://www.saastr.com/the-state-of-seed-today-10-key-learnings-from-cartas-latest-data/

3. ValueAdd VC, Pre-Seed to Series B Round Sizes 2026 (citing Carta and PitchBook-NVCA Venture Monitor). https://valueaddvc.com/blog/startup-funding-rounds-in-2025-whats-normal-at-pre-seed-seed-a-and-b

4. Causo Hub, Down rounds 2026: the real rate, source by source (citing Carta time-between-rounds data). https://hub.causo.ai/guides/down-rounds-and-bridges-this-year-2026

5. ValueAdd VC, Down Rounds: Mechanics, Consequences, Survival (citing PitchBook). https://valueaddvc.com/blog/what-is-a-down-round-mechanics-consequences-and-how-founders-survive-them

6. Carta, Bridge Rounds Got a Boost in Q2. https://carta.com/data/bridge-rounds-q2-2025/

7. Carta, State of Private Markets: Q1 2026. https://carta.com/data/state-of-private-markets-q1-2026/

8. Speedrun, 8 Takeaways from Carta's State of Seed Report. https://speedrun.substack.com/p/8-takeaways-from-cartas-state-of