What valuation should you raise at

Valuation is the last number you work out. First you decide why you're raising and how much that takes. Then you decide how much of the company you're willing to sell for it. Divide the first by the second and that's your post-money valuation. A fund that wants to invest will come back with its own number anyway.

Most founders start at the other end. They look for a number, find one that feels right and build the round around it. A low one feels like being told you're worth less than the founder down the road. But you can do a good deal at a low valuation and a bad deal at a high one. The number on its own doesn't tell you which.

Part of what makes it hard is that the question is abstract. Anyone can name any price, and the answer's different for every company. So don't spend weeks on it. Spend the time on the two numbers you can actually work out.

How much to raise

The amount comes from why you're raising. Before you think about how much, write down what the money has to prove, by when, and what that costs. The reason you're raising goes into that part. Raising more because things are going well and there's room for it is how rounds get too big.

A range is fine, and it'll probably move. I've seen it many times. A team sets out to raise two or three million and closes one. The round takes longer than planned, as rounds always do, and revenue comes in along the way. At some point chasing the full three stops making sense when one covers the next stretch. The bigger round comes six months later, with revenue behind it, at a higher price.

How much of the company to sell

The share you sell is the number that matters, and the valuation follows from it. At seed and Series A, the median round on Carta sells between 19% and 20% of the company. Pre-seed usually sells less. Carta's median cap in 2025 for SAFE rounds of $1 million to $2.5 million was about $15 million. Raise $1.5 million on that and you've sold 10%.

Pick a range you can live with for this round, say 10% to 20%. If you need $1.5 million, that's a post-money of $7.5 million to $15 million. Take away the $1.5 million and you have the pre-money, the number most term sheets quote.

Your floor is the lowest valuation you'd take. Work it out from the smallest amount that still gets you to the milestone and the most you'd sell. If an offer comes in under it, either the milestone gets smaller or you walk. In a priced round, count the option pool too. The lead will usually ask for it to be topped up before the money goes in, and that comes out of your side.

Remember there's another round after this one, and your investors want it as much as you do. Sell too much now and there's not enough left to sell later.

Why the valuations you look up don't fit your round

Every valuation you can look up belongs to somebody else's round.

On 30 September we asked an AI assistant with web search what a made-up company should raise at. The company was an AI agent for accounts payable in the US, with four paying customers and $6,000 a month in revenue. It said $2 million on an $18 million cap, and much lower for the same company in Europe. It sized the round from a milestone first, the right way round. Then it gave a table of scenarios, because it couldn't see what actually sets the price: whether the customers renew, who the founders are, who else wants in. The number in the middle came from medians of other people's rounds.

Medians hide most of what matters. On Carta, the median pre-money seed valuation for a SaaS startup in the first quarter of 2026 was $33.3 million in the Bay Area. Outside the Bay Area and New York, it was $12.7 million. Same dataset, same quarter, same kind of company, and one number is more than two and a half times the other. In Europe, PitchBook puts the median for pre-seed and seed together at €6 million for the same quarter.

The top of the market is mostly AI money. On the largest pre-seed SAFEs, the top tenth of caps reaches $100 million, and Carta ties that to investor enthusiasm for AI. AI companies took 49% of all pre-seed dollars in the first half of 2026. So a fintech without an AI story, and most AI companies too, are looking at somebody else's market.

Announced valuations are worse, because announcing is a choice. From what I've seen, founders announce a valuation when it's high. When the terms were bad, they often don't announce at all. J.P. Morgan's note in the Q2 2026 PitchBook-NVCA Venture Monitor says undisclosed terms often coincide with structured rounds, down rounds and flat rounds. So if five of the fifteen companies in your niche announced a valuation, you're looking at the five that wanted you to see it.

Then there's everything behind a number: the stage, the team, the traction, where it was raised, from whom and when. The exception is a founder everyone already knows, who can name a price and get it. For everyone else, my own view is that it's easier to raise from your local VCs, whatever local means for you. It can take longer, and the price will be the local one.

What a high first price costs you

The first price you set is the one the next round gets measured against. If the next investor won't meet it, you either take a lower price or they walk.

With friends and angels, the cheques are smaller, and you usually get less done with them than the price assumed. Then you go to a fund with a valuation it would never have offered at that stage, and it doesn't come in. I see it more in Europe, where the medians are lower.

Timing does the same thing. Before the product's live, a lot of funds will tell you it looks great and they want to see it in customers' hands first. If you've already set a high price, they won't come back and meet it. Raising before launch usually means a lower number, and I think it's worth accepting that.

Then there's the 30% on top. You work out a number you could probably raise at and add 30% to leave room to negotiate. The round grows to match, and now you're pitching $2.5 million for a product nobody's seen. On Carta, few pre-seed rounds go above $2.5 million at all, and one in three was under $250,000 at the end of 2025.

Whether to put a valuation in your deck

Who sets the price decides where the number goes. In a round a fund will lead, leave the valuation out of the deck and the first call. A fund that wants to invest will propose a price or work one out with you. Why some rounds never close covers that conversation. Give the amount range if they ask, keep your floor to yourself and let the fund name its price first.

A SAFE is different, and it's how 93% of pre-seed rounds on Carta were raised in the second quarter. Without a lead, you name the cap yourself, the way Nobody will lead your round describes. Work it out from the share you're aiming for, and keep the floor to yourself.

"we're raising [amount] on a post-money safe with a [cap] cap. that's [share] of the company for this round."

If a fund asks what valuation you have in mind before there's an offer, give it roughly. It'll hold up if you know why you're raising and how much you need.

"we're raising around [amount] and aiming to sell about [x] percent. if you'd lead, i'd rather hear your number first."