Question 1

How do you value a pre-revenue startup?

Accepted Answer

Pre-revenue companies are valued on potential rather than performance, using a small set of established frameworks: comparable transactions (what similar companies raised at), the scorecard method (comparing team, market and product against a regional average), the Berkus method (assigning value to each de-risking milestone), and cost-to-duplicate. In practice the number is set by negotiation and by how competitive your round is, far more than by any formula. The PitchBook-NVCA Venture Monitor puts the median US seed pre-money at $18.3M (Q3 2026). That is a median across all seed rounds, not a figure for pre-revenue companies.

Market figures quoted above are drawn from PitchBook-NVCA Venture Monitor, Q3 2026 (As of September 30, 2026) and update automatically when the benchmark dataset is refreshed.

Run a valuation Ask something else

Question 2

What is a typical Series A valuation right now?

Accepted Answer

The median US Series A pre-money valuation is $60.3M as of As of September 30, 2026, up from $48.1M in 2025. Series B sits at $187M. Those are medians across the whole market. An AI company at Series A prices closer to $83M while a non-AI company prices closer to $44M (AI split from the Q2 2026 edition). Source: PitchBook-NVCA Venture Monitor, Q3 2026.

Question 3

Why is there such a large gap between AI and non-AI valuations?

Accepted Answer

Because capital has concentrated to an extreme degree. 82.7% of US venture deal value in 2026 through Q3 went to AI companies. That concentration pulls every headline average upward and makes blended figures misleading for ordinary companies. It is why we publish the AI and non-AI splits separately rather than a single median.

Question 4

Is this a formal valuation I can use for a 409A or a fairness opinion?

Accepted Answer

No, and you should not present it as one. This is a directional estimate for planning, negotiation preparation and benchmarking. A 409A valuation must be produced by a qualified independent appraiser to satisfy safe-harbour requirements, and costs vary widely with company complexity. A fairness opinion is a separate exercise again. If you need either, we can point you to someone who does them.

Question 5

Where does your market data come from?

Accepted Answer

Every benchmark on this site comes from the PitchBook-NVCA Venture Monitor, published quarterly by PitchBook Data, Inc. And the National Venture Capital Association and available free to the public. We transcribe the labelled values from its charts and record the page number alongside each figure. Where the report plots a series without labelling it, we show "Not published" rather than estimating. The same figures are passed to the valuation engine as anchors, so the tool and the Benchmarks page cannot contradict each other.

Question 6

What moves a valuation the most?

Accepted Answer

In rough order: revenue growth rate, market size, the quality and track record of the team, unit economics (LTV:CAC and gross margin), defensibility, sector heat, and the competitiveness of your process. The last one is underrated. A founder with three term sheets consistently prices better than an identical company with one.

Question 7

Do you sell my information?

Accepted Answer

No. We do not sell, rent or share lead data. Your valuation inputs are used to produce your report and are visible only to us. If you indicate you are considering a sale, a sell-side advisor at FIH.com may contact you. That is the only outbound use, and it is stated plainly before you submit.

Question 8

What is FIH.com's relationship to this site?

Accepted Answer

FIH.com is a sell-side M&A advisory firm and operates this site. The valuation tool is free and open to anyone. If your company is at a size and stage where a sale is realistic, FIH may offer to talk, on a success basis, with no retainer and no fee unless a transaction closes. If you are early or simply benchmarking, you will not be pitched.

Question 9

How accurate is the estimate?

Accepted Answer

It is an estimate, and its honesty is in the range and the confidence score rather than the headline figure. Thin inputs produce a wide range and a low confidence score, which is the correct behaviour. A narrow range built on three data points would be false precision. The more of the metrics step you complete, the tighter and more defensible the output.

The things founders ask

Considering a transaction?