How do you value a startup with no revenue?

Pre-revenue valuation is a negotiation anchored by method, not a calculation. Three approaches carry weight with investors. Berkus assigns a capped value to each of five de-risking milestones — idea quality, prototype, quality of team, strategic relationships and early sales — which suits pre-seed. The Scorecard method takes the median pre-money for comparable rounds in your region and stage, then adjusts up or down on team, market size, product and competition. The venture-capital method works backwards from a plausible exit value and the return multiple the fund needs, which is how investors themselves think. Run all three: the spread tells you how defensible your number is, and the method your counterparty prefers tells you which story to prepare.

• Berkus: milestone-based, best pre-seed, naturally caps at a few million
• Scorecard: comparables-driven, requires honest regional benchmarks
• VC method: exit-value backwards — mirrors how the investor decides

What revenue multiple is realistic in 2026?

Multiples compressed hard after 2021 and have not returned. For private software businesses, growth rate and retention now drive the multiple more than revenue scale: efficient growth with strong net revenue retention still commands a premium, while flat or heavily discounted revenue trades near the low end. Services and agency revenue is valued on profit rather than revenue, usually a small EBITDA multiple, because it does not scale without headcount. Marketplaces are valued on net revenue (take rate), never on gross merchandise volume — quoting GMV as revenue is the single fastest way to lose credibility in a valuation conversation. Whatever multiple you use, state the comparable set and the date, because a multiple without a source is an opinion.

How accurate is a calculator like this?

A calculator is a structured estimate, not a valuation, quote or forecast. Its accuracy depends entirely on the quality of the inputs you feed it and on how well the benchmark ranges match your stage, geography and sector. Treat the output as a range with roughly ±30% around it, and use it for three things: sanity-checking a number you already have in mind, comparing scenarios against each other, and finding which single input moves the result most. That last use is the valuable one — it tells you which assumption is worth spending real research time on before you commit money.

• Compare scenarios rather than trusting a single absolute number
• Identify the input with the largest swing and verify that one first
• Re-run quarterly — benchmarks and costs move faster than most plans assume

Which inputs do founders most often get wrong?

Four consistently: underestimating time to first revenue, excluding founder salary or opportunity cost, ignoring taxes and payment processing on the revenue line, and assuming a straight-line ramp where real growth is lumpy. The combined effect is usually a 30–50% optimistic bias. A quick correction: take your revenue assumption, halve it, take your timeline, double it, then check whether the plan still works. If it does not survive that adjustment, the plan depends on everything going right — which the failure data shows is the least likely of all scenarios.

What should you do with the result?

Turn the number into a decision and a threshold. If the output is a cost, it becomes the amount you need in the bank before you start, plus a buffer of at least three months. If it is a valuation or funding figure, it becomes the opening range for a conversation, not a price. If it is a runway or break-even number, it becomes a calendar date with a review attached. Numbers that never become dates or thresholds do not change behaviour — and the point of running the calculation was to change what you do next week, not to produce a slide.