How Are Startup Valuations Calculated? 5 Methods (2026)
At seed, VCs don't calculate a valuation from scratch. They reverse-engineer it from an ownership target and recent comparable deals. Here's the math, and the 5 methods.
How Are Startup Valuations Calculated? 5 Methods (2026)
How are startup valuations calculated at seed? Investors rarely compute a number from first principles. They anchor to recent comparable deals and reverse-engineer the price from how much of your company they need to own. The five named methods, comparables, the VC method, scorecard, Berkus, and DCF, mostly exist to justify a number that two levers already set.
Most valuation guides hand you a menu of five methods and a spreadsheet. That is the wrong mental model for a seed round. A VC does not sit down and calculate what your company is worth. They decide how much of it they need to own, and work backward from there.
The number that lands in your term sheet is set by two things: what comparable startups priced at recently, and the ownership stake the fund needs to hit its return math. The five textbook methods matter for exactly one reason. You need to speak the language well enough to argue your own number. Here are all five, then the two that actually move the price.
How are startup valuations calculated? The 5 methods
Startup valuations are calculated with five named methods, but at seed only two of them touch your real number. Here is the full menu, ranked by how much a real seed VC leans on it:
- Comparable transactions valuation. Price off what recent, similar startups raised at. This is the anchor for almost every seed number.
- The VC method. Back-solve from a target exit value and the return multiple the fund needs. Used occasionally as a sanity check.
- Scorecard valuation method. Start from a regional average pre-money and adjust up or down for team, market size, product, and traction. Popular with angel groups.
- Berkus method. Assign a dollar value (often up to $500k each) to five risk-reduction milestones like a solid team or a working prototype. An angel-stage framework.
- Discounted cash flow (DCF). Project future free cash flows and discount them to today. Effectively unusable for a company with no cash flows.
The next section shows which of these a seed investor actually reaches for, and which ones are there to fill a spreadsheet.
The 5 startup valuation methods, and who really uses them
Comparable transactions valuation is the only method on the standard menu that a seed VC leans on directly. The other four are frameworks advisors teach, not levers that set a real price.
| Method | What it does | Who actually uses it at seed |
|---|---|---|
| Comparable transactions | Prices off recent similar deals | Every VC, implicitly |
| VC method | Back-solves from a target exit and return | Some, as a sanity check |
| Scorecard method | Adjusts a regional average by factors | Angel groups, advisors |
| Berkus method | Assigns dollars to five risk milestones | Angels, rarely priced VCs |
| DCF | Discounts projected free cash flows | Almost no one pre-revenue |
Don't build a DCF for a pre-revenue startup. It projects a precision you do not have, and handing a VC a ten-year cash-flow model signals you do not understand how seed pricing works. The scorecard and Berkus methods are more honest about their guesswork, but even they get overruled the moment a real comparable deal exists.
How VCs actually value startups at seed
How VCs value startups at seed comes down to two numbers the textbook methods never mention: a comparable price and an ownership target.
The comp sets the ceiling and floor. Investors know what similar companies at your stage and sector priced at in the last few quarters, and they will not stray far from it. Carta reports the median seed pre-money valuation reached $16M in its Q4 data (Carta's State of Private Markets). Benchmarks vary by dataset: OpenVC, summarizing AngelList figures, puts pre-seed nearer $10M and seed around $20M. Your job is to find the comps that make your number look like the market, not an outlier.
The ownership target sets the actual price. Most seed leads need 15% to 20% of the company to make the fund's return math work. CRV notes that most seed rounds result in dilution in the high teens, roughly 20% (CRV). After a seed round the median founding team still owns about 56.2% collectively (Carta's Founder Ownership Report), which tells you how much investors and option pools have typically taken by then.
Your post-money valuation is the VC's check divided by the ownership they need. Everything else is theater.
The formula: pre-money from check size and ownership target
The single formula that predicts your headline number is the check size divided by the ownership the fund needs. Run it before any meeting so you are not surprised by the offer.
post_money = check_size / ownership_target
pre_money = post_money - check_size
Example: $3M check, VC needs 20%
post_money = 3,000,000 / 0.20 = 15,000,000
pre_money = 15,000,000 - 3,000,000 = 12,000,000
Here is the recipe in five steps:
- Ask the ownership target. Most seed funds need 15% to 20% (CRV). Ask the partner directly; good ones will tell you.
- Take the check size. The median seed deal value is roughly $3M (PitchBook).
- Divide check by ownership for post-money. $3M / 0.20 = $15M post-money.
- Subtract the check for pre-money. $15M - $3M = $12M pre-money.
- Sanity-check against comps. Carta's median seed pre-money is $16M (Carta), so a $12M pre sits just under market and is easy to defend.
If your ask implies an ownership stake below what a fund needs, either the check goes up or the valuation comes down. That negotiation, not a valuation formula, is where the real number gets decided.
Pre-revenue valuation methods: no revenue, no DCF
Pre-revenue valuation methods exist because you cannot run a DCF on a company with no cash flows. The two most cited are the scorecard valuation method and the Berkus method.
- Scorecard method: Take a regional average pre-money for your stage, then adjust it up or down for the strength of the team, the size of the market, the product, and any early traction. Angel groups use it to compare deals on a common scale.
- Berkus method: Assign a dollar value to each of five risk-reduction milestones (idea, prototype, team, relationships, and launch). It caps a pre-revenue company at a defensible ceiling.
Both produce a range, and both get overruled the moment a real comparable deal exists. At the earliest stage the price is often carried by a SAFE cap rather than a priced round: in Q1 2024, 41% of checks in SAFE rounds under $1M were below $25,000 (Carta's State of Pre-Seed), so ownership gets set by caps and check sizes long before anyone runs a scorecard. Use these methods to build a defensible range, then price off comps and the investor's ownership target like everyone else.
Turn your number into a second term sheet
Knowing your number is half the job. The other half is running enough of the right conversations in parallel that a second term sheet appears, which is what actually moves the price. Causo matches you to the investors most likely to fund your stage and sector and drafts the outreach, so you can run that process without spending three weeks list-building.
Once you are in rooms, anchor to comps and let competition do the rest:
ā Good: "Recent seed rounds in our category priced at $12M to $16M pre-money. We're raising $3M at a $12M pre, and we have a second term sheet." Anchors to the market and creates pressure.
ā Bad: "We think we're worth $30M because our market is huge and the team is world-class." No comp, no leverage, and it invites the investor to mark you down.
The founder who walks in with three comps and two interested funds sets the price. The founder who walks in with a scorecard spreadsheet takes whatever is offered.
FAQ
How are startup valuations calculated? At the early stage, investors rarely compute a valuation from first principles. They anchor to what comparable startups priced at recently, then reverse-engineer your headline number from the ownership stake the fund needs. The five named methods (comparables, the VC method, scorecard, Berkus, and DCF) mostly exist to justify a number those two levers already set.
How do you value a startup with no revenue? With no revenue, you cannot run a discounted cash flow, so pricing falls back to comparable deals and the investor's ownership target. Scorecard and Berkus methods can produce a defensible range, but in practice the number is set by what similar pre-revenue startups raised at and how much equity the lead needs. Cite recent comps in your stage and sector to anchor the conversation.
How much equity do founders usually give up in a seed round? Most seed rounds land in the high teens, around 20% dilution, according to CRV. After a seed round the median founding team still collectively owns about 56.2% of the company, per Carta's 2025 data. Expect a lead fund to want 15% to 20% to make its return math work.
What is the formula to compute pre-money from investor check size and ownership target? Post-money valuation equals the check size divided by the ownership percentage the investor needs. Subtract the check to get pre-money. Example: a $3M check for 20% implies a $15M post-money and a $12M pre-money valuation. This single line predicts your headline number more reliably than any named method.
Do VCs ever rely on Berkus or Scorecard methods when pricing a seed deal? Rarely as the primary driver. The scorecard valuation method and the Berkus method are popular with angel groups and advisors for sanity-checking a range, not for setting the final price at a priced seed round. A seed VC almost always anchors to comparable transactions and their ownership target first, then uses these frameworks to explain the result.
Related on the hub
- AI Startup Valuation 2026: What AI Seed Rounds Raise At ā Related valuation guide.
- Seed valuation 2026: fair ranges, SAFE caps, and dilution math ā Related valuation guide.
- Raising a seed round for an AI agent startup in 2026 ā Related fundraising basics guide.
Run this playbook inside Causo.
Match to the best-fit partner at 1,000+ funds, draft a hyper-specific email, and send from your email ā in one place.