Five elements, up to $500,000 each, added together. We explain where the method came from, what its creator has changed since, and value a Dutch pre-seed startup with it three different ways.
The Berkus method values a pre-revenue startup by adding up to $500,000 for each of five risk-reducing elements: a sound idea, a prototype, a quality management team, strategic relationships, and product rollout or sales. Angel investor Dave Berkus devised it in the mid-1990s. In the version he published in 2009 and still uses, the elements cap a pre-revenue valuation at $2 million and a post-rollout valuation at $2.5 million.
The Berkus method is not one of the five methods Equidam runs. Those are the Scorecard method, the Checklist method, the Venture Capital method and two discounted cash flow methods (five startup valuation methods). Our Checklist method descends from Berkus, but it replaces his fixed dollar slots with a data-driven country maximum. This article is about the original: how it works, how it has changed, and what it produces for a European company today.
What is the Berkus method?
Berkus built the method around one idea, which he restated in his 2016 update: “the best way to value a start-up is to give value to those elements of progress by the entrepreneur or team that reduce risk of success” (Berkus, 2016). It deliberately ignores financial projections. In his 2009 blog post he explained why: “fewer than one in a thousand start-ups meet or exceed their projected revenues in the periods planned” (Berkus, 2009, archived).
The five elements, with the risk each one reduces in Berkus’s 2012 wording (Berkus, 2012):
| Element | Risk it reduces | Maximum per element (2009 to today) |
|---|---|---|
| Sound idea | Basic value, product risk | $500,000 |
| Prototype | Technology risk | $500,000 |
| Quality management team | Execution risk | $500,000 |
| Strategic relationships | Market risk and competitive risk | $500,000 |
| Product rollout or sales | Financial or production risk | $500,000 |
| Maximum before revenue | Four elements | $2,000,000 |
| Maximum after rollout | All five elements | $2,500,000 |
What is the maximum valuation under the Berkus method?
Berkus gives two ceilings. The $500,000 figures, in his words, “are maximums that can be ‘earned’ to form a valuation, allowing for a pre-revenue valuation of up to $2 million (or a post rollout value of up to $2.5 million)” (Berkus, 2009). Many popular guides give $2.5M as the maximum pre-money valuation for a pre-revenue startup. That is the post-rollout ceiling. The fifth element, rollout or sales, is what lifts a company from $2M to $2.5M, so a company with no product in the market should not reach past $2M.
Are the five amounts added, or is it a ladder?
They are added. Each element is a separate bucket worth anything from $0 to $500,000, and the valuation is the sum of the five. One widely read guide presents the elements as a ladder ($500K to $1M for the team, $1M to $1.5M for the prototype, and so on), as if each step stacked on the one before it. Berkus’s table has no such bands, and when a reader asked him directly in 2021, he replied: “Yes, add the scores to obtain the valuation” (Berkus, reply of 21 September 2021).
In formula form:
Valuation = Σ (score for each element × maximum per element), with each score between 0% and 100%.
Who created the Berkus method, and how has it changed?
Berkus has kept revising the method, mostly in blog posts and replies to readers, and since 2016 each change has moved it further from a fixed $500,000 table.
| Date | What changed | Source |
|---|---|---|
| Mid-1990s | Berkus devises the method to value start-ups “without having to analyze the projected financials” | Berkus, 2009 |
| 2001 | First published widely in Winning Angels by “Harvard’s Amis and Stevenson” | Berkus, 2009 |
| After the dot-com bubble | Maximums reduced “as the investment market adjusted from the craziness of the bubble” | Berkus, 2009 |
| 4 Nov 2009 | Blog post sets out five elements at $500,000 each, $2M/$2.5M caps | Berkus, 2009 |
| 25 Mar 2012 | Reposted with extended risk labels (product, competitive and financial risk) | Berkus, 2012 |
| 4 Nov 2016 | “After 20 years” update: the matrix is “too restrictive” | Berkus, 2016 |
| 5 May 2021 | Reply to a reader: divide the local average by five | Berkus, reply of 5 May 2021 |
| 23 Nov 2022 | Reply on inflation: $500,000 is “only a guide not a rule” | Berkus, reply of 23 Nov 2022 |
| 21 Nov 2024 | Repost: “nothing wrong” with higher maximums or different tests | Berkus, 2024 |
| July 2026 | Reply: the $500,000 is “completely arbitrary” | Berkus, reply of 5 July 2026 |
The 2016 update changed the most. Berkus wrote that “the original matrix is too restrictive, and should be a suggestion rather than a rigid form,” and that users should “assign maximum values to each that are appropriate to the industry and to the times.” He noted that the US average angel valuation was already above his cap, and gave an example: “in Silicon Valley, a ‘big data’ startup might competitively call for a $1.5 million maximum value per element, while the same startup in Nebraska might find $500,000 appropriate” (Berkus, 2016).
In 2021 a reader asked how to use the method in Indian rupees. Berkus answered: “To convert to another currency or location, first determine the average valuation of pre-revenue startups in your area. Divide by five and that should be the maximum you would pay for each of the items in the matrix” (Berkus, reply of 5 May 2021). The rule is easy to mistake for part of the 2016 update, but it is a reply in the comments, written five years later.
In a reply in July 2026, he went further and wrote that “the $500,000 per element number is completely arbitrary, dependent upon geographic location, type of business, and early proof of viability. In the end, it is what an investor will agree to that matters most” (Berkus, reply of 5 July 2026).
What is the $20 million revenue test?
The method has one filter that comes before any scoring. Berkus uses it only for companies he believes could, if successful, reach meaningful revenue in five years: “Today, for me, that hurdle number is $20 million” in gross revenue at the end of the fifth year (Berkus, 2016). It is a revenue threshold in dollars, and it decides whether the method applies at all. The scores never add up to it, and it caps nothing. Some guides describe it as a “soft cap of EUR 20m valuation in the 5th year”, which mixes up both the currency and the measure.
How do you choose a value inside each $0 to $500,000 range?
Berkus gives no scale for choosing a number below each maximum. He leaves it to the investor, “allowing the investor to put much lower values into each test” (Berkus, 2009). Most guides skip this step, and it is where most of the method’s subjectivity sits.
We suggest scoring each element against the evidence an outsider could check. This rubric is ours, not Berkus’s:
- 0%: nothing exists yet, or only the founders’ claim.
- 25%: early evidence that only the founders can vouch for.
- 50%: evidence an investor can verify, but the risk is still mostly open.
- 75%: strong outside evidence, and most of the risk on this element is gone.
- 100%: the risk this element covers is essentially retired.
Scores between the steps are fine. Write down the reason for each, so an investor can challenge a specific score rather than the whole number.
Berkus method example: how would a Dutch pre-seed startup be valued?
Take Glasveld, a hypothetical Dutch pre-seed company building sensors and software that tell greenhouse growers when to irrigate. Two founders work on it full time: a CEO with nine years at a greenhouse climate-systems supplier and an embedded-systems engineer. Neither has founded a company before, and nobody handles sales yet. Two of fifteen growers interviewed have run the prototype free for four months, and one says she would pay for it. The hardware is hand-built. A regional horticulture installer has signed a non-binding letter of intent to offer the system to its clients. There is no revenue.
First, the $20M test. The investor has to believe a company selling to greenhouse growers across Europe could pass $20M in revenue by year five, or Berkus would not use the method at all. We assume the investor does.
Scoring each element with the rubric above:
- Sound idea, 70%. Growers confirmed the problem and one says she would pay, which an investor can verify with a phone call. Nobody has paid yet, so product risk is not retired.
- Prototype, 60%. The system works in two real greenhouses, which is more than a lab demo. It is hand-built, so the technology risk of producing it in volume remains.
- Quality management team, 60%. Deep domain knowledge and the technical skill to build the product, both full time. No prior startup and no one yet responsible for sales.
- Strategic relationships, 40%. The letter of intent is real and checkable, but it is non-binding and there is only one partner.
- Product rollout or sales, 0%. Nothing has been sold, so this element earns nothing. That also keeps the result inside the $2M pre-revenue cap.
| Element | Score | Berkus standard ($500,000 max) |
|---|---|---|
| Sound idea | 70% | $350,000 |
| Prototype | 60% | $300,000 |
| Quality management team | 60% | $300,000 |
| Strategic relationships | 40% | $200,000 |
| Product rollout or sales | 0% | $0 |
| Total | 230% | $1,150,000 |
The arithmetic: 70% + 60% + 60% + 40% + 0% = 230%, and 230% × $500,000 = $1,150,000. As a check, $350,000 + $300,000 + $300,000 + $200,000 + $0 = $1,150,000.
Reading this as €1,150,000 would be a mistake. Berkus’s figures are US dollars set for the US angel market, and his own answer for other markets is the divide-by-five rule.
For comparison, our Checklist article scores Frostline, a US pre-seed company, under Berkus’s original caps and gets $975,000. That figure carries the Checklist’s operating-stage score over to rollout. With rollout at zero, as we suggest below, it would be $875,000.
What happens when you apply Berkus’s divide-by-five rule?
The rule needs “the average valuation of pre-revenue startups in your area”. There is no single official number, so we show two.
The first is the typical European pre-seed valuation reported in our H1 2026 Valuation Delta: $3.30M. Divided by five, that is $660,000 per element. The second is the average angel, pre-seed and seed round in the Netherlands in our July 2026 parameters, €6,540,000 (P6.3 table). Divided by five, that is €1,308,000 per element.
| Anchor (“area average”) | Per element (÷ 5) | Glasveld at 230% |
|---|---|---|
| Berkus standard | $500,000 | $1,150,000 |
| Valuation Delta, Europe, H1 2026: $3.30M | $660,000 | $1,518,000 |
| Valuation Delta, United States, H1 2026: $5.16M | $1,032,000 | $2,373,600 |
| P6.3 average, Netherlands: €6,540,000 | €1,308,000 | €3,008,400 |
The scores are identical in every row. Only the anchor changes, and the result runs from $1.15M to €3.0M. The last row is in euros, so read the table as a range of answers, not a currency conversion. The two Equidam anchors also measure different things. Valuation Delta figures are the results of full Equidam valuations for pre-seed companies preparing their first institutional round, not prices of closed rounds (Valuation Delta). The P6.3 average comes from closed angel, pre-seed and seed rounds (Data Sources), so it includes seed deals. An investor using Berkus’s rule has to pick one, and that choice moves the valuation more than any single score.
The US row shows the geography point Berkus made in 2016. The same company, with the same evidence, would come out 56% higher on the US Valuation Delta anchor than on the European one ($2,373,600 against $1,518,000), because the typical US pre-seed valuation is higher.
Have inflation and the market left the Berkus caps behind?
Inflation alone explains only part of the gap. Using the US consumer price index (BLS, CPI-U, series CUUR0000SA0), with 2009 and 2016 annual averages of 214.537 and 240.007 (select those years and tick annual averages in the BLS tool), and 334.980 for August 2026:
| Berkus figure | In 2009 dollars, restated for Aug 2026 | In 2016 dollars, restated for Aug 2026 |
|---|---|---|
| $500,000 per element | about $781,000 | about $698,000 |
| $2M pre-revenue cap | about $3.12M | about $2.79M |
| $2.5M post-rollout cap | about $3.90M | about $3.49M |
The typical US pre-seed valuation in Valuation Delta, $5.16M, is 2.6 times Berkus’s $2M pre-revenue cap and about 1.65 times that cap restated for inflation since 2009. The typical European figure, $3.30M, is only about 6% above the inflation-adjusted cap. On the Valuation Delta figure, inflation alone roughly explains the European gap. In the US, the market has moved much further than prices have. Divided by five, though, the European figure gives $660,000 per element, less than the $781,000 that inflation alone would suggest.
Berkus himself rejects inflation-indexing as the fix. When a reader argued in 2022 that $500,000 from 1999 would be worth about $900,000, he answered: “This is an art not a science, driven by so many factors that a single number like $500,000 is only a guide not a rule. Is a $500,000 value worth over $900,000 today? Not at all. Too many factors listed above have depressed startup values recently” (Berkus, reply of 23 November 2022). The European figure bears him out and the US figure does not. That is his wider point. The local market sets the number per element, and a price index cannot see it.
Where does the Berkus method break down?
The fixed caps are the first problem. Unless the user adjusts them, the method cannot produce a pre-revenue valuation above $2M, which is below the typical pre-seed valuation in every region in our H1 2026 Valuation Delta. Berkus’s 2016 and 2024 posts invite users to raise them.
Geography and sector are not in the original matrix. Berkus added both as manual adjustments: higher maximums for Silicon Valley than for Nebraska, and replacing a test to fit the business, such as “a drug development company needing various stages of FDA approval, substituted for ‘prototype'” (Berkus, 2024).
It only covers the pre-revenue stage. “Once a company is making revenues for any period of time, this method is no longer applicable” (Berkus, 2024). And it does not suit every business. In July 2026 Berkus wrote that “some types of businesses cannot easily fit into the Berkus Method of valuation. Lifestyle businesses and service enterprises come to mind” (Berkus, reply of 5 July 2026).
Scoring is subjective, and academic reviews say so. Montani, Gervasio and Pulcini describe “high subjectivity in the financial valuation of the five considered risk factors” (International Business Research, 2020), and Botello Velasco and González-Bueno conclude that the method is too subjective to give more than an idea of a possible valuation that reflects the investor’s preferences (Finance, Markets and Valuation, 2020).
The rollout element may not be worth what the matrix assumes. The Angel Capital Association found that “for Pre-Seed, those companies shipping product and earning revenues did not command higher valuations compared to those earlier in the cycle” (ACA, 2024). It is one dataset, but enough to question an automatic jump from $2M to $2.5M.
Divide-by-five has a ceiling problem of its own. If the whole matrix tops out at the local average, a company that scores 100% on every element can only ever be worth an average company. Before revenue it tops out at four-fifths of the average.
How is the Berkus method related to the Checklist method?
Our Checklist method keeps Berkus’s structure, additive blocks under a ceiling, but sets the ceiling from recent funding rounds in each country and splits it across five weighted criteria. The Checklist article explains the differences and runs both on the same company.
Berkus vs Scorecard, Checklist, Risk Factor Summation and First Chicago: what is the difference?
| Method | Starting point | How it adjusts | Limit on result | Fits best | In Equidam’s five? |
|---|---|---|---|---|---|
| Berkus | Zero | Up to $500K per element, 5 elements | $2M pre-revenue, $2.5M after rollout | Pre-revenue only | No |
| Scorecard | Regional average valuation | Weighted scores on 6 criteria | None built in | Pre-revenue and early stage | Yes |
| Checklist | Country maximum from recent rounds (US $17M) | Share of the maximum earned on 5 weighted criteria | The maximum | Pre-revenue and early stage | Yes |
| Risk Factor Summation | Regional average valuation | ±$250K per step on 12 risks | None built in | Pre-revenue, as a cross-check | No |
| First Chicago | Scenario forecasts | Probability weights on best, base and worst case | None built in | Companies with forecastable outcomes | No |
The Checklist maximum is from the P6.3 table. Risk Factor Summation figures are from Bill Payne’s Flathead Beacon column (2011), and the First Chicago description follows the UC San Diego startup toolkit. Our guide to valuation methods to use and avoid covers the wider family.
How should a founder use the Berkus method?
Use it as a map of the risk you have retired, not as your asking price. Scoring yourself honestly against the five elements shows where the next milestone adds the most. For Glasveld, turning the letter of intent into a paid contract and designing the hardware for volume production would move two of the lowest scores.
For the number itself, compare it with a full Equidam valuation. There, a pre-revenue company with a prototype is typically at the development stage, where the Scorecard and Checklist methods carry 30% each and the VC method 16% (sample report, appendix p.20). The VC method works backwards from a possible exit, discounted at a required return that is 76.32% a year at the development stage (Help Center: required ROIs). If your Berkus number and those three disagree sharply, find out why before an investor asks.
If you use Berkus in a conversation with investors, say which maximum per element you used and where it came from. Take the anchor from your own market and name it. On our H1 2026 Valuation Delta figures, a US anchor would lift a European company’s result by about 56%, and on every anchor we show, Berkus’s original $500,000 gives the lowest number. And score rollout at zero until something has been sold.
Our guide to valuing pre-revenue startups covers the rest of the toolkit.
What else do people ask about the Berkus method?
Is the Berkus method still used?
Yes, mostly by angels as a quick first view of a pre-revenue company, and in adapted form it runs inside Equidam valuations as the Checklist method. Berkus called it “fairly universal as it has become” in July 2026 (Berkus, reply of 5 July 2026). With the original $500,000 slots, it produces figures well below today’s pre-seed valuations.
Berkus or Scorecard: which should you use?
Both, if you can. Berkus builds a value from zero out of retired risks, while the Scorecard method adjusts the average recent round in your region, so it needs no separate inflation or geography fix. Bill Payne, who devised the Scorecard method, advised angels to “use multiple methods” (Flathead Beacon, 2011).
Can you use the Berkus method for SaaS, biotech or service businesses?
Software fits the five elements as they are. For biotech, Berkus suggests replacing the prototype test with regulatory milestones (Berkus, 2024). Lifestyle and service businesses, he says, do not fit.
Can the Berkus method give a precise valuation?
It is not designed to. Berkus calls it “an art not a science”, and the result depends on the maximum you choose and on your own scores. Its value is that every part of the number can be seen and challenged.
What is the next step?
Inside an Equidam valuation, the Berkus idea lives on in the Checklist method, next to the Scorecard, the VC method and two DCF methods, using data from more than 160,000 companies valued (Equidam internal data) and 30,000+ public comparables across 90+ countries and 600+ industries (Data Sources).
To see which traits feed the Checklist method, the Help Center guide to the Checklist lists them. When you want all five methods run with the current anchors for your country and a report you can share with investors, you can compare Equidam plans.