The short answer to how to value a startup is that no single method produces a correct price. Startup valuation methods are frameworks for organizing assumptions, and the right one depends on the company's stage, the quality of its financial history, and the purpose of the valuation. Pre-revenue companies lean on qualitative methods such as the Scorecard method and the Berkus method. Revenue-generating companies can anchor on comparables. Companies with credible forecasts can add the venture capital method or discounted cash flow. The practical move is to run two methods that read the evidence from different angles, compare the outputs, and defend a range rather than a point. This article explains the Scorecard method, the Berkus method, the venture capital method, comparables and DCF, works one example of each, shows which stage each fits, names each one's weakness, and gives a sanity-check sequence you can run before you invest or raise at a stated pre-money valuation.
What Pre-Money Valuation Actually Means
Before comparing methods, fix the arithmetic they feed. AngelList defines a pre-money valuation as the value of a company before a new outside investment. Post-money valuation adds the new capital. Investopedia states the relationship as: Pre-Money Valuation = Post-Money Valuation - Investment Amount. So a company raising $1 million at a $6 million post-money valuation has a $5 million pre-money valuation.
That identity is also how ownership is checked in a simple priced round. If an investor puts in $2 million for 20% post-money ownership, the post-money valuation is $2 million divided by 20%, which is $10 million, and the pre-money valuation is $8 million. Existing holders collectively keep 80% immediately after that simplified round. This matters because founders often meet valuation through an ownership proposal rather than a stated number: size divided by desired percentage gives the implied valuation even when nobody says it out loud.
Keep two separate questions apart. The pre-money and post-money identity is definitional arithmetic and it holds for a simple priced round. Ownership allocation is a separate calculation, and it is the one that gets complicated. A pre-money valuation is a judgment figure rather than a measurement: it can rest on financial performance, comparable transactions, the stage of the business, and the perceived strength of the founding team, among other inputs. Then the instruments enter. Option-pool increases, SAFEs and convertible notes all affect the ownership that a given headline pre-money produces, and this article does not explain how they are treated. The supplied sources do not establish those mechanics, and restating them from memory would put unsupported rules into a piece that asks you to check its arithmetic. The working point stays narrow, and it is worth holding onto anyway: the same stated pre-money can produce different ownership outcomes for the same amount invested, so the terms have to be read next to the number, and the simple identity above should not be extended to instruments it was not derived for.
The Scorecard Method Explained
The Scorecard method values a startup relative to a benchmark. It takes a benchmark pre-money valuation from recent comparable early-stage rounds, then adjusts that baseline using weighted factors on which the company is stronger or weaker than the comparison group. The supplied Waveup excerpt supports the final calculation of its published example, but it does not publish a factor list or a set of weights, so the factors and weights below are this article's own working set, stated explicitly so the arithmetic can be checked. The factors used here are: strength of management team, size of opportunity, product or technology, competitive environment, marketing, sales channels and partnerships, need for additional investment, and other. Each factor needs a weight, and the weights are judgment choices that must be stated so a reviewer can reproduce the result. The weights used in the example below are fixed values, not ranges: management 30%, opportunity 25%, product 15%, competition 10%, sales 10%, financing 5%, other 5%. They sum to 100%.
Worked example, hypothetical and labeled as such: a pre-seed SaaS company uses a benchmark pre-money of $5 million, taken as the median of regional pre-seed rounds. Scores against the baseline are management 1.25, opportunity 1.5, product 1.0, competition 0.75, sales 1.0, financing 1.0, other 1.0. Weighted at the fixed weights above, that gives 0.375 + 0.375 + 0.15 + 0.075 + 0.10 + 0.05 + 0.05, a sum of 1.175. Pre-money = $5M × 1.175 = $5.875M. Waveup's published example reaches the same arithmetic: Sum = 1.175. Pre-money = $5M × 1.175 = $5.875M.
Best fit: pre-revenue and seed companies with enough context for relative comparison. Weakness: the benchmark carries straight into the answer. A weak or mismatched regional median produces an irrelevant result even when the scoring is careful, and the factor weights remain judgment calls. Source the median from real transactions, footnote it, and never invent it. State which statistic you used. A mean and a median of the same comparison set are different statistics, and while they can land on the same figure, they often do not, so a reader who cannot see which one was chosen cannot reproduce the result.
The Berkus Method Explained
The Berkus method assigns a monetary value to each qualitative success factor and adds them up. The five factors used here are basic value, technology, execution, strategic relationships in the core market, and production and consequent sales. The supplied Corporate Finance Institute passage establishes the conventional sizing, not the category names: it describes an approach that normally allocates up to $500,000 per success factor for a theoretical maximum pre-money valuation of $2.5 million. The category list above is this article's working description of the method, and it should be read that way.
Worked example, hypothetical: a pre-revenue B2B SaaS founder with a working MVP, two warm strategic-partner LOIs, a strong founding team, and no paying customers yet might score $400K (idea) + $400K (prototype) + $450K (team) + $300K (partners) + $0 (sales) = $1.55M pre-money.
Best fit: the first angel check or friends-and-family round, where there is no revenue to anchor on and no dependable financial history. Weakness: the factor values are judgment-based, so two reviewers can land far apart on the same deal. The $500,000-per-factor allocation and the $2.5 million maximum describe the conventional capped version of the method as Corporate Finance Institute presents it. They are a sizing convention, not a universal ceiling on what the method can be asked to estimate, and a version built with different factor values simply produces a different, and more judgment-heavy, result. The useful part is the structure rather than a fixed dollar cap carried over from another market or era. Define the categories before choosing their values so the model follows the evidence rather than a desired total.
The Venture Capital Method Explained
The venture capital method runs the calculation backward from the investor's exit. Focal states the mechanism as: Present post-money valuation = expected exit value ÷ required return multiple. Subtract the new investment and you have the implied pre-money valuation. The visible inputs are an expected exit value, a required return multiple, and an investment amount. Two further quantities sit behind them and need explicit treatment. The time horizon is embedded in the required return: a 10× requirement over a five-year hold and a 10× requirement over an eight-year hold are different assumptions, so state the hold period next to the multiple. Dilution is not in the simple formula at all: the calculation prices the entry round only, and later financing rounds reduce the investor's ownership without changing this arithmetic.
Worked example, hypothetical: assume an expected exit equity value of $80 million, a required return of 10× over a hypothetical five-year hold, and a $2 million investment. $80 million ÷ 10 = $8 million post-money. $8 million − $2 million = $6 million implied pre-money. Hold the return requirement, the five-year hold period and the investment constant and vary only the exit: a $60 million exit gives $6 million post-money and $4 million pre-money; a $100 million exit gives $10 million post-money and $8 million pre-money. The illustrative range is $4 million to $8 million pre-money, and the spread comes entirely from the exit assumption. The simplifications are part of the result. The exit figure is an equity value at exit, so no cash or debt bridge is needed. The required return is treated as a total multiple over the stated hold period rather than an annual rate. Future rounds are ignored, so the implied pre-money is the entry price before any subsequent dilution. And one exit value stands in for a whole distribution of outcomes, including outcomes below the invested amount.
Best fit: Waveup's stage guide places the venture capital method at seed with early traction, and pairs it with comparables at Series A, which is where an articulated exit case starts to sit alongside real revenue. Weakness: the exit value and the required return dominate the result, so a desired valuation can be reverse-engineered from an optimistic exit story. Neither the return multiple nor the exit value is a universal rule, so both must be stated as assumptions rather than facts, and the hold period belongs beside them.
Comparables and Revenue Multiples Explained
Comparables value a company by applying a multiple observed for similar companies or transactions to its own operating metric. For a recurring-revenue business the basic form used here is Valuation = ARR × revenue multiple, where ARR is annual recurring revenue measured consistently with the comparison set. The supplied sources do not publish that formula or the enterprise-to-equity bridge used below, so both are stated as standard conventions this article applies rather than as quoted source text. Before the multiplication, fix the basis. Revenue multiples circulate on an enterprise-value basis and on an equity-value basis, and the two are not interchangeable. State which basis your observations use, state the measurement date of each observation, and state whether the metric is historical or forecast. Mixing any of those breaks the comparison silently.
Worked example, hypothetical: a company with $600,000 ARR and a defensible illustrative multiple range of 7× to 10× gives $600,000 × 7 = $4.2 million and $600,000 × 10 = $6.0 million. Label the output: under this illustration the peer observations are enterprise values, so the result is an enterprise value, not a pre-money equity figure. Bridge before comparing it with anything: equity value = enterprise value + cash − debt. One turn of multiple moves the answer by $600,000, which shows where the real argument sits: not the multiplication, but whether the peer set justifies the upper end.
Best fit: revenue-generating companies with genuinely relevant market references, especially Series A onward. Weakness: sparse, stale or poorly matched comparisons make the range misleading. Screen peers for sector, stage, geography, business model, scale, growth and margins, and record the source, observation date and matching rationale for every observation. One caution is worth keeping in view: a comparable records what someone else paid for a different business at a different moment. It does not tell you what this business is worth.
Discounted Cash Flow Explained
Discounted cash flow forecasts the cash a business generates and discounts each period's amount to present value at a rate reflecting the risk of those cash flows. The expression used below, Value = Σ [FCFₜ ÷ (1 + r)ᵗ] + [Terminal value ÷ (1 + r)ⁿ], is a standard discounted-cash-flow formula stated here as this article's own statement of the method; the supplied sources do not publish it. FCFₜ is free cash flow in period t, r is the discount rate, and terminal value stands for cash flows beyond the explicit forecast period. Both bases need stating. The cash-flow basis in the example is unlevered free cash flow: cash generated by operations after operating costs and capital expenditure, measured before interest, dividends, debt repayments and any new financing. Because financing flows are excluded from the measure, the discounted result is an enterprise value, and cash and debt are bridged afterward to reach equity value. The discount-rate basis matters just as much, because r has to reflect the risk of these specific cash flows, and for an early-stage forecast there is no observed market rate to read off. The rate is an assumption, and it belongs in writing next to the answer.
Worked example, hypothetical and deliberately limited. A company is forecast for five years with free cash flow of −$1.0 million in year one, $0 in year two, $1.0 million in year three, $2.0 million in year four and $3.0 million in year five. The discount rate is 30%, used here as an illustrative rate for a risky forecast rather than a sourced figure. Terminal value is set to zero, which is a deliberately conservative assumption: the model values only the five explicit years and treats the business as if it produced nothing after year five. The present values are −$1.0M ÷ 1.30 = −$0.77M; $0 ÷ 1.30² = $0; $1.0M ÷ 1.30³ = $0.46M; $2.0M ÷ 1.30⁴ = $0.70M; $3.0M ÷ 1.30⁵ = $0.81M. Sum: about $1.19 million of enterprise value. Bridge to equity: with $0.5 million of cash and no debt, equity value is about $1.69 million, and with no other claims outstanding that is the pre-money value before any new investment. The scope is narrow on purpose, and it is stated with the number: no terminal value, no financing after year one, and a discount rate chosen rather than observed.
The zero-terminal-value assumption is what keeps the answer small. Add a terminal value of $5 million at the end of year five and its present value is $5M ÷ 1.30⁵ ≈ $1.35 million, which lifts enterprise value from about $1.19 million to about $2.54 million and equity value to about $3.04 million. One assumption added after the forecast moves the answer by more than the entire explicit forecast produced. The example also assumes the plan is funded without further equity raises. If the company raises additional equity before year five, the new shares dilute existing holders, and that dilution is not in the arithmetic above. The supplied sources do not establish how other forms of additional capital would be treated, so the point stays at that conditional.
Best fit: companies with enough operating history to support explicit forecasts. Waveup's guide states that DCF is the right primary method at Series B and later, when you have 3+ years of revenue history and a defensible margin path. That is one guide's stage preference, not a universal instruction for every company before Series B. The real test is whether the forecast can be defended line by line, and revenue on its own does not make a forecast defensible. Weakness: small changes to growth, the discount rate or the terminal value materially change the answer, as the two variants above show. Where the history is not there yet, treat DCF as a cross-check rather than the anchor.
Which Method Fits Which Stage
Choose the primary method from the strongest evidence the company has today, then pick a cross-check that measures value from a different angle. The useful question is not which method is most advanced, but which requires the fewest unsupported assumptions for this company.
Waveup's stage guide gives one working map. Pre-revenue, from idea to MVP: Berkus plus Scorecard. Seed with early traction: risk factor summation plus the venture capital method. Series A with real revenue: the venture capital method plus comparables. Series B and later, with 3+ years of revenue history: DCF plus comparables. Where credible peer observations exist, comparables become the anchor and a qualitative method or an investor-return model works as the cross-check. Where an exit path is articulated, an investor-return case can be added as a third lens.
The transition is gradual rather than a switch triggered by a financing label. Qualitative factors still explain why a company sits above or below a comparable range once revenue arrives, and revenue does not automatically make a forecast reliable. A replacement-cost cross-check can be added where the cost of rebuilding the capability is informative. And a detailed DCF built on fragile forecasts can be less useful than a disciplined qualitative assessment, because the spreadsheet inherits every weak assumption fed into it.
The Weaknesses That Decide the Argument
Each method fails in a specific, predictable way, and naming it is what makes a valuation defensible.
Scorecard: the benchmark and the weights are subjective. A mismatched comparison group transfers directly into the result. Berkus: factor values are judgment-based and create false precision if treated as universal, and the conventional $500,000-per-factor allocation caps that version of the method rather than the method in general. Venture capital method: the exit value and required return dominate, so a desired valuation can be reverse-engineered from an optimistic exit story. Comparables: peer selection and observation dates decide the range, and mixing enterprise values, equity values, forecast metrics and historical metrics breaks the comparison. DCF: growth, discount rate and terminal value are material sensitivity drivers, and a spreadsheet cannot remove that uncertainty.
Equidam states that there is currently no perfect method to assess a startup's value and that each model discussed has significant limits. Read as Equidam's own conclusion, it argues for triangulation rather than for abandoning the exercise.
How to Sanity-Check a Valuation Before Investing or Raising
A sanity check is a sequence, not a single calculation. Run it in this order.
First, normalize the outputs. Label every result as enterprise or equity value, and as pre-money or post-money, with the same measurement date and financing amount. Apply the cash and debt bridge wherever an enterprise value has to be compared with an equity figure. Second, choose the primary anchor: the method resting most directly on current credible evidence. Third, add a genuinely different cross-check, not a second version of the same assumption. Fourth, investigate disagreement instead of averaging it away: is the gap coming from peer selection, revenue definition, growth forecasts, exit assumptions, required return, or dilution? Fifth, define lower, base and upper cases and state what changes between them. Sixth, translate the range into financing economics: how much capital, what milestones it funds, how much ownership moves, and on what terms.
Two practical checks follow. The first is a numerical gap test. The supplied sources do not establish any tolerance, so the bands below are this article's own working heuristic rather than a published standard, and the denominator has to be stated whenever they are used. Define the gap as the difference between the higher and lower output divided by the primary anchor: gap = (higher − lower) ÷ primary anchor. The anchor must be a positive, nonzero number; if the primary anchor is zero or negative the ratio is not interpretable and the test cannot be run. On that denominator, treat the bands as review triggers rather than verdicts. A gap of 25% or less is a prompt to carry the outputs as one band while still confirming that both rest on the same definitions and dates. A gap between 25% and 50% is a prompt to revisit at least one set of assumptions before a number is quoted. A gap above 50% is a prompt to identify which assumption set is driving the divergence, starting with the exit assumption and the peer selection. A narrow gap does not establish that the valuation is reliable, because two methods can agree while sharing the same wrong assumption. A percentage gap without a named denominator is not interpretable.
The second check is the ownership math, run independently of any valuation model. Investment divided by post-money equals the investor's ownership share, and pre-money divided by post-money equals the collective share of the existing holders immediately after a simple priced round, on the stated assumption that those holders own all pre-round equity and no new option pool is created at closing. Change that assumption and the retained share changes even though the headline pre-money does not.
Also pressure-test the price against the plan. First Citizens warns that a high valuation comes with high expectations, and that missing the milestones needed to justify an inflated valuation can put a company into a down round. Evaluate valuation alongside four linked decisions: how much capital the company needs, what milestones that capital can realistically fund, how much ownership the round transfers, and what terms accompany it. Finally, remember that practice varies: Equidam reports a survey of 885 institutional venture capitalists at 681 firms by Gompers, Gornall, Kaplan and Strebulaev, in which 63% use a cash-on-cash multiple and only 22% use NPV methods. The supplied excerpt does not give the study's year or publication details, so the figures are cited here as Equidam reports them. Knowing how the other side may reason is part of the check.
A Worked Reconciliation Across Three Methods
The point of running more than one method is the reconciliation, not the average. Consider a hypothetical early-stage AI software company with a working product, a small base of paying customers, and $600,000 in ARR, raising $2 million, with short retention history and material customer concentration. Every figure is illustrative. Normalization comes first: the revenue multiple as applied here produces an enterprise value, the qualitative model and the venture capital method produce equity values, and all three are stated pre-money, before the $2 million. The illustration assumes the company holds $300,000 in cash before the round and no debt.
The qualitative model is this article's own illustrative construction, not a published model. It borrows the five-factor structure described above and departs from the conventional capped version in two ways: the factor values exceed the $500,000-per-factor convention that Corporate Finance Institute describes, and it applies two explicit risk deductions after the factor sum. Factor values: basic value, meaning the business concept, $1.2 million; technology, meaning product and technical progress, $1.4 million; execution, meaning team and delivery capability, $1.6 million; strategic relationships $0.8 million; production and consequent sales, meaning early traction, $1.0 million. Total $6.0 million. Deductions: $500,000 for customer concentration and $300,000 for short retention history, giving $5.2 million. The output is treated as a pre-money equity estimate, so no cash or debt bridge is applied; state that choice, because a factor set calibrated to operating assets would need the same bridge as the multiple. Its weakness is visible in the construction: the factor values and the two deductions are analyst-set, so the number records judgments rather than observations. Focal's supplied excerpt reports a resulting qualitative estimate of $5.2 million pre-money for a case, without disclosing component values; the breakdown above is this article's own and should not be read as Focal's.
The revenue-multiple lens: $600,000 ARR × 7× to 10× gives an enterprise value range of $4.2 million to $6.0 million. Bridge it: equity value = enterprise value + cash − debt, so $4.2M + $0.3M − $0 = $4.5 million at the low end and $6.0M + $0.3M = $6.3 million at the high end. Applying the multiple to current ARR does not by itself make the result a pre-money figure. That classification depends on the valuation date and on how the new money is treated. Here the example states the treatment explicitly: ARR is measured as of the valuation date before the round closes, and the $2 million of proceeds are excluded from the metric and from the $300,000 cash balance used in the bridge, so the result is a pre-money equity range of $4.5 million to $6.3 million. If the metric were measured after closing, or if the proceeds already sat in the bridged cash balance, the classification would change. If the peer observations had been equity values, no bridge would be needed, but the basis would still have to be stated.
The venture capital method: an $80 million exit equity value ÷ a 10× required return over a hypothetical five-year hold = $8 million post-money. Subtracting the $2 million investment gives $6 million pre-money. Varying only the exit, with the five-year hold and the 10× requirement held constant, gives $4 million pre-money at a $60 million exit and $8 million pre-money at a $100 million exit, an equity-basis range of $4 million to $8 million.
Read together: the comparables range and the venture capital range overlap between $4.5 million and $6.3 million, and the qualitative estimate of $5.2 million sits inside that overlap. That overlap is the working band. The lower revenue scenario is not dropped; it is converted: the $4.2 million figure is an enterprise value, and carrying it into a pre-money comparison without the $300,000 cash bridge would understate the equity claim by that amount, so after the bridge it becomes the band's floor rather than a competing number. The lower exit scenario, $4 million pre-money from a $60 million exit, falls below that floor, so it is retained as a downside case instead of part of the band. It becomes the operative number only if the evidence behind a specific assumption changes, and the path has to be traced: if the retention evidence stays thin, the forecast that supports an $80 million exit weakens, which lowers the exit assumption and therefore the venture capital method's output; if the peer set weakens, the comparable range moves down with it. A weak retention record does not move either number by itself. The upper exit scenario, $8 million pre-money from a $100 million exit, sits above the comparables ceiling of $6.3 million, so it is an upside position that requires defended exit assumptions. The band is an overlap, not an average, and an investor can challenge the factor values, the peer set or the exit case without rejecting the whole analysis. The final price is still negotiated; it reflects evidence, capital needs, return requirements, demand for the round, market conditions and terms.
What Valuation Arithmetic Cannot Tell You
A valuation model organizes assumptions. It does not verify them, and it does not tell you whether the business behind the numbers is sound. The quality of any of these five methods rests on inputs that valuation work does not test: whether there is a credible path to paying customers, whether the execution plan matches the available resources, and whether the economics improve as the business grows.
That is where a structured company analysis fits, before you commit capital or publish a number. DDScore turns private-company materials into a structured 0 to 100 Due Diligence Score and a written report across twelve fixed dimensions, cross-checking submitted materials against each other and against current public information. It surfaces strengths, areas for development, material risks, evidence limitations and analysis confidence, plus questions tied to the evidence gaps that matter for further human review. A report generally takes 10 to 30 minutes to produce.
Use it as an input to the valuation conversation, not as a substitute for it. It does not provide investment advice or replace full diligence or human judgment, so the price, the range and the decision stay with you. Keep it out of the pricing role entirely: it is a structured read of the evidence, not a valuation model, and a score is not a price. What it can do is show which assumptions in your model rest on thin evidence: the traction claim with no supporting document, the market claim contradicted by current public information, the financial plan whose inputs do not reconcile. For a founder, that means seeing the questions an external reviewer may raise before the room raises them. For an investor, it means entering the discussion with the material risks and evidence gaps already separated from the unanswered questions. The valuation range is then built on a case you have actually examined.
Explore DDScore for Founders · Explore DDScore for Investors