How a company is valued.
There is no such thing as the value of a company. There is a reference range the market will pay, a set of buyer-specific values above and below it, and a number you need – and none of them is the first inbound offer. This chapter is the methods professionals use: what each one measures, what each one hides, how they are used together, and what actually moves a software multiple.
- A valuation is a range produced by several methods that check one another; a buyer’s willingness to pay is a decision. They are related, not identical.
- Financial reference range, strategic value, the number you need and the first inbound offer are four different numbers – and what reaches shareholders is a fifth.
- Every multiple is applied to an earnings or revenue figure a buyer will rebuild – run that analysis first.
Every method answers a different question. None answers “what is it worth?”
Valuation standards recognise three approaches: the market approach (what similar things sell for), the income approach (what the future cash is worth today) and the cost approach (what it would cost to build again).1 In practice six methods do the work: two from the market approach, one from each of the others, and two transaction-specific analyses – the financial buyer’s return model and the early-stage methods – that are not valuations of the company so much as of a plan for it.
- Trading comparables – what listed peers are priced at.
Enterprise value divided by revenue or by EBITDA, for a set of listed companies like yours, applied to your numbers. Fast, current and widely used: in a survey of valuation professionals, 84% said they always or almost always use EV/EBITDA when they use multiples.2 What it hides: the peers are usually larger, more liquid and more diversified, and a trading multiple contains no premium for control. A multiple is a price, not a value; it tells you what a market is paying today for companies that are not yours.
- Precedent transactions – what buyers have paid for companies like yours.
Multiples from completed acquisitions of comparable private or public companies. They include the premium buyers pay for control, which is why they usually sit above trading multiples. What they hide: the deals are old by the time they are known, the disclosed ones are a biased subset, and each price carried a structure – earn-outs, stock, escrow – that the headline multiple ignores. Unlisted companies have historically changed hands at discounts of 15–30% to comparable listed acquisitions – a discount that Officer attributes largely to sellers’ need for liquidity, and Koeplin and colleagues partly to the fundamentals that keep companies private.3 Premiums for control have historically averaged 20–30% for listed targets, and Damodaran rejects any rule of thumb: the value of control depends on how much better the company could be run.4
- Discounted cash flow – what the future is worth today.
Forecast the cash the company will generate, discount it at a rate that reflects its risk, add a terminal value for everything after the forecast. The method that values a plan rather than someone else’s price, and the one most naturally adapted to model a specific buyer’s cash flows and synergies explicitly. What it hides: the terminal value commonly makes up the majority of the result – three-quarters in Damodaran’s stylised example, and anywhere from 56% to more than 100% in McKinsey’s own textbook cases – so a DCF is only as good as its assumptions about years nobody can see.5
- Leveraged-buyout analysis – the financial buyer’s ceiling.
Not a valuation of the company but of a fund’s plan for it: given the debt available, the exit assumed and a target return of about 25% a year and 2.5 times money, what is the most a fund could pay? Under those assumptions the answer is the financial buyer’s ceiling; change the assumptions and the ceiling moves. The arithmetic, worked →
- Replacement cost – build or buy.
What it would cost a strategic buyer to reproduce your product, team and customer base, and how long it would take. Under valuation standards, a buyer will pay no more for an asset than the cost of obtaining one of equal utility;1 in technology, the time is often worth more than the money, and some customer positions cannot be rebuilt within the time available or at acceptable risk. The method rarely produces the number, but it produces the argument.
- The venture method and its cousins – for companies too early for the others.
Forecast an exit value, discount it at a rate that reflects the odds of getting there at all, or score the company against funded peers. Useful below the revenue at which multiples mean anything; largely irrelevant to a company worth ten million euros or more, except as a reminder that early investors on your cap table were valued this way and hold preferences accordingly.
The multiple values the business. You are paid for the equity.
An enterprise-value multiple produces an enterprise value: the value of the operating business to all its capital providers. What reaches shareholders is what remains after debt, debt-like items and the working-capital adjustment – and then, if the company has investors, after the waterfall. Illustrative:
| Step | Illustrative | What decides it |
|---|---|---|
| Enterprise value | 8.0× €5.0M EBITDA = €40.0M | The multiple and the definition of EBITDA it is applied to |
| Plus cash | + €1.5M | “Cash-free, debt-free”: you keep the cash, you settle the debt |
| Less debt | − €7.5M | Bank debt, shareholder loans, leases, accrued interest |
| Less debt-like items | − €2.0M | The contested list: deferred revenue, accrued bonuses, transaction costs, tax, pension deficits |
| Equity value | = €32.0M | Before the working-capital adjustment against the peg, and before the waterfall |
The argument is almost never about the multiple alone. It is about what counts as debt. Deferred revenue is the sharpest example in software: the buyer argues it is debt-like, because the cash has been collected and the service still has to be delivered at the buyer’s cost; the seller argues it is ordinary working capital; the usual settlement treats part of it as each.6 Leases moved onto the balance sheet under IFRS 16 raise EBITDA and add a liability at the same time, so the EBITDA definition and the debt definition have to be set together or the price is wrong in one direction.6 A founder who has not decided these definitions before the letter of intent is likely to discover them in the sale agreement, one at a time, mostly as reductions. The price mechanism, in the structure chapter →
Growth, retention, margin, scale and cash generation interact – and not linearly.
For recurring-revenue software, listed-company multiples are an important market reference. They are neither a private-company price list nor a ceiling. Median revenue multiples of listed SaaS companies at the end of 2025, grouped by the three metrics buyers price first:7
| Revenue growth | EV / revenue | Net revenue retention | EV / revenue | Gross margin | EV / revenue |
|---|---|---|---|---|---|
| Up to 10% | 2.4× | Below 90% | 1.5× | Below 60% | 3.3× |
| 10–20% | 5.8× | 90–100% | 2.6× | 60–70% | 4.1× |
| 20–30% | 12.7× | 100–110% | 4.5× | 70–80% | 5.1× |
| Above 30% | 8.5× | Above 110% | 8.0× | Above 80% | 6.4× |
Listed medians frame the conversation; they are not the price. Across the 2,698 software acquisitions recorded in 2025, the median on deals with disclosed terms was 4.0 times trailing revenue while the average was 6.9 – a market shaped like a barbell, with a small number of exceptional companies commanding exceptional prices and a long tail that did not.7 Size matters on its own: one European adviser’s analysis of disclosed software deals finds the median revenue multiple for $50–100 million transactions roughly twice that for $20–50 million ones.8 Across all sectors, private-equity buyers of companies worth $10–500 million paid about 7.2 times EBITDA on average in 2025.9
The metrics themselves have benchmarks. In a 2025 benchmark of 583 private B2B software companies, the medians were 26% annual recurring-revenue growth, 101% net revenue retention, 88% gross retention and an 81% subscription gross margin;10 the “Rule of 40” – growth rate plus profit margin should exceed forty – began as one investor’s remark at a board meeting in 2015 and has become a shorthand many buyers apply first.11 Multiples move by the quarter; these are the shape of the relationship, dated to the sources in the footnotes, and the reason a buyer’s first question is about your retention rather than your revenue.
A valuation is a range. A strategic value is a different number. The first offer is an anchor.
Professionals present valuation as a “football field”: the range each method produces, side by side, so that the methods check one another and the overlap becomes the defensible range. What the chart cannot show is the second lens – what specific strategic buyers could justify, mechanism by mechanism, above or below the reference range. That dispersion is where the same company shows materially different values to different owners, and it is the reason a process is run rather than a valuation simply announced. Strategic Value Dispersion →
The anchor is the practical hazard. Anchoring is among the most replicated findings in judgment research: estimates are biased toward whatever number was named first, even a number known to be arbitrary.12 In negotiation experiments, whichever side made the first offer obtained the better outcome – and the advantage disappeared when the other side focused on the first mover’s alternatives, or on its own target.12 That is a large part of the case for holding a valuation you can defend before a buyer names a number: not to win the argument about the anchor, but to be immune to it.
The number underneath the number.
Every multiple is applied to an EBITDA or a revenue figure, and among the first things a buyer’s advisers do is rebuild that figure. A quality-of-earnings analysis asks how sustainable, predictable and cash-backed the earnings are: which revenue recurs and which was a one-off, which costs are normal and which were the owner’s, how earnings convert into cash, what working capital the business actually consumes.13 It is transaction-focused and forward-looking, which is why an audit does not replace it and why buyers commonly commission one even when the accounts are audited.
The usual adjustments run both ways. Owner compensation is normalised to a market rate; genuine one-offs are removed; a contract signed in month nine may be shown at its run-rate – though expect a buyer to accept it, if at all, as a separate pro-forma line rather than inside adjusted EBITDA. Capitalised development cost is a recurring fight in software: buyers argue that capitalisation hides cash the business spends every year and deduct it from EBITDA; sellers argue that their listed peers capitalise too. Whoever runs the analysis first sets the terms of that argument. Of the letters of intent that broke on one deal platform in 2025, a fifth broke on the gap between the seller’s earnings and the buyer’s quality-of-earnings work.14 Due diligence, in the next chapter but one →
Within the first month of a mandate, once your documents are in, we build the full valuation: comparable multiples, precedent transactions, discounted cash flow, and where each likely buyer sits in that range. Institutional grade, defended, and yours to keep. How the valuation changes the negotiation →
1 International Valuation Standards, IVS 105 Valuation Approaches and Methods: market, income and cost approaches; the cost approach rests on the principle that a buyer will pay no more than the cost to obtain an asset of equal utility.
2 Mukhlynina & Nyborg, “The Choice of Valuation Techniques in Practice”, Critical Finance Review 9, 2020: survey of valuation professionals; 84% always or almost always use EV/EBITDA when they use multiples; roughly half prefer multiples and half discounted cash flow.
3 Officer, “The price of corporate liquidity: Acquisition discounts for unlisted targets”, Journal of Financial Economics 83(3), 2007: average discounts of 15–30% for unlisted targets against comparable listed acquisitions; discounts widen when the seller needs liquidity. Koeplin, Sarin & Shapiro, Journal of Applied Corporate Finance 12(4), 2000: 20–30% lower earnings multiples for US private acquisitions. US samples from the 1980s–2000s; an association, not a rule for your company.
4 Damodaran, The Value of Control (NYU Stern): historical US premiums of 20–30%, and the argument that no standard premium can be justified. McKinsey (2004) cites 10–35%.
5 Damodaran, “Myth 5.5: The Terminal Value ate my DCF” (2016): 75% of value in the standard example; Koller, Goedhart & Wessels, Valuation, 6th ed., exhibit 12.1: continuing value from 56% to 125% of total value across four illustrative companies with an eight-year forecast.
6 Grant Thornton Belgium, Impact of debts on the purchase price (2022): the worked bridge above is adapted from its example; the three tiers of debt-like items are its classification; Deloitte (2019) and Baker Tilly (2025) on aligning IFRS 16 EBITDA with the net-debt definition. Practice, not a market statistic.
7 Software Equity Group, 2026 Annual SaaS Report (March 2026): 108 listed SaaS companies, medians at the fourth quarter of 2025 (overall median 5.3× trailing revenue); the above-30% growth bucket is small, which is the likely reason it sits below the 20–30% bucket. M&A: 2,698 transactions in 2025, median 4.0× and mean 6.9× trailing revenue on the subset with disclosed terms. Listed and disclosed data; not a price for a private European company.
8 Aventis Advisors, SaaS Valuation Multiples 2015–2026 (updated August 2026, Mergermarket data): 543 disclosed software transactions; median EV/revenue 4.5× over the period, 3.8× in 2025. Disclosed-deal sample.
9 GF Data (ACG), 2025 full-year report: 297 private-equity-backed transactions of $10–500 million total enterprise value, all sectors, average 7.2× EBITDA. US, sponsor buyers only.
10 Benchmarkit with Pavilion, 2025 B2B SaaS Performance Metrics Benchmarks (583 companies, surveyed early 2024, mostly US and under $50 million ARR). Self-selected sample.
11 Brad Feld, “The Rule of 40% For a Healthy SaaS Company” (February 2015), attributing the rule to a late-stage investor at a board meeting and intending it for companies at scale. A heuristic, not a finding.
12 Tversky & Kahneman, “Judgment under Uncertainty”, Science 185, 1974 (the arbitrary-anchor experiments); Galinsky & Mussweiler, “First Offers as Anchors”, Journal of Personality and Social Psychology 81(4), 2001: three experiments in which the party making the first offer obtained the better outcome, an advantage eliminated when the recipient focused on the opponent’s alternatives or reservation price, or on its own target. Laboratory negotiations.
13 Grant Thornton Switzerland, Quality of earnings in M&A transactions (April 2026): scope and purpose of a quality-of-earnings analysis as distinct from an audit.
14 Axial, Dead Deal Report (January 2026): 75 broken letters of intent in 2025; quality-of-earnings or EBITDA discrepancies cited in 21.3%. Small, US, self-reported.
You should know your number before any buyer knows your name.
Thirty minutes with a partner: which methods apply to your company, what the reference range looks like today, and which buyers could rationally sit above it. Under NDA, no engagement attached.
Reading this with a live transaction in front of you? Ask a partner about valuation → No engagement, and no need to name the company.