SaaS Valuation Multiples in the UK: EBITDA and ARR Benchmarks

The short answer

For profitable private SaaS and software businesses in the UK and EU — the £1m–£30m enterprise-value band where most SME deals actually happen — the benchmark TrueValue’s valuation engine applies is roughly 8x to 14x EBITDA, with 14x the current default for a pure SaaS model and 10x–12x for software businesses whose revenue is less predominantly recurring. Those figures are benchmarks, not prices: they describe where comparable transactions have clustered, and where any individual business lands depends on its growth, churn, customer concentration and the quality of its reported earnings.

Where EBITDA does not exist — an early or pre-profit SaaS business, or one reinvesting heavily into growth — the engine falls back to a revenue multiple, and deliberately a conservative one: 1.5x revenue is the current default for SaaS, 1.2x for fintech and 1.0x for technology businesses more broadly. The distinction between the two measurement bases is the single most useful thing to understand about SaaS valuation, because applying the wrong one moves the implied price by a wide margin before any negotiation begins.

EBITDA multiples by software sector

TrueValue’s valuation engine carries a benchmark table of mid-market EBITDA multiples for UK and EU SMEs, built around an enterprise-value band of £1m–£30m. The software and technology rows are the ones that matter here, and the spread between them is not noise — it reflects how much of each business’s revenue is recurring, how sticky the product is, and how capital-intensive delivery is.

Two things to notice. First, the premium sits with genuinely recurring revenue: the gap between SaaS at the top of the table and software development further down is a statement about how much of the revenue comes back without being re-sold. Second, these are base multiples for a typical business in each row — the factors that move a multiple within its range apply here exactly as they do in any other sector.

ARR multiples versus EBITDA multiples

The two multiples answer different questions about different businesses. An EBITDA multiple prices the profit a business already makes: apply 14x to £500k of EBITDA and you have a £7m enterprise value grounded in demonstrated earnings. An ARR (or revenue) multiple prices a business whose profit does not yet support that arithmetic — typically because it is loss-making, near break-even, or reinvesting everything into growth. It is a bet that the revenue will eventually convert into profit at software margins, discounted for the fact that this has not happened yet.

This is why TrueValue’s engine treats the revenue multiple as a fallback rather than a feature: it only reaches for a revenue multiple when usable EBITDA does not exist, and the defaults are conservative — 1.5x revenue for SaaS against 14x EBITDA. The conservatism is the point. A 14x EBITDA multiple on a profitable business and a 14x revenue multiple on an unprofitable one describe entirely different risk positions, and the widely quoted ARR multiples from US venture-backed deal coverage belong to that second, far riskier population: companies growing at venture scale, sold to buyers who can fund years of losses. Those multiples do not transfer to a profitable UK SME with £400k of EBITDA and single-digit growth, and quoting one as a benchmark for the other is how sellers arrive at expectations no buyer can fund.

The practical rule: if the business has stable, credible EBITDA, value it on EBITDA and use the revenue figure as a cross-check. If it does not — if EBITDA is negative, negligible, or inflated by owner adjustments — value it on revenue, state clearly that you are doing so and why, and expect the multiple to be a fraction of the headline ARR numbers in the press. How to calculate EBITDA covers the adjustments that decide whether an EBITDA figure is credible enough to multiply in the first place, and EBITDA vs SDE covers the owner-managed case where the two diverge.

What moves a SaaS multiple within the range

A sector multiple prices a typical business; your business is not typical. The factors that decide where it sits inside — or outside — the 8x–14x band are the same ones the engine reads from the accounts and the deal record:

• Revenue durability. Net revenue retention, contract length and churn decide how much of this year’s revenue returns next year. A business where 90%+ of revenue recurs on annual contracts supports the top of the range; one where logo churn eats a fifth of the base each year does not.
• Customer concentration. A SaaS business whose largest customer is 40% of revenue is priced as a much riskier proposition than the same revenue spread across hundreds of accounts — see customer concentration risk for how the discount is reasoned about.
• Growth and its quality. Growth funded by broad adoption across the customer base supports a higher multiple than growth bought with heavy discounting or one large win that does not repeat.
• Earnings quality. Adjusted EBITDA that depends on aggressive add-backs — owner remuneration, one-off costs, personal expenses — is worth less than the same headline figure reported cleanly. Buyers price the risk that the adjustments do not survive due diligence.
• Owner dependency. A founder who is the product, the sales team and the support desk is a transition risk that buyers discount explicitly, particularly below the size band where professional management is assumed.

There is also a market-context point worth stating plainly: the transaction evidence behind the 8x–14x band comes from UK and EU private SaaS and software trade sales in the £1m–£30m enterprise-value range. Below that band — micro-SaaS, lifestyle software businesses, sub-£100k EBITDA — the comparable-transaction evidence thins out sharply, and what businesses are actually advertised for is a materially different question from what buyers eventually pay.

How to benchmark a specific business

A benchmark range becomes a valuation argument through four steps, and the order matters:

1. Establish which earnings figure the multiple applies to. Cleaned, normalised EBITDA for a profitable business; ARR or revenue for one that is not. Mixing the two — an EBITDA-sourced multiple applied to revenue, or an ARR multiple applied to EBITDA — produces an implied price with no evidence behind it.
2. Start from the sector benchmark, then justify the movement. Take the 8x–14x band and place the business inside it explicitly: which factors push it up, which pull it down, and why. A valuation that says “14x because SaaS” without the placement argument is a rule of thumb wearing a suit.
3. Test the price the multiple implies against the buyer’s own funding. A multiple the market paid is not a price your structure can carry — senior debt capacity, equity commitment and covenant headroom set a ceiling that sits below the comparable range more often than sellers expect.
4. Sanity-check against the other methods. A multiple-based value that sits far from what a discounted cash flow of the same projections supports is a signal that either the multiple or the projections carry an assumption nobody has said out loud — see the five valuation methods for how the methods are meant to corroborate each other.

How TrueValue applies these benchmarks

The EBITDA-multiple method in TrueValue’s valuation engine reads the sector benchmark table described above, normalises the sector name the user typed onto the right row, and adjusts for the range the comparable evidence supports. The comparable-transactions method draws on recorded transaction profiles for UK and EU private SaaS and software trade sales in the £1m–£30m enterprise-value band, using the median of a genuine comp set rather than the mean, and never substituting a valuation built from fewer than three relevant comparables. Where EBITDA is unavailable the revenue multiple enters as the conservative fallback it is. Because the engines are deterministic, the same inputs produce the same valuation every time, and the assumptions — which table row, which multiple, which earnings figure — are visible in the output rather than buried in a spreadsheet.