Revenue multiples explained: how private startups are really valued
Almost every private valuation comes down to one line of arithmetic — revenue times a number. That number carries most of the risk in a round, and it is more art than science. Here is how to read it, what pushes it up and down, and how to tell a justified multiple from hype.
A revenue multiple values a company as its revenue times a number. A business with £10m of annual recurring revenue, valued at a 20x multiple, is worth £200m on that basis. The whole game is deciding what that number should be — and that decision, not the revenue, is where almost all the judgement and risk sit.
Startups are usually valued on revenue rather than profit because most are not yet profitable. The multiple is meant to reflect how good that revenue is — how fast it is growing, how much sticks, and how defensible it is. Read the fundamentals behind the number, not just the number.
You will hear it stated as a fact. "They raised at 30x ARR." "The comparable companies trade at 8x revenue." "It is expensive at 40x." The multiple gets quoted like a temperature reading — precise, objective, settled. It is none of those things. It is a compressed opinion about the future, dressed up as a single number.
This piece unpacks that number properly. What a multiple is, why private startups are valued on revenue instead of profit, the difference between the versions you will meet, and — the part that matters — how to work out whether a given multiple is earned or simply hoped for. We will use round, clearly hypothetical figures throughout. None of them are real market data.
- A multiple is revenue times a number. The revenue is usually the easy part; the number is the argument.
- Startups are valued on revenue because most have no profit. Profit-based multiples break when earnings are negative.
- ARR is stricter than revenue. It counts only recurring, predictable income — which is why it earns a richer multiple.
- Growth, margin and retention drive the number. The Rule of 40 is a quick sanity check that blends growth and profitability.
- A high multiple is a claim, not a fact. Test it against the fundamentals before you accept it.
What a multiple actually is
A multiple is simply the ratio between what a company is worth and some measure of what it produces. Value on top, a business metric on the bottom. The best-known one in public markets is the price-to-earnings ratio: share price divided by earnings per share. A stock at 20x earnings is priced at twenty times its annual profit.
Turn it around and the multiple becomes a valuation tool. If you know that companies like this one tend to trade at 20x earnings, and you know this company earns £5m, you can estimate its value at £100m. That is the whole mechanism — find the right multiple, apply it to the right number, read off a valuation. The difficulty is never the multiplication. It is choosing the multiple.
Why startups are valued on revenue, not profit
Here is the problem with earnings multiples for young companies: most startups do not have earnings. They deliberately run at a loss, spending on engineering, sales and marketing to grow as fast as they can. A company might be adding customers quickly while burning cash every month.
You cannot build a sensible price-to-earnings ratio on a negative number. Divide a valuation by a loss and you get nonsense. So the market moves down the income statement to the largest number it can actually trust — revenue — and values the company against that instead. Revenue is real, it is growing, and it is a reasonable proxy for the size and momentum of the business.
This is a temporary arrangement. As a company matures and its profit becomes real and stable, valuation gradually shifts back towards profit and cash flow. Revenue multiples are the language of the growth phase, when the profit that justifies today's price still lives in the future.
The versions you will meet: EV/Revenue and the ARR multiple
Two forms come up constantly, and they are close cousins.
EV/Revenue divides enterprise value — the value of the whole business, including debt and net of cash — by revenue. It is the cleaner academic measure because enterprise value strips out how the company happens to be financed. You will see it in analyst notes and comparable-company tables.
The ARR multiple is the one you will hear most in private venture rounds. It divides the company's valuation by its annual recurring revenue. When someone says a company "raised at 30x ARR," they mean its valuation was thirty times its recurring revenue at the time of the round. For a subscription business, ARR is usually the number everyone anchors to.
ARR versus revenue: why the distinction matters
Not all revenue is equal, and the multiple should reflect that. Annual recurring revenue is the subscription income a company can reasonably expect to keep — the contracted, repeating money that renews month after month. Plain revenue includes everything: recurring subscriptions, yes, but also one-off setup fees, consulting, hardware, professional services and other income that may not come again next year.
Recurring revenue is worth more per pound because it is predictable. A company that knows it will collect the same money next year is far more valuable than one that has to win all of its sales again from scratch. So two companies can report identical total revenue and deserve very different multiples: the one that is mostly recurring earns the richer number.
The multiple is not paid on the revenue. It is paid on how likely that revenue is to still be there next year — and the year after.
This is also where valuations get quietly inflated. A company may present a headline "ARR" figure that includes revenue which is not truly recurring, or annualise a single strong month into a full-year number. When you see an ARR multiple, it is always worth asking what is actually inside the ARR.
Forward versus trailing multiples
The last trap in the definition is which revenue the multiple sits on. A trailing multiple uses revenue the company has already earned — the last twelve months. A forward multiple uses revenue it is expected to earn — often the next twelve months.
The distinction is not cosmetic. Consider a hypothetical company with £10m of trailing ARR, growing fast, that expects £20m next year. A £200m valuation is 20x trailing but only 10x forward. The same price looks dramatically cheaper on the forward number. Founders and their advisers, unsurprisingly, tend to quote whichever version flatters the round.
Neither is a lie, but you must know which you are being shown. A forward multiple bakes in growth that has not happened yet. If the growth disappoints, a "cheap" forward multiple turns out to have been an expensive trailing one all along.
How to read "valued at 30x ARR"
Put the pieces together and the phrase decodes cleanly. "Valued at 30x ARR" means the company's valuation equals thirty times its annual recurring revenue. If its ARR is £10m, the implied valuation is £300m. If its ARR is £50m, it is £1.5bn.
The instinctive question is "is 30x a lot?" It is the wrong first question. The right one is: what has to be true for 30x to make sense? A multiple that high is a bet on years of rapid, durable, high-margin growth. If the company is growing quickly, keeping almost all of its customers, and expanding what those customers spend, 30x can be defensible. If it is growing slowly or leaking customers, 30x is a story, not a valuation.
What drives a multiple up or down
A handful of fundamentals do most of the work. The same £10m of ARR can reasonably support a single-digit multiple or a very high one depending on these.
| Factor | Pushes the multiple UP | Pushes the multiple DOWN |
|---|---|---|
| Growth rate | Fast, sustained growth (doubling or more) | Slow or decelerating growth |
| Gross margin | High margins (software-like, 70%+) | Thin margins (services or hardware-heavy) |
| Net revenue retention | Above 100% — existing customers spend more each year | Below 100% — the base is leaking |
| Market size (TAM) | Large, expanding addressable market | Small or saturated market |
| Moat | Defensible — switching costs, network effects, data | Easily copied, commoditised |
| Revenue quality | Mostly recurring, contracted, diversified | One-off, concentrated in a few customers |
| Rule of 40 | Growth plus margin comfortably above 40% | Well below 40% |
Net revenue retention deserves a note because it is the quiet hero of the list. It measures what happens to the money from a cohort of existing customers over a year — after some churn, but including upgrades and expansion. Above 100% means the company would grow even if it never signed a single new customer, because its current base spends more each year. That property is rare, and it justifies a lot of multiple.
The Rule of 40 ties two of these together into one quick test: a healthy software company's growth rate plus its profit margin should reach at least 40%. Fast growth can excuse losses; strong margins can excuse slower growth; but a company that fails on both is hard to justify at a rich multiple. It is a guide, not a law, and it flatters some business models more than others — but it is a fast gut-check.
How AI has distorted multiples
Since the current wave of artificial-intelligence investment, revenue multiples in and around AI have detached from their historical ranges. Rounds have reportedly been done at multiples that would have looked absurd a few years ago, on companies with short histories and revenue that is growing at unprecedented speed but is also young and, in some cases, of uncertain durability.
Two forces are at work. First, some AI companies genuinely are growing faster than any prior software cohort, and a forward multiple on triple-digit growth mathematically looks less insane than the trailing number suggests. Second — and this is the part the pitch decks leave out — scarcity and fear of missing out inflate prices well beyond what the fundamentals alone would support. When capital chases a small number of names, multiples expand regardless of whether the retention and margins have been proven.
The discipline here is the same as always, only harder to hold: separate the companies whose fundamentals justify the number from those whose scarcity justifies the number. The first can survive a correction. The second is priced for perfection. We go into this in more depth in our guide to investing in AI startups, linked below.
The other methods, briefly
Multiples are not the only way to value a company, but for private startups the alternatives are weaker.
- Comparable companies. Look at what similar businesses are valued at and apply their multiple. This is really just a structured way of choosing a multiple — useful, but only as good as the comparables, which are often not that comparable.
- Discounted cash flow (DCF). Project the company's future cash flows and discount them to today. Rigorous in theory, but nearly useless for early-stage startups: it demands confident long-range forecasts for a business that may not have a stable product yet. Small changes in assumptions swing the answer wildly. For young companies, a DCF is often false precision.
- The last-round mark. The simplest of all — the company is "worth" whatever it last raised at. This is convenient and widely quoted, but it can be stale the moment the market moves, and it reflects one negotiation, not an ongoing market price.
A worked example (hypothetical)
Take a purely illustrative company — call it Northwind. It has £10m of ARR and is growing at 80% a year. Its gross margin is 78%, its net revenue retention is 118%, and its revenue is almost entirely recurring. On a 20x ARR multiple, its implied valuation is £200m.
Is 20x justified? On these hypothetical fundamentals, it is at least arguable. Growth of 80% plus a positive contribution comfortably clears the Rule of 40. Retention above 100% means the base compounds on its own. High margins mean the revenue converts efficiently to eventual profit. The multiple is high, but the quality behind it is high too.
Now change one thing. Suppose the same £10m ARR is growing at 15%, retention is 92% and half the revenue is one-off consulting. The identical £200m valuation — still 20x — is now indefensible. Same revenue, same multiple, completely different reality. That is the entire lesson: the multiple only means something in the context of what sits beneath it. These are invented figures used to illustrate the mechanics, not a real company.
Before we take any deal to members, we take the round's multiple apart. We ask what is really inside the ARR, whether the multiple quoted is forward or trailing, and what growth and retention would have to hold for years to make the number reasonable. We compare it against genuinely similar companies, not flattering ones, and we write down the case for why the price could be wrong. If a multiple only works on the founder's most optimistic forecast, we say so plainly in the memo. The number has to survive scrutiny before members ever see it.
How to sanity-check a multiple
You do not need a valuation model to pressure-test a number. You need a few blunt questions.
- Forward or trailing? Always establish which revenue the multiple sits on. A cheap-looking forward multiple may hide an expensive trailing one.
- What is inside the ARR? Is it genuinely recurring, or padded with one-off income and annualised good months?
- Does growth support it? A high multiple demands high, durable growth. If growth is slowing, the multiple should be too.
- Does the base leak? Net revenue retention below 100% undercuts almost any premium multiple.
- Would it survive a re-rating? If comparable multiples across the market fell by half, would this company still be a reasonable investment? If only the multiple holds it up, that is fragile.
A justified multiple has a story you can check. A hype multiple has a story you are asked to believe. The difference is whether the fundamentals underneath it can bear the weight — and that is knowable, if you are willing to look.
Private valuations are far less solid than they appear. A company's "value" is often just the price of its last funding round, which can be months or years old and reflects one negotiation rather than an open market. Marks can be stale, optimistic, or shaped by terms — liquidation preferences and other protections — that flatter the headline number while disadvantaging ordinary shareholders. A multiple derived from a stale or engineered mark inherits all of those flaws. Treat any single private valuation as an estimate with a wide error bar, not a fact.
Frequently asked questions
What is a revenue multiple?
A revenue multiple values a company as a multiple of its revenue — you take the company's revenue and multiply it by a number to arrive at a valuation. The most common forms are EV/Revenue, which divides enterprise value by revenue, and the ARR multiple, which divides valuation by annual recurring revenue. It is a shorthand for how much the market will pay today for each pound of a company's sales.
What is a good revenue multiple for a startup?
There is no single good number, because the multiple should reflect growth, margins and retention. Slower-growing software companies have historically traded in low single-digit to high single-digit revenue multiples, while fast-growing companies can command far higher. What matters is not the headline number but whether the growth, gross margin and net revenue retention behind it justify it. A high multiple on weak fundamentals is a warning, not a badge.
Why are startups valued on revenue instead of profit?
Most startups are not yet profitable, because they spend heavily on growth. A profit-based multiple like a price-to-earnings ratio needs positive earnings to work, and dividing by a negative or near-zero number produces a meaningless figure. Revenue is the largest reliable number on the income statement, so it becomes the anchor. As a company matures and profit becomes stable, valuation usually shifts towards profit and cash-flow measures.
What is the difference between an ARR multiple and a revenue multiple?
A revenue multiple can use any revenue figure, including one-off and project revenue. An ARR multiple uses annual recurring revenue only — the predictable, subscription-based revenue a company can reasonably expect to keep. Because recurring revenue is more durable, ARR multiples are the standard for subscription and software businesses. Two companies with the same total revenue can deserve very different multiples if one is mostly recurring and the other is mostly one-off.
What is the Rule of 40?
The Rule of 40 is a rough test for software companies: revenue growth rate plus profit margin should add up to at least 40 percent. A company growing 50 percent a year while burning 10 percent passes; one growing 20 percent with a 10 percent margin does not. It is a shorthand for whether a business is balancing growth and profitability well, and companies that clear it have tended to earn higher multiples. It is a guide, not a law.
This article is educational and general in nature. It is not financial advice, a recommendation, or an offer or solicitation to invest. All figures and worked examples are hypothetical and illustrative, used only to explain the mechanics of valuation, and are not real market data. Private company valuations are uncertain, often based on stale or optimistic marks, and may not reflect what a stake could actually be sold for. Private investments are illiquid and high-risk, and you may lose all of your capital. Capital is at risk.