What Is a Good Burn Multiple? Benchmarks by Stage, and Where the Bands Come From
By Vishesh Khurana — Oct 14, 2026
Burn multiple is the number investors ask for when they want to know how efficiently you are buying growth. It is one division, it is easy to compute from numbers you already have, and it is now standard in seed and Series A diligence alongside the Rule of 40 and net revenue retention.
The short answer to what counts as good: under 1.5x is good, under 1x is excellent, and anything above 3x needs an explanation. But the honest version of the answer is longer, because the bands everyone quotes were published as a rule of thumb rather than measured from data, and because the band that counts as good moves with your stage. Both of those facts change how you should read your own number.
The formula
Burn multiple = net burn / net new ARR. David Sacks published it in a 2020 essay on Craft Ventures' site, framed as "how much is the startup burning in order to generate each incremental dollar of ARR?" A burn multiple of 1.5x means you spent $1.50 to add $1.00 of new recurring revenue. Higher is worse.
Worked example: you start a quarter with $1.2M in the bank and end it with $900,000, having raised nothing in between. Net burn is $300,000. ARR went from $400,000 to $600,000, so net new ARR is $200,000. Burn multiple is 300,000 / 200,000 = 1.5x.
The two inputs, defined carefully
Net burn is cash leaving the business minus cash coming in, over a period. The cash-based definition — opening bank balance minus closing bank balance, adjusted for anything you raised or repaid — is the one to use, because it cannot be flattered by revenue recognition timing. Accounting net loss is not net burn.
Net new ARR is ARR at the end of the period minus ARR at the start, which means it is already net of churn and downgrades. This is where most people get the metric wrong: they divide by gross new ARR, which ignores the customers who left, and produce a number that looks roughly twice as good as reality. If a customer churned in the period, they belong in the denominator.
The two inputs must cover the same period. Dividing an annualised burn by a quarterly ARR movement overstates efficiency by roughly 4x. Pick a quarter, compute both sides of the fraction over that quarter, and say which quarter it is.
The bands everyone quotes
The scale below is the one repeated across every venture glossary and finance blog. Sacks published it as a rules-of-thumb chart alongside the essay rather than as a table of measured results, and the essay names no survey, no sample and no data source behind the numbers.
Under 1x — amazing. You are adding ARR faster than you are burning cash; growth is close to self-funding.
1x to 1.5x — great. Efficient, investable growth. This is where a well-run company with real revenue should sit.
1.5x to 2x — good. Acceptable, but the number should be improving rather than drifting.
2x to 3x — suspect. Expect hard questions, and expect to need a specific plan for how it comes down.
Above 3x — bad. You are spending too much for the growth you are getting. Sacks' own illustration of a terrible burn multiple is $5M burned in a quarter to add $1M of net new ARR, which is 5x.
That the chart was never backed by a published dataset is the most useful thing to know about these bands. It means they are a convention rather than a measurement — a shared expectation that investors and founders have converged on, which is real and worth meeting, but not evidence about your business. Nobody has published a large, sourced burn-multiple benchmark table; the numbers circulate because everyone repeats them. Read the bands as the language your investor already speaks, not as a law of nature.
The bands move with your stage
Sacks' own text is explicit that stage changes the read, and this is the part founders skip. He notes that a seed-stage company might run a burn multiple of 3x, because it has only just started selling, and that the expected number improves to around 2x after a Series A. He calls a 2x burn multiple "reasonable for an early-stage startup" and tolerates weak multiples mainly "in the earliest years, when sales are incipient and the product is still being built".
So the same 2.5x is unremarkable at seed and a problem at Series A. The mechanical reason is arithmetic: net new ARR is a difference between two small numbers early on, and the sales and marketing spend that produces it lands before the revenue does. A company at $200,000 of ARR adding $50,000 in a quarter is doing something very different from a company at $4M adding $1M, even if both spend the same to get there.
Our own seed-stage SaaS benchmarks use a looser band for exactly this reason: below 2x is strong at seed, 2x to 4x is acceptable, and above 4x is a warning. That is not a contradiction of the Sacks scale, it is the same scale read at a stage where the denominator is noisier. If your board deck cites one of the two, say which and why.
The direction matters more than the level
A 2.4x trending down to 1.6x over three quarters is a better story than a 1.4x that has crept up to 2.1x. Investors read the slope because it tells them whether the efficiency is structural or whether you simply have not spent the money yet. Compute the number for the last four quarters and put them next to each other before you look at any band.
There is one exception, and it is a real one: a company deliberately investing ahead of a market it can see clearly, with the cash to fund it, can justify a bad-looking quarter. That argument works exactly as long as the cash lasts and no longer.
How to compute yours in five minutes
Pull three numbers from your bank rather than your P&L: the balance on the first day of the quarter, the balance on the last day, and any financing that moved in or out of the account during it. Net burn is opening minus closing, plus anything you raised, minus anything you repaid.
Pull two numbers from your billing system: ARR at the start of the quarter and ARR at the end. The difference is net new ARR, and the gap between that and your new sales for the quarter tells you how much churn you are absorbing.
Divide, then label the quarter. Write "Q3 2026: 1.8x" rather than an unattributed number, because the metric is meaningless without the period attached, and because next quarter you will want the series.
If your net new ARR is negative — you shrank — the metric is undefined rather than infinite, and that is a signal in itself: no amount of capital efficiency fixes a shrinking revenue base, and the conversation should be about the churn, not the burn.
Where the metric stops being useful
Burn multiple assumes growing recurring revenue. A pre-revenue company has no denominator and should not use it. A business with heavy one-off or services revenue has ARR that understates what it earns, so the metric will look worse than the business is. A marketplace recognised gross versus net has two different numbers depending on which way you count. And a company that has stopped growing in order to preserve cash will show a superb burn multiple while it dies, which is why this number is never the only one on the page.
Where this leaves you
A burn multiple is a diagnostic, not a target. If yours sits above the band for your stage, the useful question is which line of spending is not converting into ARR yet, and whether waiting one more quarter for it is affordable — and that last part is a runway question, not an efficiency one.
We built FounderConsole to answer it in one place. Connect Stripe, QuickBooks or your bank and you get net burn and net new ARR from your own data, a burn multiple for the period, and the runway distribution that tells you how long you can afford to keep the current spend. Where the data cannot honestly support a metric we mark it as an estimate or show N/A rather than printing a plausible-looking number.
To see the runway half without an account, the free runway calculator needs no signup, and the default-alive check answers the related question of whether you reach profitability on the cash you already have. If either one tells you what you needed to know, that is a fine outcome.