Burn multiple measures how much cash you burn to create each euro of new recurring revenue. David Sacks popularized it in 2020, and it has since become the efficiency number investors reach for first, because it compresses the entire company, product, sales, pricing, retention, into one ratio: what does a euro of growth cost you?
How is burn multiple calculated?
burn multiple = net burn / net new ARR
Both terms measured over the same period, usually a quarter or a year. Net burn is spending minus cash revenue. Net new ARR is the annual recurring revenue you added: new plus expansion, minus contraction and churn.
A company that burns €300,000 in a quarter while adding €200,000 of net new ARR has a burn multiple of 1.5: it pays €1.50 for each recurring euro. Lower is better, because the euro of ARR keeps arriving every year while the €1.50 is gone once.
What is a good burn multiple?
A burn multiple under 1 is exceptional, under 1.5 is great, up to 2 is good, and anything above 3 signals trouble at any stage past seed. The ranges Sacks proposed still anchor how investors read the number:
| Burn multiple | Verdict |
|---|---|
| Under 1 | Exceptional, growth funds itself |
| 1 to 1.5 | Great |
| 1.5 to 2 | Good |
| 2 to 3 | Suspect, the model needs a story |
| Over 3 | Trouble at any stage past seed |
Stage matters, and Sacks said so himself: the framework is aimed at venture-stage companies, and a very early company with a small revenue base can post a high multiple without it meaning much. What the bands assume is a company past first revenue, measured over a quarter or more.
Burn multiple benchmarks by stage
The bands do not change by stage; how strictly to read them does. This is the honest way to apply the same Sacks ranges across a company's life:
| Stage | How to read the bands |
|---|---|
| Pre-revenue | The multiple is undefined (no net new ARR to divide by). Track burn and runway instead. |
| Seed, first revenue | Read loosely. A small ARR base makes the ratio jumpy, and one good or bad quarter swings it. The signal is the trend: the multiple should improve fast as revenue compounds. |
| Series A and B | The bands apply as written: under 1 exceptional, 1 to 1.5 great, 1.5 to 2 good, 2 to 3 suspect, above 3 trouble. This is the stage the framework was built for. |
| Growth stage | Read strictly. At this scale the ratio is stable, so a multiple stuck above 3 is not noise: the machine converts cash to revenue badly, and more cash will not fix the conversion. |
Two honest caveats. Benchmarks loosen in downturns, when everyone's growth slows and the same burn buys less ARR across the whole market. And a single quarter proves little at any stage; investors read the direction across several. You can put your own quarter against the bands in the free burn multiple calculator.
Why churn is the hidden driver
Because the denominator is net new ARR, churn attacks the multiple twice: you spend to win the customer, then lose the ARR they carried. A company adding €100k of new ARR while churning €40k posts €60k net new; the same burn now produces a multiple almost twice as bad. This is why efficiency conversations so often turn out to be retention conversations, and why revenue concentration deserves watching before one big customer can move the number alone.
A worked example
A startup's quarter:
| Amount | |
|---|---|
| Total spend | €270,000 |
| Cash revenue | €150,000 |
| Net burn | €120,000 |
| New ARR | €90,000 |
| Expansion ARR | €25,000 |
| Churned + contraction ARR | -€35,000 |
| Net new ARR | €80,000 |
| Burn multiple | 1.5 |
A 1.5 lands in "great" territory. The same company with €60k of churn instead of €35k would post 2.2 on identical spending, which is the whole lesson of the metric in one row.
How to improve it
The formula offers exactly two levers, and each splits once:
Lower net burn. Cut spending that does not produce ARR (the three-pass method), or raise cash revenue from what you already sell, where pricing is usually the fastest move.
Raise net new ARR. Sell more, expand accounts, or fix the churn leak. The cheapest of the three is almost always churn.
Keep it computable
Burn multiple is a derived number: it needs your burn and your MRR movements to be current before it can be true. If those live in a spreadsheet from two months ago, your multiple is fiction. Track burn and the four MRR movements as they happen, which is what Plainhub keeps live, and the quarter-end burn multiple becomes a division you can do on a napkin: one honest number, ready before any investor asks.