Burn multiple: formula, calculator and 2026 benchmarks
What is the burn multiple?
The burn multiple is how much cash a startup burns to add each dollar of new annual recurring revenue: burn multiple = net burn ÷ net new ARR over the same period. Burning $2M to add $1M of ARR is a 2x multiple. Lower is better; under 1x is excellent.
Key facts
- David Sacks of Craft Ventures introduced it in April 2020: "Burn Multiple = Net Burn / Net New ARR". (Sacks, 2020)
- Sacks's bands for venture-stage startups run from under 1x (amazing) through 1–1.5x (great) and 1.5–2x (good) to 2–3x (suspect) and over 3x (bad). (Sacks, 2020)
- The median burn multiple was 1.0 in Benchmarkit's 2025 data, across 62 companies. (Benchmarkit, 2026)
- Bessemer's "ideal growth profile" pairs 100% revenue growth with a 1.2x burn multiple. (Bessemer, 2023)
- ICONIQ found top-quartile companies keep burn multiples under 1.0x after scaling past $10M of ARR. (ICONIQ, 2025)
Burn multiple calculator
Enter the cash burned over a period (cash out minus cash in, excluding new funding) and the net new ARR added over the same period, after churn and downgrades. Use a quarter or a year; a single month is noisy.
Want your burn multiple from real numbers? CX Cash is being built to track burn, runway and net new ARR together, from your bank and billing data.
Get started free (opens in a new tab) See burn multiple and efficiency metrics in CX Cash
How to calculate the burn multiple
Burn multiple = Net burn ÷ Net new ARR
In words: divide the cash the company burned over a period by the recurring revenue it added, net of churn, over the same period.
- Net burn: Cash out minus cash in from operations over the period, leaving out new equity or debt. ICONIQ uses free cash flow instead.
- Net new ARR: New ARR plus expansion, minus contraction and churn, over the same period.
Worked example
An illustrative startup that burned $3M over the last 12 months and added $2M of net new ARR after churn. The results below are computed by the calculator's tested code.
| Input | Value |
|---|---|
| Net burn over 12 months | $3,000,000 |
| Net new ARR over the same 12 months | $2,000,000 |
Burn multiple: 1.50x Net new ARR per $1 burned: 0.67x
Burn multiple = 3,000,000 ÷ 2,000,000 = 1.50x
A burn multiple of 1.50x: $1.50 of cash for each $1 of new ARR, at the line between Sacks's "great" and "good". Had churn been $500,000 lower, net new ARR would be $2.5M and the multiple 1.20x, Bessemer's ideal for a fast grower (our arithmetic).
In a spreadsheet
- Burn multiple, with net burn in B2 and net new ARR in C2: =B2/C2
- Net burn from cash, with cash at the start in B2, at the end in C2 and funding raised in D2: =B2-C2+D2
- Net new ARR, with new, expansion, contraction and churned ARR in B2:E2: =B2+C2-D2-E2
- Bessemer's efficiency score, the inverse: =C2/B2
How founders and investors read the burn multiple
If you're the founder
- It's the question investors ask in a downturn: how much cash does each dollar of growth cost? Sacks's example: burning $2M in a quarter to add $1M of ARR is "reasonable for an early-stage startup"; burning $5M for the same growth means it "should probably cut costs immediately".
- It catches problems anywhere in the business. Sacks: "Any serious problem will eventually impact the Burn Multiple by either increasing burn, decreasing net new ARR, or (most tricky) increasing both but at disproportionate rates."
- Churn hurts it twice: lost ARR lowers the denominator while the cost of replacing it raises the numerator.
Questions you'll be asked
- Is that net burn or free cash flow, and over what period?
- How much churn and contraction sits inside net new ARR?
- How has the multiple moved quarter by quarter?
- How much runway does today's burn leave?
If you're the investor or LP
- Read it with growth and stage. Bessemer pairs 100% growth with a 1.2x multiple; a 2x multiple at that growth is a different company from 2x at 20% growth.
- Check what's in net burn. Large annual prepayments lower burn for a quarter without changing efficiency; one-off costs raise it.
- It stops working once a company makes money. For profitable companies the Rule of 40 is the better lens.
Questions to ask the company
- How is net burn defined, and does it include financing or one-offs?
- What were gross new ARR and churn in the period?
- How does the multiple compare across the last four quarters?
- What would cutting the least efficient spend do to growth?
What is a good burn multiple?
Under 1x is excellent and under 2x is acceptable for a fast-growing startup; above 3x is a warning. Benchmarks improve with scale, because existing revenue covers more of the fixed costs.
Show the numbers
| ARR band | Median | 25th percentile (best quarter) |
|---|---|---|
| Under $5M | 1.00x | 0.80x |
| $5M–20M | 1.30x | 0.60x |
| $20M–50M | 1.00x | 0.30x |
| $50M–100M | 0.50x | 0.20x |
Benchmarks
| Segment | Measure | Typical | Top quartile | As of | Sample | Source |
|---|---|---|---|---|---|---|
| Measured burn multiples · For this metric lower is better, so the "top quartile" column shows the 25th percentile. | ||||||
| B2B SaaS and AI-native | Burn multiple | 1.0 (75th percentile 2.0) | 0.4 | CY 2025 | 62 companies (Benchmarkit) | Benchmarkit, 2026 |
| Growth-stage software, past $10M ARR | Burn multiple of the top quartile | Under 1.0x | — | 2025 | ICONIQ portfolio and public companies | ICONIQ, 2025 |
| Rules of thumb | ||||||
| Venture-stage startups (Sacks) | Bands | <1x amazing; 1–1.5x great; 1.5–2x good; 2–3x suspect; >3x bad | — | Apr 2020 | VC rule of thumb | Sacks, 2020 |
| Fast-growing SaaS (Craft) | Acceptable range | Under 1 amazing; under 2 quite good | — | Oct 2021 | VC rule of thumb | Craft Ventures, 2021 |
| Early-stage companies (Bessemer) | Attractive range | About 1x to 1.5x | — | Jan 2024 | VC rule of thumb | Bessemer, 2024 |
| 100% growth (Bessemer) | Ideal pairing | 1.2x | — | Apr 2023 | VC rule of thumb | Bessemer, 2023 |
Scale lowers the multiple. Benchmarkit's median falls from 1.3 at $5M–20M of ARR to 0.5 at $50M–100M; companies under $5M "show the highest Burn Multiple".
ICONIQ expects further improvement: "We believe burn multiples are gradually improving across segments as a result of AI automation and labor" arbitrage.
Compiled Oct 2, 2026. Next review: June 2027, after Benchmarkit's 2027 report. Left out: ICONIQ 2026 State of Scaling (The report prohibits reproduction of its figures); Public companies (None discloses a burn multiple, and most large SaaS companies are cash-generative). Download these benchmarks as CSV.
Burn multiple vs efficiency score, magic number, Rule of 40 and runway
| Metric | Answers | Ignores | Use it when | It misleads when |
|---|---|---|---|---|
| Burn multiple | Cash burned per $1 of net new ARR | Whether the company is near profitability | Judging a cash-burning company's growth efficiency | The company is profitable or barely has ARR |
| Efficiency score | Net new ARR per $1 burned, the inverse | The same as the burn multiple | Bessemer's and Insight's reporting | Mixed up with the burn multiple's direction |
| SaaS magic number | New annual revenue per $1 of sales and marketing | Spend outside sales and marketing | Deciding on sales and marketing budgets | One quarter is unusual |
| Rule of 40 | Growth rate plus profit margin | How much the growth cost in cash | Scaled or profitable companies | Applied to early-stage startups |
| Runway | Months until the cash runs out at today's burn | Whether the burn buys growth | Planning the next raise | Burn is about to change sharply |
Burn multiple = 1 ÷ efficiency score. Over a year, net burn equals the burn multiple times net new ARR, so a company adding $5M of ARR at 1.5x burns $7.5M (our arithmetic).
Common mistakes: how the burn multiple gets flattered
- Counting new funding against burn. Net burn is operating cash out minus cash in, before any equity or debt raised. Including a raise makes the multiple meaningless.
- Using gross new ARR. Sacks's denominator is net of churn. Leaving churn out flatters the multiple exactly when churn is the problem.
- Reading a single month. Lumpy deals and annual payments swing monthly figures. Use a quarter or a trailing year.
- Letting prepayments hide burn. A big annual prepayment cuts burn for the quarter it lands, without any change in efficiency. Look at the trailing year.
- Comparing net burn with free cash flow. ICONIQ divides free cash flow by net new ARR. Capital expenditure and other items make the two numerators differ.
- Applying it before revenue. Sacks: "if the startup is pre-revenue, net new ARR will be zero so the ratio will not even compute."
Where the burn multiple came from and how it's regarded today
| When | What happened | Source |
|---|---|---|
| Sep 2025 | ICONIQ benchmarks the burn multiple by ARR band and finds top-quartile companies under 1.0x past $10M of ARR. | ICONIQ, 2025 |
| Jan 2024 | Bessemer says early-stage companies should aim for a burn multiple of about 1x to 1.5x. | Bessemer, 2024 |
| Jan 2024 | Dave Kellogg predicts "strong attention to cash burn efficiency (burn multiple)". | Kellogg, 2024 |
| Jun 2023 | Kellogg notes that Insight's "efficiency rule" is "what most people call the burn multiple". | Kellogg, 2023 |
| Apr 2023 | Bessemer's ideal growth profile pairs 100% growth with a 1.2x burn multiple. | Bessemer, 2023 |
| Oct 2021 | Sacks and Ethan Ruby make it part of Craft's Series A metric set. | Craft Ventures, 2021 |
| Apr 2020 | David Sacks publishes "The Burn Multiple" as the COVID downturn hits. | Sacks, 2020 |
| Feb 2019 | Bessemer introduces its efficiency score, net new ARR over net burn, for companies under $30M of ARR. | Bessemer, 2019 |
| Mar 2016 | Kellogg proposes the "Hype Factor", capital raised divided by ARR, the cumulative forerunner. | Kellogg, 2016 |
How it's regarded today
- A standard downturn metric. Bessemer, Craft, ICONIQ and Insight all track it, under one name or its inverse, alongside growth.
- Improving with scale and AI. ICONIQ: "burn multiples are gradually improving across segments as a result of AI automation and labor" arbitrage (ICONIQ, 2025).
- Judged with growth. Benchmarkit's own guidance: a higher burn multiple "at fast growth and with significant runway may be a deliberate and rational investment strategy".
- The Burn Multiple (Sacks, 2020). The original definition, the bands, and why it doubles as a product-market-fit signal.
- 2026 SaaS and AI Metrics Benchmarks (Benchmarkit, 2026). Burn multiple quartiles for 2025 and medians by ARR band.
- State of Software 2025: Rethinking the Playbook (ICONIQ, 2025). How the burn multiple falls as companies scale, and the top-quartile line past $10M ARR.
What experts say about the burn multiple
“we still think in terms of the business achieving an attractive burn multiple (ideally ~1x-1.5x)”
“The beauty of the Burn Multiple is that it’s a catch-all metric.”
“For fast-growing SaaS companies, a Burn Multiple of less than one is amazing, but anything less than two is still quite good.”
Burn multiple FAQ
What is the burn multiple?
A capital-efficiency ratio: net cash burned divided by net new ARR over the same period. It answers Sacks's question, "how much is the startup burning in order to generate each incremental dollar of ARR?"
How do you calculate the burn multiple?
Take net burn over a quarter or a year (cash out minus cash in, excluding funding) and divide by net new ARR over the same period. $3M burned for $2M of net new ARR is a 1.5x burn multiple.
What is a good burn multiple?
Under 1x is excellent and 1x to 1.5x is great by Sacks's bands; Bessemer considers about 1x to 1.5x attractive early on. The median in Benchmarkit's 2025 data was 1.0. Over 3x is a warning.
What is the difference between burn rate and burn multiple?
Burn rate is how much cash you lose per month. The burn multiple divides that loss by the ARR it bought, so it measures efficiency, not just speed of spending.
Why use net new ARR rather than new ARR?
Because churn has to be replaced. Net new ARR subtracts churn and downgrades, so a leaky business shows a higher, more honest multiple.
What if the company is profitable?
Then net burn is negative and the multiple stops meaning much. Use the Rule of 40 or free cash flow margin instead.
What is the difference between the burn multiple and Bessemer's efficiency score?
They're inverses. The efficiency score is net new ARR divided by net burn, so higher is better; the burn multiple flips it, so lower is better. A 1.5x burn multiple is a 0.67x efficiency score.
Related metrics and guides
Sources
Every link was opened and checked. Archived copies guard against links that move or disappear.
Primary data
- 2026 SaaS and AI Metrics Benchmarks. Benchmarkit, 1 Jun 2026. Sponsored report; band figures read from its chart; small sample. Archived copy
- State of Software 2025: Rethinking the Playbook. ICONIQ Analytics, Sep 2025. Investor's portfolio and public-company data; quoted, not reproduced. Archived copy
Practitioner
- The Burn Multiple. David Sacks (Craft Ventures; reposted on Bottom Up), 23 Apr 2020. Reposted from Craft Ventures' Medium publication; bands are in an image. Archived copy
- The SaaS Metrics That Matter. David Sacks and Ethan Ruby (Craft Ventures), 18 Oct 2021. Archived copy
- State of the Cloud 2023. Bessemer Venture Partners, 11 Apr 2023. Archived copy
- The Rule of X. Byron Deeter and Sam Bondy (Bessemer Venture Partners), 2 Jan 2024. Archived copy
- Bessemer Venture Partners' 2019 State of the Cloud. Bessemer Venture Partners (SlideShare), 6 Feb 2019. Slide deck; read from the archived copy. Archived copy
- Introducing a New SaaS Metric: The Hype Factor. Dave Kellogg (Kellblog), 22 Mar 2016. Archived copy
- Interpreting the Insight 2023 Sales KPI Report. Dave Kellogg (Kellblog), 18 Jun 2023. Archived copy
- Kellblog Predictions for 2024. Dave Kellogg (Kellblog), 2 Jan 2024. Archived copy
Changes to this page
- · Major · Page published. Benchmarks compiled 2 October 2026.
Cite this page
Dominique Bouillet, "Burn multiple: formula, calculator and 2026 benchmarks", CX Cash, updated Oct 22, 2026, https://cxcash.com/metrics/burn-multiple
CX Cash builds software for founders and investors. This page is education, not investment advice. Third-party figures link to their source, and our own arithmetic and illustrative examples are labelled as such.