Startup Finance
What Is a Burn Multiple, and Why Bessemer's Threshold Matters
By Byron MaloneLast updated
The formula and why it exists
In a 2020 essay, David Sacks — operator, investor, and co-founder of Craft Ventures — argued that runway was answering the wrong question. "How long can you survive?" tells you about the size of your cash reserve. It says nothing about whether you are using that cash well.
Burn multiple re-frames the question: how much fuel are you spending per mile of forward progress? A plane burning 10,000 liters to cross the Atlantic is flying efficiently. A plane burning 10,000 liters to taxi to the runway is not. The distance traveled is what changes the denominator — not the burn itself.
Formally:
Burn Multiple = Monthly Net Burn ÷ Monthly Net New ARR
where:
Monthly Net Burn = Total Cash Out − Total Cash In
Monthly Net New ARR = Current Month ARR − Previous Month ARR
Worked example: a company spends $700K in a month and collects $200K in revenue. Net burn = $500K. ARR goes from $2.4M to $2.64M — monthly net new ARR = $240K/12 = $20K on a monthly basis, or equivalently: the $240K increase in annualized contract value represents $20K of new monthly revenue. Wait — let's keep the units consistent with how the market actually uses this metric. In practice, burn multiple is usually computed on ARR-normalized figures: net burn for the month ($500K) divided by net new ARR added in that month ($240K annualized). That gives a burn multiple of 2.1×. A simpler way to state it: for every dollar of new ARR the company locked in this month, it burned $2.10.
The metric is unitless. That makes it useful for comparing companies across size brackets — a $1M ARR company and a $20M ARR company can both be at 1.5× for very different absolute burn levels.
Bessemer's four-tier framework
Bessemer Venture Partners codified Sacks's framework into a four-tier classification in their 2021 State of the Cloud report, drawing on data from their SaaS portfolio. The thresholds reflect what they observed in practice across companies that raised successfully versus those that struggled to close a next round.
| Burn Multiple | Tier | What it signals |
|---|---|---|
| < 1× | Excellent | Every $1 burned generates more than $1 of new ARR |
| 1 – 2× | Good | Capital-efficient growth; well within fundable range |
| 2 – 3× | OK | Workable, but the trend matters — is it improving or sliding? |
| > 3× | Concerning | High capital destruction relative to revenue progress; raises questions about unit economics |
Source: Bessemer Venture Partners, State of the Cloud (2021). Thresholds are portfolio-derived benchmarks, not regulatory rules. They compress in higher-cost-of-capital environments — more on this below.
What burn multiple captures that runway does not
Runway is a time budget. It tells you when the money runs out if nothing changes. Paul Graham's 2015 framing in “Default Alive or Default Dead?” sharpened this distinction: a company with 6 months of runway and fast-enough revenue growth can be default alive — meaning it reaches profitability before cash hits zero even without raising again. But "runway" alone cannot tell you that. You need to layer in the growth rate.
Burn multiple bakes capital productivity into a single number. Consider two companies:
- Company A:24 months of runway, 4× burn multiple. It has time, but it is burning $4 for every $1 of new ARR. When the cash runs low, investors will see a company that has systematically over-spent on growth. Raising again will be difficult.
- Company B:6 months of runway, 0.8× burn multiple. The clock is tight, but every dollar burned is producing more than a dollar of new ARR. The company is structurally healthier. It is likely default alive, and even if it needs to raise, the story is compelling.
Two companies at the same absolute revenue and burn can have radically different burn multiples depending on their growth velocity. That is the variable runway ignores. Burn multiple forces it into the frame.
Where burn multiple breaks down
Burn multiple is a sharp metric when applied correctly. It is a misleading one when applied to companies or situations it was not designed for. Four failure modes to know:
Pre-revenue and low-revenue companies
The denominator — net new ARR — collapses when a company has little or no ARR to grow from. Going from $0 to $5K ARR is not the same category of event as going from $2M to $2.24M ARR, but both could produce an absurd-looking burn multiple (>100×) that tells you nothing useful. The metric is designed for post-product, post-initial-traction companies. A reasonable floor is $1M ARR before burn multiple becomes a meaningful scorecard item.
Lumpy revenue — enterprise and project-based SaaS
A company that closes a $600K annual contract in January shows a burn multiple near zero for January and a very high one for the following eleven months. Monthly burn multiple can be noisy to the point of uselessness for companies where deal flow is lumpy. The fix: measure on a trailing twelve months (TTM) or quarterly basis, which smooths the distortion without losing the directional signal.
Non-SaaS businesses: hardware, marketplaces, consumer apps
Burn multiple is a SaaS-native metric. ARR is the right denominator for businesses with recurring subscription contracts. Hardware companies, consumer apps with in-app purchases, or two-sided marketplaces have fundamentally different revenue structures. Forcing their economics into an ARR-denominated metric will produce a number that sounds precise but measures the wrong thing. Use burn multiple for what it was calibrated against: recurring software revenue.
Discount-driven growth gaming
A company can improve its burn multiple by slashing prices — more customers, higher near-term ARR, lower burn multiple number. This is a real optimization risk in pre-fundraise periods. The counter is to read burn multiple alongside net revenue retention (NRR): if NRR is declining while burn multiple looks healthy, the company may be buying growth it cannot keep. Per Pavilion's Q4 2023 SaaS Pulse Survey, median NRR for SaaS companies at Series B was 107%; companies with NRR below 100% faced materially harder fundraising conditions regardless of burn multiple. Burn multiple and NRR together tell the full story that either alone omits.
The 2026 rate environment makes the thresholds tighter
Bessemer's tier thresholds were calibrated on 2021 portfolio data — a period when the Federal funds rate was near zero and venture capital was priced accordingly. Capital was cheap, runway was abundant, and investors were willing to tolerate burn multiples in the 2–3× range as the cost of growth at scale.
In 2026, with the Fed funds rate near 4%, the cost of capital has structurally reset. LP letters from major SaaS-focused funds explicitly reference burn multiple thresholds of 2× or below as table-stakes for new commitments. What Bessemer labeled "OK" in 2021 (2–3×) is now "borderline" in most term-sheet conversations. A burn multiple of 1.5× — solidly "good" in 2021 — is now approximately the threshold that separates companies that close a competitive round from those that settle for a down-round or inside-round at unfavorable terms.
The metric itself is unchanged. The bar has moved. Track your burn multiple over time, not just at a single point — a 2.8× burn multiple trending down toward 1.5× tells a better story than a 1.8× burn multiple that has been flat for six quarters.
The bottom line
Burn multiple is a direct measure of capital productivity. It does not replace runway — you still need to know how much time you have. But it answers the harder question: is the time you are spending actually buying you progress? A company can have 36 months of runway and a terrible burn multiple. It can have 8 months of runway and an excellent one. Investors who ask about burn multiple are probing the second question, not the first.
Know the metric's limits before you lean on it. It is a SaaS metric, it breaks below $1M ARR, it gets noisy with lumpy revenue, and it can be gamed with aggressive discounting. Paired with NRR, it gives you a clear picture of whether your growth is worth what you are paying for it.