Burn multiple is net burn divided by net new ARR. See the formula, Sacks' stage rules of thumb, and how to calculate it from your books before a raise.
Burn multiple equals net burn divided by net new ARR. David Sacks defined it on April 23, 2020 as the inverse of Bessemer's efficiency score: how many dollars of cash you spend to add one dollar of annual recurring revenue. A 2x multiple means $2 of net burn for each $1 of ARR added.
A board asks for this number in a raise. The formula is short. The inputs are where the books usually fail.
Net burn divided by net new ARR for the same period. David Sacks introduced the ratio on April 23, 2020. A higher multiple means more cash spent for each dollar of new ARR. A lower multiple means more efficient growth.
Sacks presents burn multiple as the inverse of Bessemer's efficiency score. HiBob's explainer restates the same relationship and credits the formula to him.
| Metric | Formula |
|---|---|
| Burn multiple | Net burn ÷ net new ARR |
| Bessemer efficiency score | Net new ARR ÷ net burn |
| Hype ratio | Capital raised (or burned) ÷ ARR |
The hype ratio compares capital raised or burned with the ARR balance. Burn multiple compares one period's net burn with that period's net new ARR, so a later quarter can improve after earlier spend was wasted. Sacks' point is that the multiple does not lock in sunk costs.
Run it on a month, a quarter, or a year. Both inputs have to cover that same window. ARR stays an annual run-rate. The change in the ARR balance is already annualized. Do not multiply that change by 12 or by 4 a second time.
Sacks sets the examples at a board meeting after Q1. Both use his formula on one quarter.
He also marks the gap between those two results. If the growth required extraordinary investment, which he defines as 3x burn or more, he treats that as a sign that product-market fit is weaker than the growth rate suggests, or that some other problem sits in the business.
On his reading, the company that adds $1 million of ARR by burning $2 million looks stronger than the company that adds the same ARR by burning $5 million. In the first case the market is pulling the product. In the second the company is pushing it.
The stage figures below are one investor's rule of thumb, taken from the text of Sacks' essay. They are not a survey of closed rounds. The essay also contains a benchmark chart published as an image. This page does not transcribe that chart.
Sacks says the multiple should improve as the company matures:
If the multiple worsens as the company matures, he treats that as a problem even while headline growth is still rising in dollars.
HiBob describes the same direction. A higher multiple is common at seed, while the company is still building the product and the first customers. From Series B onward, HiBob says investors expect the multiple to fall as revenue growth outpaces net burn. HiBob points to an a16z chart for numeric bands by ARR. That chart is an image, and this page does not transcribe it.
Seed around 3, about 2 after Series A, tighter after Series B, and toward 0 at profitability. That is Sacks' text from April 23, 2020. Another investor can ask for a tighter number in the same round. Put the formula and the source on the slide next to the result.
Sacks calls the ratio a catch-all. A serious problem shows up by raising burn, cutting net new ARR, or moving both at rates that do not match. These are his drivers, paraphrased from the essay:
Cutting cost improves the next period's multiple, because the ratio uses that period's burn. Dropping unprofitable revenue can do the same when the cash you stop spending is larger than the ARR you give up.
Inference and model API bills move with usage. That is Sacks' gross-margin case inside a usage-priced product. ARR can rise while the cash cost of serving that usage rises with it. Track the cost and the margin is visible. Leave it in a general software account and the multiple gets worse with no line in the deck that explains it.
The formula uses net burn. Gross burn is a different total.
| What it counts | In the burn multiple | |
|---|---|---|
| Gross burn | Cash operating spend in the period, including model and API bills | No |
| Net burn | That cash spend minus cash collected from customers | Yes. This is the numerator |
| Net new ARR | Change in annual recurring revenue over the same period | Yes. This is the denominator |
Gross burn and net burn are cash measures. A model invoice increases gross burn when you pay it. It increases net burn unless customer collections rise by the same cash in that period. ARR can move when a contract is signed, before that cash arrives, so a usage spike can worsen the multiple in the quarter the API cost hits.
Classify inference, hosting, and the support that delivers the product as cost of goods sold. Sacks' point is mechanical. If the cost of delivery scales with revenue and the business has no operating leverage, a higher ARR balance does not improve the burn multiple on its own.
Run four checks before the number goes in a board deck.
Read net burn on the statement of cash flows. Operating cash flow is the cash the period's operations produced or consumed. Accrual net loss on the profit and loss statement includes non-cash charges and misses timing. A prepaid annual contract, an unpaid vendor bill, and a capitalized cost move profit and cash on different dates.
Use the operating section. An equity wire increases the bank balance and sits in financing cash flow. It does not reduce burn. The cash flow statement separates operating, investing, and financing for this reason. If a one-time equipment purchase is in investing cash flow, decide whether that cash belongs in the burn you are defending, and say which choice you made.
Net new ARR = new ARR + expansion ARR − churned ARR − contraction ARR.
That rollforward is the one used for annual recurring revenue. ARR and recognized revenue are different totals. ARR is the annualized recurring run-rate. Recognized revenue is what the accounting period earned. An annual contract can add its full annual value to ARR in the month it starts and add only one month of that value to recognized revenue. Using recognized revenue in the denominator understates net new ARR and inflates the multiple.
Sacks' 2x example pairs one quarter of net burn with the ARR added in that quarter. Monthly burn over a quarterly ARR bridge, or yearly burn over a one-month ARR change, is a different ratio. Label the window on the slide: month, quarter, or year.
ARR is already an annual figure. In Sacks' quarter, "added $1 million of ARR" means the ARR balance rose by $1 million during those three months. Divide that quarter's net burn by that change.
Annualize burn only when the denominator is the ARR change across that same longer window. Annualizing the burn and leaving the ARR change on the shorter window multiplies the result by the number of periods you stretched. An ARR change that is already a change in annual run-rate does not get multiplied again.
A startup financial model should compute the multiple from the cash schedule and the ARR bridge on the same columns. If the hiring plan and the ARR bridge use different months, the multiple on the output tab is not the multiple in the books.
These ratios sit next to burn multiple in a board pack. Each one asks a different question. The table states the definition. The worked versions stay on the linked pages.
| Metric | Question it asks | Where to read the calculation |
|---|---|---|
| Burn multiple | How much net burn did this period's net new ARR require? | This page |
| Bessemer efficiency score | How much net new ARR did each dollar of net burn produce? This is the inverse. | Sacks, April 23, 2020 |
| Rule of 40 | Do the revenue growth rate and a profit margin, added together, clear 40? | SaaS financial metrics that matter |
| CAC payback | How many months of gross profit does it take to recover the cost of winning a customer? | SaaS financial metrics |
The Rule of 40 adds a growth rate to a margin. A revenue multiple is the price paid per dollar of revenue. The two names measure different things.
CAC payback can explain a weak burn multiple. Sacks' sales-efficiency case is a high acquisition cost or falling sales productivity. Payback is the narrower metric for that motion. Burn multiple still includes delivery cost, churn, and the rest of operating spend.
The 5-day Finance Ops Diagnostic checks the inputs. It is a $3,500 fixed fee, delivered in 5 business days. The review maps the current stack and handoffs, the spreadsheet dependencies, the missing owners, and a cleanup path, so net burn and the ARR bridge can be tied to the same close.
The diagnostic does not include tax preparation or a fundraise narrative. It checks whether the two numbers in the formula are the numbers in the books.
This is general information from published definitions. It is not a valuation, a financing commitment, or advice for one company's round. Investors in that round can set a tighter bar than the rules of thumb above.