What is 'burn multiple' and when should you worry about yours vs. celebrate it in 2027?

A quarter that looks like a disaster and isn't

Picture a Series-B company sitting at roughly $24M ARR. The CFO opens the quarterly finance review with a number that lands badly: the burn multiple for the quarter just closed came in at 2.29x. Last quarter it was 1.07x. The board's growth investor, who reads a dozen decks a week and has internalized the post-ZIRP grading scale, sees a company that went from "great" to "suspect" in ninety days. The temperature in the room drops.

Here is what actually happened. The company burned $3.2M in the quarter, essentially flat against the prior quarter's $3.0M. Nothing structural changed in the cost base — no hiring spree, no new office, no marketing splurge. What changed was the denominator. A large enterprise deal that had been forecast to close in the final two weeks of the quarter slipped across the boundary by eleven days because the customer's security review needed one more sign-off. That single deal was worth roughly $1.3M of net new ARR. Without it, net new ARR for the quarter landed at $1.4M instead of the ~$2.7M that would have been normal. Divide flat burn by a denominator that lost nearly half its mass and you get a multiple that nearly doubles.

The following quarter, the slipped deal closed alongside the normal pipeline. Burn: $3.1M. Net new ARR: $4.2M. Multiple: 0.74x — "amazing" by the same grading scale that had condemned the company one quarter earlier. If the board had panicked at 2.29x and forced a hiring freeze, they would have damaged a healthy engine on the basis of a calendar accident. If they had celebrated at 0.74x and greenlit an aggressive hiring plan, they would have been building on an equally false reading.

The honest number is neither. Across the full four quarters — $12.6M of total burn against $11.3M of total net new ARR — the trailing-twelve-month multiple is 1.11x. That is the verdict: a healthy, capital-efficient company whose quarterly numbers are noisy because enterprise deals close on human timelines, not fiscal ones.

This scenario is the reason the question is framed as *when to worry versus when to celebrate* rather than *what number is good*. A burn multiple is a reading on an instrument, and instruments have noise. The discipline that separates a useful metric from a destructive one is knowing which readings are signal, which are artifact, and what you should do about each. That discipline is a RevOps problem as much as a finance one, because the systems that produce the denominator — CRM close dates, ARR rollups, churn and contraction tagging, the boundary between committed and consumption revenue — all live in the RevOps stack. If those systems are sloppy, your burn multiple is a precise-looking number computed from mush.

There is a second, quieter lesson in this scenario. The CFO who walked into that room with a 2.29x and no context handed the board a problem to solve. The CFO who walks in with "the quarterly multiple is 2.29x, driven entirely by one $1.3M deal slipping eleven days, TTM is 1.11x, and here is the deal closing next week" hands the board an explanation. Same underlying business, radically different meeting. Reporting the metric alongside the mechanism is not spin — it is the difference between a number and information.

How the mechanism actually works

The formula looks trivial and the definitions are where teams quietly lose half a turn in either direction.

Net burn is the change in cash and cash equivalents over the period, adjusted to strip out financing activity. In plain terms: how much did your bank balance shrink, ignoring money you raised or debt you drew? Start a quarter at $40M, end at $36M, no raise, no debt draw — net burn is $4M. But if you raised $20M mid-quarter and still ended at $36M, your operating net burn is $24M, not $4M. The raise masked the real consumption. This is the single most common misreporting error, and it is not always deliberate; a finance team pulling the number straight from a cash-balance delta without an adjustment produces a flattering result by accident.

Net new ARR is ending ARR minus beginning ARR. It is *net* because it already absorbs churn and contraction. Start a quarter at $20M ARR, sign $3M of new and expansion business, lose $1M to churn and downgrades, and your net new ARR is $2M — not $3M. Using gross new ARR in the denominator is the second most common error, and it flatters the multiple by pretending churn does not exist.

Worked cleanly, a quarter looks like this: beginning cash $40.0M, ending cash $35.5M, no capital raised, so net burn is $4.5M. Beginning ARR $24.0M, $4.0M of new and expansion signed, $1.0M churned and contracted, ending ARR $27.0M — so net new ARR is $3.0M. The burn multiple is $4.5M divided by $3.0M, or 1.5x. That is "great" territory by current standards.

Notice how fragile that number is. Had a $5M financing tranche landed mid-quarter and gone uncounted, ending cash would read $40.5M, net burn would read negative $0.5M, and the company would report a meaningless *negative* multiple. Discipline in the numerator is everything, and the same is true of the denominator: strip out anything that is not genuinely recurring and likely to renew before you compute.

The window you choose matters as much as the definitions. A single month is pure noise and should never reach a board. A single quarter is a leading-edge trend detector, useful for spotting a turn, distorted by deal timing. The trailing twelve months is the board-grade headline — it smooths seasonality and lumpiness, at the cost of being slow to reflect a genuine recent turn. Report both: TTM as the headline, the quarterly series as a trend line beneath it. Never let one quarter, good or bad, drive a strategic decision on its own. The exception is a sustained directional move — three consecutive quarters of a rising multiple is signal, not noise.

What makes the metric worth the definitional care is its comprehensiveness. Magic Number watches only sales and marketing spend against new revenue — it is blind to an oversized R&D org, G&A bloat, office leases, and the customer-success headcount you hired to hold churn down. A company can post a beautiful Magic Number and still be a cash bonfire. CAC payback measures only the time to recoup acquisition cost and says nothing about the rest of the P&L or gross-margin drag. Rule of 40 combines growth and profit but uses a margin definition that flexes with capitalized software development, stock-based compensation treatment, and one-time classification. Raw net burn tells you the speed of the bleed but not whether the spending is productive — a company burning $10M a quarter while adding $20M of net new ARR is a rocket; a company burning $3M while adding $500K is a problem, and raw burn cannot tell them apart.

Burn multiple folds the entire cash story into the numerator and the entire durable-revenue story into the denominator. There is nowhere to hide, which is exactly why David Sacks of Craft Ventures published the concept in April 2020, in the first weeks of the COVID demand shock, when fundraising froze and every founder needed one honest answer to "is our spending actually working?" Cash is cash; the bank statement does not negotiate.

Real numbers, ranges, and stage-adjusted bars

The widely used grading scale, popularized by Craft Ventures and echoed across growth-investor diligence templates, runs: below 1.0x is amazing, 1.0x to 1.5x is great, 1.5x to 2.0x is good, 2.0x to 3.0x is suspect, and above 3.0x is a capital-efficiency emergency. Those are trailing-twelve-month figures, not quarterly ones.

A note on the arithmetic edge cases, because they cause real confusion. A negative burn multiple is undefined and meaningless, not "infinitely good." If you are cash-flow positive with no net burn, you do not have a burn multiple — you have graduated past the metric, and you should say so in words rather than reporting a number. If you are *losing* ARR while burning cash, the ratio also goes negative, and that is a five-alarm fire, not an achievement. Never let a negative number sit on a slide without a sentence explaining which of the two situations produced it.

The flat table is a starting point, not a verdict, because the bar tightens materially with stage. At seed or pre-product-market-fit, under roughly $1M ARR, the multiple is statistical garbage — a single $50K contract swings it by a full turn. At Series A, roughly $1M to $5M ARR, 2.0x to 3.0x is genuinely acceptable. At Series B, roughly $5M to $15M, expect 1.5x to 2.5x. At Series C, $15M to $40M, efficiency is now expected: 1.0x to 2.0x. At Series D and beyond, below 1.5x and trending toward 1.0x, because a path to free cash flow must be visible.

The tightening is structural, not an arbitrary investor preference. A Series-A company is *building* its go-to-market machine — hiring its first reps, testing its first paid channels, writing its first repeatable playbook. Almost none of that spending has produced revenue yet, because there is a multi-quarter lag between hiring a rep and that rep carrying full quota. The multiple reflects investment in a machine that does not yet run. By Series C, the machine should be built — the playbook repeatable, reps ramping predictably, channels with known efficiency curves — so continued high burn now reflects an inefficient *running* machine rather than an unfinished one. The same number means "under construction" at one stage and "broken" at another. A practical heuristic: each subsequent stage should lower your acceptable ceiling by roughly half a turn, and within eighteen months of a contemplated IPO the market expects a modeled path below 1.0x.

The flat table misses one more thing, and it is the most important refinement in this entire entry. Two companies can both post 2.0x and be in completely different health. Company A burns $8M to add $4M of net new ARR, and its net new ARR grew 80% year over year, up from roughly $2.2M. Company B burns $8M to add $4M — the identical multiple — but its net new ARR grew 15%, up from roughly $3.5M. The flat benchmark stamps both "suspect."

Now ask the question that matters: what happens if each is forced to cut burn 30% next year? Company A's growth is driven by a compounding engine — product pull, expansion, an acquisition channel still scaling — so a 30% cut might slow net new ARR growth from 80% to 55%, and the multiple falls toward roughly 1.2x. Company B's growth is almost entirely *bought*; a 30% cut likely reduces net new ARR close to proportionally, leaving the multiple stuck near 2.0x while growth collapses toward single digits. Same number today, opposite futures. This is why a sophisticated investor always asks for the growth rate and the growth *trajectory* in the same breath. A burn multiple without a growth rate next to it is half a sentence.

Pair the multiple with runway and you get a decision matrix rather than a grade. Below 1.5x with more than thirty months of runway, consider investing *more*. Below 1.5x with under eighteen months, cash is tight but efficiency is proven — raise calmly. Between 2.0x and 3.0x with thirty-plus months, there is time to improve without panic. The same 2.0x to 3.0x with under eighteen months is an emergency: you will be forced to raise, sell, or cut from a position of weakness into a market that prices inefficiency punitively. The rule of thumb: above 2.0x with under eighteen months of runway, the multiple stops being a metric to improve and becomes a problem to solve this quarter. Eighteen months is the practical floor because a fundraise itself consumes three to six months, and you never want to be raising with fewer than nine to twelve months of cash visible to a new investor.

Trade-offs, alternatives, and the levers that actually move it

When you decide the multiple needs to come down, the levers are finite and they sort cleanly into two families: grow the denominator or shrink the numerator.

Denominator levers are the low-risk ones. Reducing churn and contraction grows net new ARR directly — it is pure upside, because the *net* in net new ARR is where leakage shows up first. Increasing expansion and net revenue retention is similarly low-risk and carries near-zero incremental acquisition cost; you already paid to land the customer. Raising prices on new business grows ARR per deal quickly, at the medium risk of slowing new-logo velocity while the sales team recalibrates.

Numerator levers move faster and carry more risk. Improving gross margin — infrastructure costs, support load per account, delivery efficiency — shrinks burn slowly but safely. Slowing hiring pace shrinks burn fast, at the medium risk of slowing the growth you are trying to make efficient. Cutting genuinely underperforming programs is fast and low-risk if they are truly underperforming. Reducing headcount is the fastest and the most dangerous, because it damages morale and capacity in ways that show up two quarters later in the denominator.

The ordering principle is unambiguous: exhaust the low-risk denominator levers before reaching for the high-risk numerator ones. Reducing churn improves the multiple *and* the business. Cutting headcount improves the multiple while potentially destroying the engine you are trying to make efficient. A multiple improved by means that shrink the denominator faster than the numerator is not an improvement at all — it is a company getting smaller and calling it discipline.

The way to make that ordering concrete is a glide path rather than a target. A Series-B company at 2.2x should not declare "get to 1.5x next quarter" — that is unachievable and demoralizing. It should commit to a quarter-by-quarter path with a named lever and a named owner at each step: Q1 from 2.2x to 2.0x by pausing non-quota-carrying hiring (COO owns it; shrinks the numerator without touching pipeline). Q2 from 2.0x to 1.8x by launching a formal expansion motion (customer success owns it; grows the denominator at near-zero CAC). Q3 from 1.8x to 1.65x by repricing new business (sales and finance own it; tests demand elasticity). Q4 from 1.65x to 1.5x through a gross-margin program on infrastructure (engineering owns it; shrinks burn without slowing growth). The early steps use the lowest-risk levers; the riskier repricing and structural cost work come later, and can be deferred or softened if the earlier steps over-deliver. A board that receives a credible lever-by-lever glide path will tolerate a currently-high multiple far more readily than a board that hears "we know it's high, we're working on it."

The alternatives to burn multiple are worth understanding as companions rather than substitutes. Rule of 40 answers whether cash efficiency and the growth-profit balance agree; when the two disagree — a healthy 45 Rule of 40 alongside a 3.0x multiple — trust the burn multiple, because cash does not have accounting policies. The usual explanation is capitalized engineering payroll improving reported margin while not a single dollar of cash behaves differently, or an "adjusted EBITDA" that excludes stock-based comp. SBC is non-cash so it does not hit the burn numerator directly, but a company leaning on it is borrowing from the cap table to flatter a metric. A CFO who proactively says "our Rule of 40 is 45 but our burn multiple is 2.6x, the gap is capitalized engineering, here is the reconciliation" earns credibility; one who lets the board find the gap loses it.

CAC payback tells you whether the acquisition-specific story matches the whole-company story — if payback looks fine but the multiple is ugly, the burn is happening outside acquisition. Net revenue retention tells you how much of your efficiency is durable expansion versus exposed new-logo acquisition. Gross margin tells you whether delivery cost caps the achievable multiple regardless of go-to-market work. Read alone, the burn multiple is a blunt label; read alongside these, it is a diagnostic instrument. Treat it as the headline of a paragraph, not the whole paragraph.

There is also a set of companies for which the standard bar is simply the wrong lens. Deep infrastructure — databases, developer platforms, security infrastructure, data clouds — legitimately runs high multiples for years, because revenue arrives as a delayed step-change once a platform crosses an adoption threshold rather than proportionally to spend. Snowflake, MongoDB, and Datadog all ran early burn profiles that would fail a naive 1.5x test and all three became category-defining public companies. The right diagnostic there is the *trajectory* of the multiple plus platform-adoption leading indicators, not the absolute level. Hardware-attached and IoT models carry inventory and manufacturing working capital that inflates the numerator for reasons unrelated to go-to-market efficiency; compute a software-segment multiple separately, or benchmark against hardware-attached peers. Pre-PMF companies should not be steered by the metric at all — the exploration spending that kills bad hypotheses is *correct* even though it produces an ugly ratio. And a deliberate, time-boxed land grab can justify 3x for four quarters, provided three conditions hold: the window is explicitly dated, runway covers the whole window plus a buffer, and there is a pre-committed plan that brings the multiple back below 1.5x when the window closes. Absent all three, "land grab" is a story that justifies inefficiency.

Common pitfalls and how to avoid them

The first pitfall is treating a rising multiple as a fundraising problem. It is not. When your multiple climbs from 1.8x to 2.4x to 3.1x across three quarters while net new ARR stays flat or falls, you are watching a growth engine break in real time — each incremental dollar is less productive than the last. The cause is almost always one of three things, and they require different responses. Sales-capacity saturation: you hired reps faster than you generated pipeline, so new reps are starving. Test it by measuring average net new ARR per *fully ramped* rep over three quarters; if that is falling while headcount rises, pause hiring and rebuild pipeline before adding anyone. Channel exhaustion: your best acquisition channel hit diminishing returns and you are now buying expensive marginal customers. Test it by decomposing net new ARR by channel and computing cost per dollar of net new ARR per channel; if your historically best channel's cost is spiking, reallocate rather than doubling down. PMF erosion: a competitor, a pricing gap, or a feature hole is quietly raising churn and contraction, shrinking the *net* even as gross bookings hold. Test it by separating gross new ARR from churn and contraction; if gross is flat but net is shrinking, no amount of sales spend fixes a denominator draining out the bottom. Raising capital into any of these produces a larger, faster, more expensive failure, and a far worse position at the next raise because the new investors watch the multiple deteriorate on their watch.

The second pitfall is a denominator propped up by revenue that is not really recurring. A $2M three-year prepaid deal with a logo unlikely to renew, a large professional-services contract miscategorized as ARR, a usage spike from one customer that will not repeat — each makes your reported multiple better than reality. The test is mechanical: strip out anything not genuinely recurring and likely to renew, then recompute. If the multiple jumps from 1.6x to 2.4x after that scrub, the 1.6x was a story you told yourself. Usage-based businesses need this discipline most, because consumption revenue is volatile and a quarter inflated by one customer's seasonal spike produces a deceptively low multiple that reverses next quarter. Separate *committed* recurring revenue from *variable* consumption revenue and report both.

The third pitfall is the mirror image: a false celebrate. Execute a pricing or packaging change and the multiple can drop sharply in the very next quarter because a wave of existing customers signed early renewals at the new rate. That inflates net new ARR in the transition quarter and deflates it in the two that follow. A genuinely accretive reprice shows a flat-or-better transition quarter *and* holds or improves two quarters later. A pull-forward shows a sharply better transition quarter and a noticeable sag afterward. The honest test is the same one used everywhere in this entry: read the TTM multiple across the full transition, not the single quarter that contains the change. A company moving from per-seat to usage-based pricing should expect net new ARR to be *lumpier* afterward, not necessarily lower — the right way to celebrate that transition is to show TTM holding steady while the *committed* floor under the usage model keeps growing.

The fourth pitfall is a set of reporting errors that a rigorous finance team eliminates on a checklist. Financing activity left in the numerator, hiding real burn behind a fresh raise. Gross rather than net ARR in the denominator, pretending churn does not exist. One-time revenue counted as recurring. A cherry-picked window — reporting the best quarter instead of TTM. Mismatched periods, with burn measured over one window and ARR over a slightly different one. Deferred-revenue timing, where a quarter heavy with annual prepayments looks artificially cash-rich. A team that reports honestly will sometimes post an uglier number than a less rigorous peer; that honesty is itself a credibility asset with investors who have seen every flattering trick.

The deeper reason to insist on the honest number is that the burn multiple is most valuable as an internal steering instrument, and a miscalibrated instrument steers you off a cliff. If your true multiple is 2.6x and you have convinced yourself it is 1.6x by counting a raise against burn and a prepaid deal as recurring, you will make hiring plans, spend commitments, and fundraising-timing decisions appropriate to a 1.6x company. You discover the real number only when the prepaid deal does not recur, the cash runs lower than the model said, and you are raising from weakness into a market pricing your actual efficiency. The flattering number buys nothing; it delays the reckoning and worsens the position you face it from.

The fifth pitfall is presentation. Build one board slide and present it identically every quarter: the TTM headline as a single large figure, at least six quarters of quarterly multiple as a trend line, net burn and net new ARR shown *separately* so the reason for a move is visible, months of runway alongside it, and one plain-language sentence on why the number is where it is and what management is doing. Resist adding adjusted, normalized, or ex-one-time variants — the credibility of the metric comes from its starkness. And where you can, show cohort burn efficiency underneath: burn to land and grow each acquisition cohort, trended across successive cohorts. Four consecutive cohorts each more efficient than the last is the hardest celebrate signal to fake, because it survives the deal-timing noise that contaminates any blended quarter.

The last pitfall is philosophical, and it is the one worth carrying out of this page. A burn multiple is a verb, not a noun. A company at 2.2x and falling is in a fundamentally better place than a company at 1.8x and rising, even though the flat benchmark grades the 1.8x company higher. The trend, the story behind the trend, and the credibility of the plan to continue it are the real diagnostic. When you look at yours, do not ask what grade it earns. Ask which direction it is moving, whether you can explain why in one sentence, and whether the explanation would survive a skeptical question. That is when you know whether to worry or to celebrate.

Related questions

Can a burn multiple be negative?

Arithmetically yes, but it is never a grade. Cash-flow positive with growing ARR means you have graduated past the metric — say that in words. Negative because ARR is shrinking while you burn is an emergency. Never report a negative multiple as if it were a low one.

How often should I calculate it?

Monthly for internal tripwires, quarterly for board reporting, with trailing-twelve-months as the headline figure. A single month is dominated by hiring timing and payment cycles. Quarterly catches turns early; TTM is the version a serious investor anchors on during diligence.

Does burn multiple replace Rule of 40?

No — they cross-check each other. Rule of 40 reads an adjusted income statement; burn multiple reads the bank account. When they disagree materially, the gap is usually capitalized engineering cost or excluded stock-based compensation. Present both on the same page and explain any divergence proactively.

What if my ARR is under $1M?

Then the metric is statistical noise and should not steer you. One $50K contract signing or churning swings the ratio by a full turn. Pre-product-market-fit, use qualitative signals instead: early-cohort retention, organic pull, and demonstrated willingness to pay.

Which lever improves it fastest without damage?

Churn and contraction reduction. It grows the net in net new ARR, requires no new acquisition spend, and improves the underlying business rather than just the ratio. Expansion revenue is the close second, since you already paid to land those customers.

FAQ

What counts as a good burn multiple right now?

Under 1.0x is excellent, 1.0x to 1.5x is great, 1.5x to 2.0x is acceptable, 2.0x to 3.0x is suspect, and above 3.0x signals a capital-efficiency problem. These are trailing-twelve-month figures and reflect post-ZIRP investor expectations, which are materially tighter than what the market tolerated before 2022. Stage-adjust them: a Series A at 2.5x is normal, a Series D at 2.5x is not.

How do I calculate net burn correctly?

Take beginning cash minus ending cash for the period, then strip out every financing activity — equity raises, debt draws, venture-debt tranches. If you raised mid-period and forget to adjust, the raise masks your real consumption and can even produce a nonsensical negative burn figure. The number you want is operating cash consumption, not the change in the bank balance.

Why use net new ARR instead of gross new bookings?

Because gross bookings pretend churn does not exist. Net new ARR is ending ARR minus beginning ARR, so it already absorbs churn and contraction. A company signing $3M of new business while losing $1M to downgrades added $2M of durable revenue, not $3M — and using the larger number flatters the multiple while hiding a retention problem.

Should a rising multiple trigger an immediate cost cut?

Not immediately. One rising quarter is a prompt to investigate, not a verdict. Three consecutive rising quarters with flat or falling growth is signal. Before cutting, diagnose whether the cause is sales-capacity saturation, channel exhaustion, or retention erosion — each has a different fix, and cutting headcount when the real problem is churn makes the denominator worse.

Does the metric work for usage-based pricing models?

It works, but it needs more care. Consumption revenue is volatile, so a quarter inflated by one customer's spike produces a low multiple that reverses. Separate committed recurring revenue from variable consumption, report the multiple against the committed floor as well as the blended figure, and always read it on a trailing-twelve-month basis.

How does this connect to RevOps day to day?

Every input lives in RevOps systems: CRM close dates that determine which quarter a deal lands in, ARR rollup logic, churn and contraction tagging, and the boundary between committed and consumption revenue. If those definitions drift or are inconsistently applied, the multiple is a precise-looking number computed from unreliable inputs. Locking the definitions is a RevOps deliverable, not a finance afterthought.

Sources

• https://sacks.substack.com/p/the-burn-multiple
• https://www.craftventures.com/
• https://www.bvp.com/atlas
• https://www.saastr.com/
• https://openviewpartners.com/blog/
• https://www.mckinsey.com/industries/technology-media-and-telecommunications/our-insights
• https://a16z.com/16-startup-metrics/
• https://hbr.org/topic/subject/finance-and-investing
• https://www.investopedia.com/terms/a/arr.asp
• https://www.sec.gov/edgar/searchedgar/companysearch

Related on PULSE

• What does the Rule of 40 actually measure, and how do you explain a miss?
• What's the 'Magic Number' in SaaS and why does it matter more than CAC?
• How do you calculate true CAC payback period with multi-quarter sales cycles?
• How do you model CAC for usage-based pricing with no upfront contract value?
• What metrics belong on a board-ready unit economics dashboard, and in what order?
• How do I track burn multiple alongside other efficiency metrics?