Burn Multiple Denominator Trap: Why Net New ARR Beats Total Revenue Growth in VC Diligence

The Burn Multiple Denominator Trap, Explained

The burn multiple denominator trap is the diligence error of dividing net burn by the wrong measure of growth — typically total revenue growth, gross new ARR, or billings — instead of net new ARR. The inflated denominator makes the ratio look healthier than it is, masking churn, contraction, and one-time services revenue that should never have counted in the first place. In a 2026 capital environment where Series A median burn multiples have compressed to roughly 1.6x, picking the wrong denominator can be the difference between a term sheet and a pass.

David Sacks introduced the burn multiple in his 2020 Bottom Up essay to give boards a single number that captured capital efficiency: net burn divided by net new ARR. The metric is deceptively simple, and that is exactly why it gets gamed. Founders raising in 2026 — and the investors underwriting them — keep substituting easier-to-produce numerators and denominators that flatter the story. This guide walks through the trap, the canonical Sacks definition, the four most common denominator swaps, and how a clean ARR bridge resolves it.

The Sacks Definition: Net Burn Over Net New ARR

Per Sacks's original framework on his Bottom Up Substack, burn multiple is calculated over a defined period (usually a quarter, annualized) as:

Burn Multiple = Net Burn ÷ Net New ARR

Net new ARR is the change in ARR from the start of the period to the end. It is not a gross number. The bridge looks like this:

New logo ARR — recurring revenue from net-new customers signed in the period
Plus expansion ARR — upsell, cross-sell, and seat expansion on existing customers
Minus contraction ARR — downgrades, seat reductions, plan downshifts
Minus churn ARR — customers who fully terminated
Equals net new ARR

Sacks's published benchmarks define five tiers: amazing below 1.0x, great between 1.0x and 1.5x, good between 1.5x and 2.0x, suspect between 2.0x and 3.0x, and bad above 3.0x. CFO Advisors's 2026 read of the market shows top-quartile Series A SaaS now landing at 1.0x to 1.2x — a level Sacks called "amazing" four years ago is now table stakes for top-tier rounds.

Takeaway: Before you ever benchmark, lock the denominator definition. If your CFO and your lead investor are not computing net new ARR the same way, the ratio is meaningless.

Trap #1: Substituting Total Revenue Growth for Net New ARR

The most common substitution — and the one that fools the most boards — is dividing burn by total GAAP revenue growth instead of net new ARR. They look similar on a slide. They are not.

Total revenue includes professional services, one-time implementation fees, hardware pass-through, and usage spikes that will not repeat. None of that is recurring. A company billing $2M in onboarding services in Q4 looks like it grew $2M, but that $2M does not annualize. Using it in the denominator deflates the burn multiple in a way that vanishes the next quarter.

Sacks explicitly built net new ARR into the formula because ARR is forward-looking — it is the run-rate of recurring contracts at period end. GAAP revenue is backward-looking and contaminated with non-recurring streams. The Wall Street Prep teardown of the metric makes the point bluntly: burn multiple is a SaaS efficiency ratio, and if you do not have ARR you cannot compute it honestly.

Fix: Pull contract ARR from your billing system at the first and last day of the period. The delta is your denominator. Services revenue belongs in a separate line on the P&L and should never enter this calculation.

Trap #2: Using Gross New ARR Instead of Net

The second trap is more sophisticated and more dangerous: using gross new ARR — new logos plus expansion, but not subtracting churn and contraction. This is the "we sold a lot" denominator, and it is how mid-stage SaaS companies hide a leaky bucket.

The math is brutal. A company that books $10M in gross new ARR but loses $4M to churn and contraction has only $6M of net new ARR. If burn is $9M, the honest burn multiple is 1.5x (good). The gross-new version reports 0.9x (amazing). Same company, same quarter, two completely different stories.

This trap matters more than ever because expansion is now the dominant growth motion at scale. SaaS Capital's 2026 benchmarking data shows expansion ARR has risen from roughly 25% of total new ARR in 2022 to 40% in 2024, and above $100M ARR expansion accounts for about 67% of new ARR. When expansion is the engine, contraction is the symmetric risk — and only a net calculation captures both sides.

Fix: Demand a full ARR bridge in every board package: starting ARR, plus new logo, plus expansion, minus contraction, minus churn, equals ending ARR. If any of those four components is missing, the burn multiple is unverified.

Trap #3: Bookings or Billings as a Stand-In

Bookings (total contract value signed) and billings (invoices issued) get substituted into the denominator constantly, especially at early-stage companies that have not yet built ARR reporting. Both inflate the picture for different reasons.

Bookings include multi-year contract value. A $300K three-year deal books at $300K but contributes $100K to ARR. Using bookings in the denominator can understate burn multiple by 2-3x on enterprise deals.
Billings are timing-driven. A customer who prepays an annual contract in January generates $1.2M of billings in Q1 but only $300K of quarterly ARR contribution. A burn multiple computed against billings flatters Q1 and punishes Q2-Q4.

The Bessemer Venture Partners team — whose Efficiency Score is the inverse of burn multiple — is explicit that the metric should be normalized to recurring revenue. Their published guidance on the BVP framework defines efficiency as net new ARR per dollar of burn, and benchmarks the "best in class" cohort at above 1.5x (equivalent to a burn multiple below 0.67x).

Fix: If your data room shows bookings or billings as the denominator, rebuild the ratio from contract-level ARR before sharing it. Investors who notice the substitution will discount the entire data room.

Trap #4: Ignoring the Net Revenue Retention Signal

Even a correctly computed burn multiple can mislead if it is not paired with net revenue retention (NRR). NRR is the lagging indicator that confirms whether your net new ARR is durable or about to invert.

Snowflake is the cleanest public case study. According to its FY2022 10-K filed with the SEC, NRR was 178% as of January 31, 2022 — the highest figure ever reported by a public SaaS company at scale. By the end of Q3 FY2023 (October 31, 2022), per Snowflake's 8-K earnings filings, NRR had stepped down through 174%, 171%, and then 165%. Revenue growth decelerated from 84% to 67% year-over-year over the same three quarters. Snowflake's burn multiple in those periods looked fine because net new ARR was still strongly positive. But the NRR trajectory told diligence investors what the burn multiple alone could not: expansion was decelerating, and the denominator was about to compress.

This is why a16z, Sequoia, and other top-tier funds now diligence burn multiple and NRR together. A burn multiple of 1.2x with NRR of 130% is a real signal. A burn multiple of 1.2x with NRR of 95% is a denominator that is about to collapse. The SaasMag 2026 benchmarks reinforce the point — public SaaS companies above 120% NRR traded at roughly 9.3x median EV/revenue versus 3.1x for those below 100%.

Takeaway: Never present burn multiple in isolation. Pair it with NRR, gross retention, and CAC payback. A single efficiency ratio without retention context is one quarter of denominator drift away from a down round.

How to Build a Diligence-Proof ARR Bridge in Excel

The fix for all four traps is the same: a quarterly ARR bridge that ties to your billing system and your GL. Here is the minimum viable structure:

Tab 1 — Customer-level ARR snapshot: one row per customer, ARR at start of quarter, ARR at end of quarter, delta. Tag each row as new logo, expansion, contraction, churn, or no-change.
Tab 2 — ARR bridge summary: roll up the customer-level deltas into the five categories. Output the net new ARR number.
Tab 3 — Net burn calculation: operating cash outflows minus operating cash inflows for the same period. Exclude financing and one-time items.
Tab 4 — Burn multiple: net burn divided by net new ARR. Show the quarterly value and the trailing four-quarter average.
Tab 5 — Companion metrics: NRR, gross retention, CAC payback, Rule of 40. These contextualize the burn multiple.

Every cell should be auditable back to source data. Investors will tie out the ARR bridge against your Stripe, Chargebee, or Salesforce records. If the numbers do not reconcile, you fail diligence on the math before you ever get to the strategy.

Takeaway: The burn multiple is a one-cell output of a 500-cell model. Without the underlying ARR bridge, the cell is unverifiable. Build the bridge before you build the pitch.

Conclusion: The Denominator Is the Diligence

Burn multiple is now the most widely used single number in venture diligence, which is exactly why founders and investors keep abusing it. The denominator trap — substituting total revenue, gross new ARR, bookings, or billings for net new ARR — is the failure mode that turns "amazing" companies into "suspect" ones the moment a real CFO runs the bridge.

The defense is mechanical, not strategic. Build the ARR bridge from contract-level data. Net out churn and contraction. Strip services revenue. Pair the output with NRR and CAC payback. Do this every quarter, in the same template, and you remove the single largest source of diligence failure at Series A and B.

If you want to skip the spreadsheet build and start from a clean, audit-ready template, the ModelStack VC & Startup Toolkit includes a pre-built ARR bridge, burn multiple tracker, and full SaaS metrics dashboard wired to a customer-level ARR snapshot. It is the same structure top-tier funds expect to see in a data room — built in Excel, no subscription, free of the four traps above.