What is the SaaS magic number — and what's a good one in 2027?
Direct Answer
The SaaS magic number is a capital-efficiency ratio that asks how much new annualized recurring revenue each dollar of sales and marketing spend produces. The original Scale Venture Partners formula is current-quarter ARR minus prior-quarter ARR, multiplied by 4, divided by prior-quarter S and M spend.
In 2027, a magic number above 1.0 means lean harder on GTM, 0.75 to 1.0 is a healthy hold, and under 0.5 means the GTM motion is broken. World-class is above 1.5. The median public SaaS company sat near 0.7 entering 2027, compressed from the 2021 peak of roughly 1.1.
TL;DR
- Magic number equals net-new ARR for the quarter, annualized, divided by the PRIOR quarter S and M spend.
- The most-cited bands in 2027: under 0.5 retrench, 0.5 to 1.0 hold, above 1.0 invest, above 1.5 scale fast.
- Median public SaaS magic number in 2024 to 2026 hovered near 0.7, well below the 2021 peak of 1.1.
- CFOs in 2027 increasingly report Net Magic Number, which strips expansion ARR, because it isolates new-logo efficiency.
- The three killer mistakes: using trailing S and M, counting renewals as new ARR, and ignoring multi-quarter enterprise sales cycles.
The Formula and a Worked Example
The magic number was popularized in a 2008 Bessemer Venture Partners memo and refined by Scale Venture Partners, which is why most operators today cite the Scale version. The formula is intentionally blunt: take the change in ARR for the quarter, annualize it by multiplying by four, then divide by the sales and marketing spend from the PRIOR quarter.
The prior-quarter denominator matters because the dollars you spent in Q1 are what generated the bookings recognized in Q2. Using same-quarter S and M flatters the number during a hiring ramp and punishes it during a quarter where you front-load pipeline-generation spend.
Work it through with concrete numbers. A Series B company ends Q1 2027 with $20M ARR. They spend $3.2M on sales and marketing during Q1, which includes fully loaded AE compensation, SDR teams, marketing programs, paid media, events, RevOps tooling, and the allocated portion of customer marketing.
They close Q2 2027 at $22M ARR. The arithmetic is straightforward: net new ARR equals $22M minus $20M, which is $2M. Annualized, that is $8M.
Divide $8M by the $3.2M Q1 S and M spend and you get a magic number of 2.5. That is exceptional, and the operating implication is unambiguous: pour more dollars into the GTM motion because every dollar in is returning roughly $2.50 of annualized revenue. Companies in that range typically accelerate hiring, expand into adjacent segments, and lean into paid acquisition.
Now flip it. Same company, same $3.2M Q1 spend, but Q2 ARR lands at $20.6M instead. Net new ARR is $0.6M, annualized $2.4M, divided by $3.2M equals 0.75. That sits in the healthy-hold band. The CFO is not panicking, but the company should not be adding capacity until the next quarter confirms the trajectory.
Benchmarks by Stage
The bands move with stage. Early-stage companies should run hot because their S and M base is small and product-led pull is strong. Late-stage public companies face a math problem of scale and naturally compress toward 0.5 to 0.8 even when healthy.
The 2027 benchmarks below draw from ICONIQ Growth Operating Metrics, Bessemer State of the Cloud 2024 to 2026, and OpenView SaaS benchmark reports.
| Stage | Top Quartile | Median | Bottom Quartile |
|---|---|---|---|
| Seed (under $1M ARR) | 2.0+ | 1.2 | 0.6 |
| Series A ($1M to $10M ARR) | 1.5+ | 1.0 | 0.5 |
| Series B ($10M to $30M ARR) | 1.2+ | 0.8 | 0.4 |
| Series C ($30M to $100M ARR) | 1.0+ | 0.7 | 0.3 |
| Public SaaS ($100M+ ARR) | 0.9+ | 0.55 to 0.7 | under 0.3 |
The compression versus 2021 is real. In 2021 the median public SaaS magic number peaked near 1.1, fueled by zero-cost capital and abundant pipeline. By the 2024 to 2026 window, sales cycles stretched, win rates fell, and S and M productivity dropped roughly 30 to 40 percent across most cohorts.
The good news for 2027 operators: the bar to be top-quartile is lower than it was in the 2021 frenzy, and disciplined teams hitting 0.9 to 1.0 at scale are getting premium revenue multiples again.
The 3 Misuses That Distort the Number
The first misuse is dividing by trailing S and M instead of prior-quarter S and M. Trailing twelve-month S and M smooths out the hiring lumpiness that the metric is supposed to surface. The original Scale Venture Partners definition is explicit: prior quarter.
If you switch denominators, you are computing a different metric and your investors will catch it.
The second misuse is treating renewal ARR as new ARR. Renewals are the price of admission, not evidence of GTM efficiency. The cleanest practice is to compute Gross Magic Number using gross new ARR before churn for the broad efficiency view, then compute Net Magic Number using only net-new ARR from new logos for the truer signal on outbound and marketing-sourced motion.
CFOs in 2027 increasingly publish both, but the Net Magic Number is the one most boards now anchor on because expansion ARR has its own dedicated metric in net revenue retention, and double-counting it inside magic number obscures whether new-logo acquisition is actually working.
The third misuse is ignoring multi-quarter sales cycles. If your average enterprise deal takes 6 to 9 months to close, the S and M spend from Q1 is generating bookings that land in Q3 or Q4. A single-quarter magic number reading on an enterprise business swings violently and tells you almost nothing.
The fix is a trailing four-quarter magic number computed alongside the single-quarter view, and segmenting the calculation by motion: SMB self-serve with a one-quarter lag, mid-market with two-quarter, enterprise with three or four-quarter. Customer Magic Number, computed per logo rather than per dollar, is the third variant that helps here because it normalizes against ACV mix shifts.
Frequently Asked Questions
Net versus Gross Magic Number — which do CFOs prefer in 2027? Net Magic Number is winning. It strips expansion ARR and isolates new-logo efficiency, which is what boards want to evaluate when deciding whether to fund another sales pod. Report both, but anchor the operating conversation on Net.
Does PLG count S and M differently? Yes. Product-led companies fold product, engineering for growth surfaces, and self-serve infrastructure into a separate "investment in product-led growth" line, and the S and M denominator typically excludes those costs. Comparing a PLG magic number to a traditional outbound magic number without normalizing is misleading.
Should we report magic number monthly? No. Monthly magic numbers are noise. Quarterly is the standard cadence, and a trailing four-quarter view is the right second view. Anything tighter than that distorts more than it reveals.
Sources
- Scale Venture Partners — original Magic Number definition (2010, updated 2022)
- Bessemer Venture Partners — State of the Cloud 2024 and 2026
- ICONIQ Growth — Operating Metrics Report 2025
- Meritech Capital — Public SaaS Comp Sheet (2026 update)
- OpenView Partners — 2024 SaaS Benchmarks
- Pavilion — 2026 GTM Efficiency Study
- KeyBanc Capital Markets — 2025 SaaS Survey
- SaaStr — Magic Number Compression Analysis (Jason Lemkin, 2026)