The SaaS magic number tells you how much new recurring revenue each unit of sales and marketing spend bought you. It is the quickest check a founder can run before deciding whether to put more money into growth. To calculate it, take the change in quarterly revenue, multiply it by four, and divide by the previous quarter’s sales and marketing spend.
This guide covers where the metric came from, the magic number formula with a worked example, what the 2026 benchmark data says a good result looks like, and the four ways the number misleads. It is written for SaaS founders who know their ARR and their marketing spend but not yet whether that spend is working.
The short version
- What it measures: new annualized revenue bought per unit of last quarter’s sales and marketing spend.
- The baseline: Scale Venture Partners, who coined the metric, call 0.7x a fairly healthy baseline.
- The 2026 data: the median across 132 companies in Benchmarkit’s 2026 report was 1.37; the bottom quartile was 0.68.
- The catch: a one-quarter number on a long sales cycle can mislead. Read it beside CAC payback period, never alone.
Table of Contents
What the SaaS magic number measures, and where it came from
The metric is a ratio. On top sits the new recurring revenue you added in a quarter, annualized. Underneath sits what you spent on sales and marketing to win it. The answer is a plain number: 1.0 means each dollar, or shilling, of sales and marketing spend bought one dollar of new annual revenue: a direct read on go-to-market efficiency.
It came out of venture capital. According to Scale Venture Partners’ own history of the metric (published 11 September 2020, checked 20 October 2026), partner Rory O’Driscoll coined it while evaluating Omniture as a possible investment around 2005. Omniture was producing more than $2 of first-year revenue for every $1 it put into sales and marketing, and the name stuck.
Investors use it because it answers one question quickly: if this company spends more on growth, does the revenue follow? Founders should use it for the same reason, before they hire the next salesperson or raise the ad budget.
It is one of several sales efficiency metrics, and it works best alongside them. It says nothing about how long a customer takes to pay back what they cost to win, or how long they stay. Those are separate questions with separate numbers.
The magic number formula, and how to calculate it
Scale’s own description, in its follow-up piece on the metric (15 September 2021, checked 20 October 2026), is annualized new GAAP revenue as a percentage of last quarter’s sales and marketing expense. It assumes that the spend in one quarter shows up as growth in the next.
Written out, the revenue version of the magic number formula is:
Magic number = (this quarter’s revenue − last quarter’s revenue) × 4 ÷ last quarter’s sales and marketing spend
Many companies now run it on ARR instead. Aleph’s summary of the 2026 benchmark report (checked 20 October 2026) gives that version as net new ARR added in the period divided by sales and marketing expense in the prior period. Both are valid. What matters is picking one basis, revenue or ARR, and never mixing the two across quarters.
A worked magic number calculation
The figures below are illustrative, for a hypothetical Nairobi-based B2B SaaS company billing in shillings. They are not from a real company.
- Last quarter’s recurring revenue: KES 12,000,000 (about US$92,300)
- This quarter’s recurring revenue: KES 13,500,000 (about US$103,800)
- Quarter-on-quarter change: KES 1,500,000, which annualizes to KES 6,000,000 (about US$46,200)
- Last quarter’s sales and marketing spend: KES 7,500,000 (about US$57,700)
- Magic number: 6,000,000 ÷ 7,500,000 = 0.80
Count everything that went into winning customers in the spend figure: salaries and commission for sales and marketing staff, ad spend, agency and freelance fees, events and tools. Leaving salaries out is the most common way a team flatters its own number.
If you want the cost side worked out first, the customer acquisition cost calculator totals the same spend lines and divides by customers won.
SaaS magic number benchmark: what good looks like in 2026
Two credible sources give two quite different reference points, and the gap between them is the useful part.
Scale’s long-run view. In its 2020 history of the metric, Scale calls 0.7x a fairly healthy baseline for most SaaS businesses, and says the long-term median for private SaaS companies had hovered around that level before trending down.
The 2026 survey data. The Benchmarkit 2026 SaaS and AI Metrics Benchmarks, published 1 June 2026 with Aleph as presenting sponsor, surveyed 342 B2B SaaS and AI-native companies. 132 of them reported a magic number for calendar year 2025. The figures below come from the full report, checked 20 October 2026.
| Measure (Benchmarkit 2026, N=132) | Magic number |
|---|---|
| Median, CY-2025 | 1.37 |
| Median, CY-2024 | 0.94 |
| 25th percentile, CY-2025 | 0.68 |
| 75th percentile, CY-2025 | 2.14 (1.27 in CY-2022) |
| Companies growing above 50%: median | 2.40 |
| Companies growing 31–50%: 25th percentile | 0.80 |
| Companies growing 11–30% | Below 0.75 |
The report treats 0.75 as the minimum for sustainable sales and marketing investment, and 0.75 to 1.0 as the floor that justifies spending more.
So why is the 2026 median nearly double Scale’s baseline? The sources do not settle it, so treat what follows as reasoning rather than finding:
- Definitions differ. Scale describes GAAP revenue, while the ARR version in Aleph’s summary of the 2026 data counts net new ARR, and ARR can move faster than recognized revenue.
- Samples differ. A voluntary benchmark survey is not the whole private market, and Benchmarkit’s own report warns, in its section on sales and marketing spend, that sample composition is material when reading total-population figures.
- Timing differs. The 2026 report shows the median rising from 0.94 to 1.37 in a single year, which is a recent shift, not a long-run norm.
The practical reading of a SaaS magic number benchmark: compare yourself with companies at your growth rate and on your definition. On the hypothetical KES example above, 0.80 clears the 0.75 floor but sits below the 2026 median and well below the 50%-growth cohort.
When the SaaS magic number says spend more, and when it lies
Read the number in bands, not as a pass mark:
- Below about 0.75: fix efficiency before adding spend. More money into a channel mix that returns under 75 cents a year per dollar usually buys more of the same.
- Between 0.75 and 1.0: the floor the 2026 report says justifies more investment. Spend more carefully, on the channels doing the work.
- Above 1.0, held for several quarters: a case for spending more, as long as retention and payback agree.
That last condition matters, because the SaaS magic number has four blind spots.
It ignores churn unless you net it out
If the top line is gross new revenue, a quarter where you won ten customers and lost eight looks excellent. Use net revenue change, or net new ARR after churn and contraction, so lost customers count against the spend that was meant to replace them.
One quarter’s lag can be too short
The formula assumes last quarter’s spend produced this quarter’s growth. On a sales cycle of several months, it may not have. A team selling to banks or large enterprises can show a poor number in a quarter of heavy investment and a flattering one the quarter after. Smooth it with a trailing average of two to four quarters.
A single quarter is noisy
One large deal, an annual renewal season or a price change can swing a quarterly figure sharply. Judge the trend across several quarters, not one reading.
It says nothing about payback or profit
A magic number above 1.0 can sit beside a long CAC payback period if gross margin is thin. The same Aleph and Benchmarkit report puts median CAC payback at 16 months, with the top quarter at six months or fewer; our guide to CAC payback period benchmarks covers it in full. For why ad-platform return figures make this worse, see ROAS vs CAC. Investors also pair it with the burn multiple, which compares cash burned with new ARR added.
This is why how we work reports CAC payback, marketing-sourced pipeline and SQL conversion by source over a window that matches the sales cycle, and refuses blended ROAS and any 30-day performance window. The magic number is a good first question. It is a poor only question.
FAQ: the SaaS magic number
What is a good SaaS magic number?
Scale Venture Partners describe 0.7x as a fairly healthy baseline. The Benchmarkit 2026 report puts the CY-2025 median at 1.37 and treats 0.75 as the minimum for sustainable investment. Compare yourself with companies at your growth rate and on the same revenue definition.
How do you calculate the magic number?
Take the change in quarterly revenue, multiply it by four, and divide by the previous quarter’s sales and marketing spend. If you run it on ARR, divide net new ARR by the prior period’s sales and marketing spend instead. Use one basis consistently.
Is the magic number the same as CAC payback?
No. The magic number measures how much new annual revenue spend bought. CAC payback measures how many months of gross profit it takes to recover the cost of winning a customer. A company can look strong on one and weak on the other, which is why both belong in the same review.
How does the magic number relate to the Rule of 40?
The Rule of 40 checks whether growth rate plus profit margin reaches 40%, a test of the whole business. The magic number looks only at sales and marketing efficiency. A strong magic number makes the Rule of 40 easier to meet, because growth is being bought cheaply.
The SaaS magic number: what to do next
Run the SaaS magic number for your last four quarters on one definition, with salaries included in the spend. If it sits below 0.75, find the channel dragging it down before you add budget. If it holds above 1.0 and payback agrees, you have a case for spending more.
If you would rather have the channel-by-channel work done for you, how we price explains how an engagement is scoped and quoted after a short call.
