# SaaS Magic Number

> The magic number formula, the 0.75 and 1.0 thresholds, why the lagged version is the only honest one, and what a sub-0.5 result tells you to change.

Source: https://saas-marketing.net/guides/saas-magic-number/
Topic: SaaS Metrics and Analytics
Type: guide
Published: 2026-09-11
Last updated: 2026-09-11
Publisher: SaaS Marketing (saas-marketing.net)
License: CC BY 4.0. Quote or republish with attribution and a link to https://saas-marketing.net/guides/saas-magic-number/

## Short answer

The SaaS magic number is annualised net new revenue divided by the previous quarter's sales and marketing spend. Under 0.5, stop adding spend and fix conversion first. Between 0.5 and 0.75 the model works but pays back slowly. Above 0.75, fund growth. Above 1.0 usually means you are underinvesting. Always put the prior quarter in the denominator, because this quarter's spend has not produced revenue yet.

## Key takeaways

- Use the prior quarter's sales and marketing spend in the denominator. This quarter's spend has not had time to close anything.
- A gross-margin-adjusted magic number of 0.75 is arithmetically the same statement as a 16 month CAC payback period.
- Expansion ARR sits in the numerator and did not come from sales and marketing, so run a new logo magic number alongside the blended one.
- Below roughly $5M ARR a single quarter is noise. One enterprise deal can swing the ratio by 0.2 with nothing else changing.
- A magic number above 1.0 is an instruction to spend more, not a trophy. Boards that treat it as a ceiling starve growth.
- The metric is blind to sales cycle length, so content-led and enterprise motions look worse than they are on a one-quarter lag.

---

Ask for the magic number in a board pack and you usually get one decimal with no denominator attached. Nine times out of ten it was calculated against the same quarter's sales and marketing spend, which credits this quarter's dollars with revenue they had no time to produce.

That matters more than it sounds, because this is the only widely used efficiency metric that judges sales and marketing as one system. Not cost per lead. Not quota attainment. Both budgets together, measured against net new recurring revenue, answering a single question: is the next dollar worth committing, and how fast does it come back?

Two things go wrong in practice. The lag on the denominator, and what is sitting in the numerator. Both are fixable in an afternoon, and both move the answer far enough to change a budget decision. Put your own figures into the [SaaS magic number calculator](/calculators/magic-number/) while you read this, because the arithmetic is more persuasive than the argument.

## The formula, written the way it should be calculated

The magic number is the annualised change in quarterly revenue divided by the previous quarter's sales and marketing spend. If you already report ARR, the ARR form is simpler and gives the same answer.

```
Magic number = (ARR at end of Q3 minus ARR at end of Q2) / S&M spend in Q2
```

```
Gross-margin-adjusted = ((ARR Q3 minus ARR Q2) x gross margin) / S&M spend in Q2
```

Use the second one. The unadjusted version pretends a dollar of revenue is a dollar of contribution, which it is not at 78 percent gross margin and definitely is not at 62 percent for a product with heavy infrastructure or human services attached. Public SaaS gross margins cluster between 70 and 82 percent, so the adjustment typically cuts your headline number by a fifth to a quarter. That is not a rounding difference.

The spend line needs a written definition, kept in the same document as the metric. Sales and marketing means loaded salaries, commissions and accelerators, ad spend, events, agency and contractor fees, content production, and the sales and marketing share of tooling. Attio, Clay, Apollo, Outreach and 6sense all belong in here. Sales engineering is a judgement call, and the only wrong answer is changing your mind halfway through the year.

A company moved its two sales engineers out of S&M and into cost of revenue in Q3, and its magic number rose from 0.64 to 0.71 with no operational change whatsoever. Nobody noticed for two quarters. Write the definition down, version it, and footnote any change in the quarter it happens.

## Why the one-quarter lag changes the answer

The lag exists because revenue lands months after the spend that caused it. In a B2B motion with a 74 day median sales cycle, the pipeline that closed in Q3 was created in Q2 by spend that was committed in Q1. Matching Q3 growth to Q3 spend is a timing error dressed as a measurement.

The direction of the error depends entirely on your spend trajectory, which is why the unlagged version is worse than useless. It is wrong in a way that tracks the thing you are trying to judge.

| Scenario | Q3 net new ARR | Q2 S&M | Q3 S&M | Lagged | Unlagged |
|---|---|---|---|---|---|
| Spend ramping 40 percent | $1.20M | $1.60M | $2.24M | 0.75 | 0.54 |
| Spend flat | $1.20M | $1.60M | $1.60M | 0.75 | 0.75 |
| Spend cut 35 percent | $1.20M | $1.60M | $1.04M | 0.75 | 1.15 |

Look at the bottom row. A company that just cut its budget by a third shows an unlagged magic number of 1.15 while nothing about its efficiency has changed. This is precisely how a team convinces itself that a reduction in force improved sales efficiency, six months before the pipeline gap arrives. Every efficiency-era board deck from 2023 and 2024 contains a version of this mistake.

The ramp row is the mirror image and it is just as damaging. A company investing ahead of growth reads 0.54, panics, and cuts the hiring plan that was about to produce the revenue.

## What is actually sitting in your numerator

Net new ARR is not the same as acquired ARR. It contains expansion from the installed base, minus contraction, minus churn, and only sales and marketing produced the first slice. At a company running 115 percent [net revenue retention](/guides/net-revenue-retention/), expansion can supply 35 to 40 percent of net new ARR while the customer success team, not the demand gen team, did the work.

That contamination pushes the blended magic number up and hides a weak acquisition engine behind strong retention. It also works in reverse: a company with 96 percent NRR is subtracting churn from its own sales efficiency score every quarter, which makes its acquisition look broken when the actual problem sits in the product.

Run two numbers. Blended magic number on total net new ARR, and a new logo magic number on new logo ARR only, against the same denominator. The gap between them is the honest picture of who is producing growth.

**0.31** Typical gap between the blended and new-logo magic number at a company running 115 percent NRR, with expansion supplying roughly 38 percent of net new ARR

## The four bands, and what each one tells you to do next

These thresholds are gross-margin-adjusted. Applying them to an unadjusted number will tell you to spend in a quarter you should be fixing conversion.

The band that gets misread most often is the first one. A magic number of 0.38 does not automatically mean marketing is failing. Check the three things that sit upstream of it before you touch a budget: win rate, average selling price, and contraction in the base. A company whose ACV fell from $31,000 to $24,000 on an unchanged pipeline will produce a collapsing magic number with a perfectly healthy demand engine.

## A worked four quarters at $12M ARR

Here is a mid-market B2B SaaS business, $28,000 average contract value, 80 percent gross margin, sales-led with a 74 day median cycle. Spend is ramping through the year.

| Quarter | Ending ARR | Net new ARR | Prior quarter S&M | Magic number | GM-adjusted |
|---|---|---|---|---|---|
| Q1 2025 | $12.0M | | $1.50M | | |
| Q2 2025 | $13.1M | $1.10M | $1.60M | 0.69 | 0.55 |
| Q3 2025 | $14.4M | $1.30M | $1.85M | 0.70 | 0.56 |
| Q4 2025 | $15.9M | $1.50M | $2.20M | 0.68 | 0.55 |
| Q1 2026 | $17.0M | $1.10M | $2.60M | 0.42 | 0.34 |

Three quarters of remarkable stability. The ratio holds between 0.68 and 0.70 while quarterly spend climbs 37 percent, which is the single most useful signal this metric produces: the machine scaled without losing efficiency. Then Q1 2026 drops to 0.42.

The naive read is that efficiency collapsed. It did not. Q4 spend jumped to $2.6M because six account executives and two demand gen hires started in November, and Q1 is the weakest closing quarter in this company's history. Reps hired in November are carrying roughly a third of quota in Q1. The spend is real, the revenue is real, and the productivity gap is a ramp, not a failure.

Cutting in response to that 0.42 would strand six months of hiring cost and produce a genuinely broken Q3. The right move is to publish the number with the ramp overlaid, flag Q3 2026 as the quarter where the hires should clear 0.65, and hold. If they do not clear it, the diagnosis changes.

The magic number cannot tell the difference between a company adding $1.5M gross and losing $0.4M, and one adding $1.1M and losing nothing. Both produce the same net new ARR. Put the [SaaS quick ratio](/glossary/saas-quick-ratio/) next to it in the same table so growth and leakage are visible at once.

## Magic number and CAC payback are the same number in different clothes

They are algebraically linked, which means they should never disagree. If yours do, one of them has a different spend definition or a different revenue base underneath it.

```
CAC payback in months = 12 / gross-margin-adjusted magic number
```

| GM-adjusted magic number | Implied CAC payback |
|---|---|
| 0.40 | 30.0 months |
| 0.50 | 24.0 months |
| 0.60 | 20.0 months |
| 0.75 | 16.0 months |
| 1.00 | 12.0 months |
| 1.25 | 9.6 months |

Benchmarkit's 2025 B2B SaaS metrics work put median CAC payback at around 16 months, with the top quartile at six months or better. Read across the table: the median B2B SaaS company is running a gross-margin-adjusted magic number near 0.75, and the top quartile is above 2.0. That reframing is useful in a planning conversation, because 0.75 sounds mediocre while 16 months sounds normal.

The practical consequence is that you only need to instrument one of them properly. Build [CAC payback](/guides/cac-payback-period/) with a clean spend definition and the magic number falls out of it for free. Then check both against your [LTV to CAC ratio](/guides/ltv-cac-ratio/), because payback tells you about cash and LTV to CAC tells you about the return, and a 16 month payback on a customer who leaves in month 20 is not a business.

## Why PLG and low-touch products distort the metric

Self-serve products break the lag assumption in both directions, and the distortion is large enough that I would not put a quarterly magic number in a PLG board pack without a caveat next to it.

The spend that produces self-serve revenue is mostly content, product surfaces and word of mouth, and it pays back over 12 to 30 months rather than one quarter. A company like PostHog or Linear that invests heavily in documentation and developer content is booking the cost now against revenue that arrives through 2027. One-quarter lagging makes that investment look like waste for six consecutive quarters, then look like a windfall.

The second problem is scale. Below about $5M ARR a single enterprise deal moves the ratio by 0.2 with nothing else changing, and monthly billing makes the ARR line jump around on renewal timing alone. At $3M ARR with $1.4M in quarterly S&M, one $180,000 deal slipping from March to April swings your reported number from 0.62 to 0.49 and back.

Three adjustments make it usable for low-touch motions:

- Report a trailing four-quarter magic number, summing four quarters of net new ARR over four quarters of spend lagged by one.
- Split the denominator into demand capture spend (paid, events, outbound tooling) and demand creation spend (content, community, product-led surfaces), and only lag the first by one quarter.
- Add a self-serve-only line that excludes sales-assisted revenue entirely, so the two motions stop hiding each other.

The honest cost of doing this: your board loses a single comparable number and gains three that need explaining. That is a real tradeoff and some boards will not accept it. If yours will not, report the blended trailing four-quarter figure and keep the split internally.

## A magic number above 1.0 is a spending signal, not a trophy

This is where I disagree with how most boards use the metric. A gross-margin-adjusted magic number above 1.0 means every dollar you put in comes back inside twelve months at full contribution. There is no financial instrument available to you that does that. Treating 1.0 as a ceiling to protect is the most expensive mistake in this entire metric.

If you are above 1.0 and growing under 40 percent, you are capacity constrained, not efficiency constrained. Deliberately spend the number down to 0.8. Hire the reps, raise the paid budget in the two channels where you can name unfilled demand, and tell the board in advance that the ratio will fall and why. A planned fall from 1.1 to 0.8 buys materially more ARR than holding 1.1.

Three conditions have to hold before you do it. Runway above 18 months after the increase. NRR at or above 100 percent, because spending into a leaking base just refills a bucket. And a named channel with evidence of unmet demand, such as impression share you are not buying or a [pipeline velocity](/glossary/pipeline-velocity/) constraint you can point at, not a general belief that more spend produces more growth.

Who should not do this: anyone under 12 months of runway, anyone whose NRR is below 95 percent, and anyone whose 1.1 came from a budget cut two quarters ago rather than from demand. In those cases the high number is an artefact and spending against it is how companies die efficiently.

## How to instrument it so the number survives an audit

**Five steps to a magic number you can defend**

**Before this number goes in a board pack**

## What to do this quarter

Calculate both versions for your last four quarters before you do anything else. If the lagged and unlagged numbers differ by more than 0.1, your spend is moving fast enough that the version you have been reporting has been telling you the wrong story, and you should say so explicitly rather than quietly switching.

Then put the magic number, [CAC payback](/guides/cac-payback-period/), NRR and the quick ratio on one page. Those four together answer whether growth is efficient, whether it repays, whether it lasts, and whether it leaks. Any one of them alone can be defended into a bad decision. The [SaaS metrics library](/saas-metrics/) has the definitions and the calculation notes for each, and the [LTV to CAC calculator](/calculators/ltv-cac-ratio/) will tell you within a few minutes whether your payback period is survivable given how long your customers actually stay.

Last thing. Decide now, in writing, what you will do at 0.45 and what you will do at 1.15. Deciding the policy before the number arrives is the only reliable defence against arguing about the metric instead of acting on it.

## Frequently asked questions

### What is the SaaS magic number?

It is a sales efficiency ratio. Take the change in annualised revenue between two quarters and divide it by the sales and marketing spend of the earlier quarter. The result tells you how many dollars of new recurring revenue each dollar of combined sales and marketing spend bought. A result of 0.8 means 80 cents of new ARR per dollar spent.

### What is a good magic number for SaaS?

Above 0.75 on a gross-margin-adjusted basis is the usual threshold for funding more growth, because it implies a CAC payback under 16 months. Between 0.5 and 0.75 the motion works but repays slowly. Under 0.5 you are buying revenue at a price that will not clear the cost of capital, and more spend makes it worse.

### Should the magic number use the current or previous quarter's spend?

Previous. In a B2B SaaS business with a 60 to 90 day sales cycle, almost none of this quarter's new ARR was produced by this quarter's spend. Matching current growth to current spend mistimes cause and effect, and it flatters any company that has recently cut its budget. The lagged version is the standard and the honest one.

### How is the magic number different from CAC payback?

They are the same information. A gross-margin-adjusted magic number of M implies a CAC payback of 12 divided by M months. A magic number of 1.0 is a 12 month payback, 0.75 is 16 months, 0.5 is 24 months. If your two numbers disagree, one of them has a different definition of spend or revenue underneath it and you should reconcile them before either goes in a board pack.

### Does the magic number work for product-led growth companies?

Poorly, without adjustment. PLG revenue arrives from content, product surfaces and word of mouth that were paid for quarters earlier, so a one-quarter lag misattributes almost everything. Self-serve businesses should run a trailing four-quarter magic number and split paid acquisition spend from content and product-led spend, otherwise the ratio moves for reasons nobody can act on.

### What should I do if my magic number is below 0.5?

Stop increasing spend and find out whether the problem is conversion, price or retention. Check win rate and average contract value first, because both sit directly in the numerator. Then check whether contraction in the existing base is eating your gross new ARR. Cutting spend raises the ratio mechanically without fixing anything, so do not treat a cut as a result.

### How often should you calculate the magic number?

Quarterly, reported as a four-quarter trailing series rather than a single point. One quarter carries too much noise from deal timing, seasonality and hiring ramps. Plot four quarters side by side with the spend line underneath, and judge the trend rather than the level. A stable 0.68 while spend grows 30 percent is a better result than a one-off 0.9.