A board member asks one question on the efficiency slide: "For every dollar you put into sales and marketing last year, how many dollars of new recurring revenue came back?" The founder pulls up a number. It says 1.1. The room nods. Everyone agrees the growth engine is efficient and it is time to spend more.
Six months later the company has doubled its sales headcount, burned through the round faster than planned, and growth has not moved the way the model promised. The 1.1 was real. It was also wrong, because the spreadsheet behind it counted bookings as ARR, compared this quarter's revenue to this quarter's spend, and blended a healthy new-business motion with an expansion motion that flattered the whole thing.
I have seen this exact movie at three companies in the last two years. The magic number is one of the most useful metrics a B2B SaaS team has, and one of the easiest to calculate in a way that lies to you. This is a RevOps problem before it is a finance problem, because the inputs live in your CRM and your billing system, and almost nobody has wired them up to tell the truth.
What the magic number actually measures
The SaaS magic number measures sales and marketing efficiency. It answers a simple question: for every dollar you spent acquiring customers, how much annual recurring revenue did you get back within a reasonable window.
It came out of Scale Venture Partners years ago as a quick way to judge whether a software company should pour more fuel on the fire or fix the engine first. A high number means each dollar of go-to-market spend is producing a lot of new ARR, so spending more is a good bet. A low number means you are buying revenue at a bad price, and adding reps or ad budget will just lose money faster.
That is the whole idea. It is a price tag on growth. The reason it matters more in 2026 than it did in 2021 is that capital is no longer free. Investors stopped rewarding growth at any cost somewhere around 2022, and the magic number is the cleanest single read on whether your growth is the efficient kind.
The magic number is only as honest as your CRM and billing data.
The formula is trivial. The inputs are where teams go wrong. Fix the inputs and the number becomes a decision tool instead of a vanity slide.
The formula, and the lag everyone gets wrong
Here is the standard version:
Magic number = (current quarter ARR minus previous quarter ARR) x 4, divided by previous quarter sales and marketing spend.
You take the change in ARR from one quarter to the next, multiply by four to annualize it, then divide by what you spent on sales and marketing the quarter before. The result is a ratio. A magic number of 0.8 means every dollar of S&M produced 80 cents of new annual recurring revenue.
The part most teams skip is the word "previous" in front of "sales and marketing spend." This is the lag, and it is the single most common mistake I see. You compare this quarter's revenue growth against last quarter's spend, not this quarter's spend. Why? Because the money you spend on a sales rep or a campaign does not turn into revenue the same week. It takes a sales cycle.
If your average B2B deal takes 90 days to close, a one-quarter lag is roughly right. If your deals take six to nine months, a one-quarter lag is too short and your number will look worse than reality, because you are crediting this quarter's revenue to spend that has not had time to pay off yet. Match the lag to your actual sales cycle. If you do not know your sales cycle length cold, that is the first data problem to fix, and we wrote a whole piece on why B2B deals take 134 days and what to do about it.
What good looks like in 2026
People love to quote 1.0 as the target. It is the wrong anchor for most companies. The median private SaaS magic number sits closer to 0.7, which is healthy and ordinary. Chasing 1.0 as a baseline makes a normal business feel broken.
Here is how I read the bands.
Below 0.5, adding sales and marketing spend mostly buys you a bigger loss. The right move is to fix conversion, pricing, or targeting before you scale. Between 0.5 and 0.75 you have a working motion that costs too much, so you tune it. Above 0.75 the engine is good and you can press the gas. Above 1.0 you are in rare company, and the question flips to why you are not spending more.
Two caveats that matter. First, stage changes everything. A Series A company at 3M ARR can post a 1.2 because two big deals landed in the same quarter, and that says almost nothing about durable efficiency. A 100M ARR company at 0.6 with nine-month enterprise cycles might be doing great. The magic number gets more meaningful as your deal flow gets less lumpy. Second, AI-native SaaS companies have been posting magic numbers above 1.0 through 2025 and into 2026, well ahead of the rest of the market, partly because of pull-driven demand and partly because they have not hit the expensive part of the curve yet. Do not benchmark yourself against them.
Why your magic number is probably wrong
The formula takes ten seconds. The reasons your number is misleading take a lot longer to fix, because they are data problems. Here are the four I run into most.
You are using bookings, not ARR
This is the big one. Bookings is the total contract value you signed. ARR is the annualized recurring portion. If you sell a two-year deal with a setup fee and some services, the booking is much larger than the ARR, and counting bookings inflates your magic number. Services revenue and one-time fees do not recur, so they have no business in a metric about recurring revenue efficiency. Strip them out. Use net new ARR, full stop.
Your S&M lag does not match your sales cycle
Covered above, but it belongs on the list because it is so common. A team with a 200-day sales cycle using a one-quarter lag is reading a number that is structurally too low. They conclude the engine is broken when it is just slow, and slow is a different problem with a different fix.
You blend new business with expansion
Net new ARR usually includes expansion from existing customers. That is fine for a company-wide read, but it hides the thing you most need to know. Expansion is cheap. Landing a new logo is expensive. If your blended magic number is 0.9 but your new-business magic number is 0.45, your acquisition engine is underwater and your existing base is carrying the whole metric. You would never know from the blended figure. Split them.
Churn is dragging your number down
Net new ARR is calculated after churn. If you are losing 20% of your base every year, you have to spend just to stand still before any of that spend shows up as growth. High churn makes a perfectly good acquisition engine look inefficient. The fix is not in the magic number. It is in net revenue retention and onboarding. But you need to know churn is the cause before you go cutting sales budget that is actually working.
There is a fifth quiet mistake worth naming: stuffing the wrong costs into the S&M line. Customer support, product engineering, and general overhead are not customer acquisition costs. Putting them in the denominator makes acquisition look more expensive than it is and pushes you to cut the wrong things. Sales salaries, commissions, marketing program spend, SDR costs, and the tooling that runs all of it belong there. Support tickets do not.
The segmented magic number is where the truth lives
The single most useful change you can make is to stop reporting one number. A blended company-wide magic number is an average, and averages hide the two things you need: which motion is efficient and which one is bleeding.
Cut it by the dimensions that actually change how you spend:
- New business versus expansion, always
- By segment, because SMB and enterprise have completely different cost structures and cycle times
- By channel, so inbound, outbound, and partner each get their own read
- By region, if you sell across geographies with different cost bases
When you do this, the picture usually breaks into a clear story. Inbound SMB might be running at 1.3 while outbound enterprise sits at 0.4. The blended number says 0.8 and tells you to keep doing what you are doing. The segmented numbers tell you to pour money into inbound SMB and either fix or shrink the outbound enterprise motion. That is a real decision, and you can only see it after you cut the data. Your customer segmentation has to be solid for this to work, which is one more reason segmentation is foundational rather than a nice-to-have.
How to build it as a RevOps system
The reason most teams calculate the magic number once a quarter in a spreadsheet, by hand, is that the data is not wired up. ARR lives in billing. Spend lives in finance. Both need to be tagged by segment and motion, and almost nobody does that tagging at the source. Here is the build I run.
The hard part is step two, not step four. Getting clean motion and segment tags on every deal and every cost line is unglamorous data work, and it is exactly where we spend most of our time on a CRM and RevOps engagement. Once the tags exist, the joining and the math are easy. We usually run the join through HubSpot custom properties for the deal side and a lightweight pipeline for the finance side, and we have written before about CRM data quality because none of this works on a messy pipeline.
If you want the automation to run on your own infrastructure rather than a stack of point-to-point connectors, that is a data and AI automation build, and the same plumbing that computes your magic number weekly can feed your other efficiency metrics too.
What the number should actually change
A metric is only worth tracking if it changes a decision. The magic number drives three real ones.
Budget. If your new-business magic number is above 0.75, the case to increase sales and marketing spend is strong, because each new dollar is paying off. If it is below 0.5, the honest move is to hold spend flat and fix conversion or targeting first. This is the same logic that should drive your sales capacity planning. Adding reps into an inefficient motion just scales the inefficiency.
Where to spend, not just how much. The segmented version tells you which channel and segment deserve the next dollar. A blended number cannot do that.
When to raise. Investors read the magic number as a proxy for whether you have a repeatable engine. Walking into a round with a clean, segmented number that holds up under questions is worth more than a blended figure that falls apart the moment someone asks how you defined ARR. Pair it with your CAC payback period and Rule of 40 and you have the efficiency story most boards actually want.
The magic number will not tell you why your engine is efficient or broken. It is a smoke detector, not a diagnosis. But a smoke detector that you have wired correctly, segmented properly, and lagged to your real sales cycle will tell you the truth, and the truth is what lets you decide whether to spend more or fix what you have first.
Not sure your magic number tells the truth?
We will audit how your ARR and S&M data feed the metric, find where it lies, and wire up a segmented version you can trust. Book a free 30-minute RevOps audit.
Book an audit →FAQ
What is a good SaaS magic number?
For most private B2B SaaS companies, anything above 0.75 signals efficient growth and a case to spend more. The median sits around 0.7, which is healthy and ordinary. Below 0.5 means you should fix conversion, pricing, or targeting before adding spend. Read the bands by stage, since a Series A company and a 100M ARR company should interpret the same ratio very differently.
How do you calculate the SaaS magic number?
Take the change in ARR from one quarter to the next, multiply by four to annualize it, then divide by the previous quarter's sales and marketing spend. The "previous quarter" part matters, because spend takes a sales cycle to turn into revenue. Use net new ARR, not bookings, and strip out services and one-time fees.
Why does the lag in the formula matter?
Money spent on a rep or a campaign does not produce revenue the same quarter. It takes a full sales cycle. The standard one-quarter lag fits a 90-day cycle. If your deals take six to nine months, lag the spend two or three quarters instead, or your number will look worse than your engine actually is.
Should I use bookings or ARR for the magic number?
ARR, every time. Bookings include total contract value, services, and one-time fees, none of which recur. Counting them inflates the number and hides how efficient your recurring-revenue engine really is. Net new ARR is the only input that gives you a decision you can trust.
What is the difference between the magic number and CAC payback?
They measure the same thing from different angles. The magic number is ARR produced per dollar of S&M, read as a ratio. CAC payback is the number of months it takes to earn back the cost of acquiring a customer. The magic number is faster to compute at the company level. CAC payback is better for per-customer and per-segment unit economics. Most boards want both.
Build a magic number you can actually trust
The formula is easy. Clean ARR, properly lagged spend, and segment tags on every deal and cost line are the hard part, and that is RevOps work. If your efficiency metrics fall apart the moment someone asks how you defined them, let's talk. We will help you wire the data so the number tells the truth, the first time and every quarter after.