SaaS Metrics: Which Numbers Feed Which

How many SaaS metrics there are

No standard set exists, and published lists run from five metrics to twenty-eight. The disagreement worth your attention sits underneath that count, where one name covers different quantities and nobody says which one they mean.

The counts themselves are easy to line up: twenty-two, twenty-eight, fifteen, thirteen, five. Google suggests a top five beside this term, so anyone answering picks a scope before picking a number.

The widest of them runs to twenty-two metrics under five headings, each one named in a line of names.

Stripe, read 7 September 2026: twenty-two metrics under five headings

1. 1Acquisition, engagement, retention, growth and economics. The headings sort the metrics, and apart from a note on net revenue retention the page leaves them unconnected.
2. 2Lifetime value and acquisition cost sit in the same line under economics, and the page gives no arithmetic linking one to the other.
3. 3Customer churn rate sits two headings above lifetime value, and churn is a term in the most common lifetime value formula.

Sorting metrics into groups makes a long list readable. It also makes every entry look like a separate measurement, and several of the entries on this page are not.

Nine different answers is not a scandal. Different readers need different depth, and a five item framework for a finance team and a twenty-two item catalogue for a product team can both be right.

A venture firm put its own count in a title in 2015 and named the trouble in the same piece. People use different definitions of the same metric, and that makes the health of a business hard to read.

Which metric feeds which

A flat inventory treats every metric as separate. A circuit shows how one changes another. Lifetime value is the clearest case: one published definition has it incorporate acquisition cost, dollar retention and gross margin. Three metrics from that same list are inputs to a fourth one on it.

The phrase that decides everything below sits three lines before his list, in the definition itself.

Sacks and Ruby, The SaaS Metrics That Matter: lifetime value, net of acquisition cost

1. 1The definition says net of CAC. Acquisition cost is subtracted before the number is reported.
2. 2The next sentence names the three things it incorporates, and Stripe carries no equivalent statement.
3. 3The same paragraph asks for lifetime value to reach three times the original acquisition cost. That target is not the same number as a 3:1 ratio of the two.

Once one metric is an input to another, a change in the first one moves the second one without anything else happening. Churn is the clearest case, because it sits in the denominator of the published lifetime value formula.

A quarter like that produces a slide saying acquisition efficiency doubled. Acquisition did nothing. Retention did all of it, and the ratio reported the result under the other department's name.

Three chains worth naming

Churn feeds lifetime value, which feeds the LTV to CAC ratio. Cost per lead feeds acquisition cost, which feeds the same ratio from the other side. Win rate and deal size feed pipeline velocity, and the same compression happens there.

None of the three is exotic. All three are invisible in a list that prints the metrics as siblings.

Look up the definition before you compare

A metric name is not a metric. Before two numbers can be compared, both have to be computed the same way. The four numbers this page sends you to are each defined in one place with one arithmetic. Read the definition first and the benchmark second.

Here is what each name on the common lists answers, and where a full definition sits if this site has one.

The benchmarks you would compare against split on this. ChartMogul aggregated more than 2,100 SaaS businesses over the twelve months to March 2023 and publishes a methodology and glossary with the formulas beside the numbers.

SaaS Capital surveyed more than 1,000 private B2B SaaS companies for its fifteenth annual edition and reported a median growth rate of 22 percent. Its research brief does not say how growth was calculated.

Two benchmark reports, similar samples, and only one of them lets you check that your number and theirs are the same quantity. Read the methodology page before you read the percentile.

The lookup takes a minute and it catches the expensive case. Two teams reporting lifetime value from two formulas will argue about the business when a divisor is what they disagree about.

Say you are handed a lifetime value of $13,333 and a benchmark of 3:1. Four checks decide whether those two numbers belong in the same sentence.

1. Write down which formula produced your number, not just the number.
2. Check whether acquisition cost is already inside it. That one line settles the section below.
3. Check the window. Twelve times a monthly churn rate is not the annual rate, because churn compounds against a shrinking base.
4. Only then compare against a benchmark, and only against one computed the same way.

Steps two and three are the ones that get skipped, and they are the two that change the answer.

What a mismatched lifetime value costs

One name carries two quantities here, and the gap between them lands on either side of the 3:1 target. One definition takes lifetime value net of acquisition cost. The other takes it as revenue times gross margin divided by churn, with acquisition cost left in. Dividing the first by acquisition cost counts that cost twice.

A third definition is in circulation with a sample behind it. Benchmarkit, whose 2024 report draws on roughly 1,000 B2B SaaS companies reporting 2023 figures, publishes lifetime value as ARPA times subscription gross margin, divided by one minus gross revenue retention.

That version has no churn denominator and no acquisition cost in it at all. Three published formulas, one name, and your ratio moves depending on which one your board read last.

For instance, run both definitions on your own company, in one quarter, with one set of inputs.

A full point of ratio, from a definition mismatch. One reading clears the 3:1 line the overview calls healthy and the other does not, on the same company in the same quarter. Which of the two reaches your board depends on which formula produced it.

Google’s own AI answer inherits the same ambiguity. In the version recorded on 27 August 2026 it defines lifetime value as total gross profit or revenue from a customer.

That is two quantities in one sentence. The same overview carries the 3:1 target, and if you take it at face value the target holds for only one of them.

Getting this wrong is not a reporting embarrassment. Your budget moves, against a number computed for a different question.

Suppose the acquisition number is the one moving your budget: that is our SaaS SEO service.

A metric definition list, the field the formula is typed intoWhere the disagreement about a definition lives

Two rows can carry the same metric name and a different formula, and the tool accepts both. A report reads whichever row someone saved. The name travels between teams and the formula field stays behind.

Where the lists go wrong

The lifetime value mismatch is one instance of a wider habit, and four of them show up around these numbers. None is an arithmetic mistake. Each is a reporting reflex that survives because a metric name looks self-explanatory, and each costs a reader more than a wrong formula would.

The second habit is the expensive one and also the easiest to fix. A formula beside a metric costs you one line and removes the entire class of error this page describes.

For example, a slide reading LTV $13,333 tells you nothing about whether acquisition cost is already out of it. The same slide reading LTV $13,333, revenue times margin divided by churn, settles that before anyone asks.

The third gets missed because churn compounds. Three percent a month is not thirty-six percent a year, and a page quoting one band against the other is comparing two different numbers under one name.

Nine counts above, and seven came from summaries. Where a summary is quoted the page says so and quotes it as far as it goes.

All four habits share one cause: a number reaches you without the arithmetic that made it. That is a travel problem, and it has a one line fix.

Report the definition beside the number

A SaaS metric travels badly. It leaves the spreadsheet where it was computed, lands on a slide, and arrives without the arithmetic that produced it. The fix costs one line per number, and it works whether or not anyone agrees on which metrics belong on the list.

Three things travel with a metric or you should not let it travel: the formula, the window, and the date your inputs were pulled.

The habit is not unreasonable to ask for, because published reports do it. Benchmarkit prints its CAC ratio as total sales and marketing expenses divided by new customer ARR, on the same page as the number.

Its 2025 page carries the formulas and does not state how many companies are behind them, so you get half the line. Half is still more than a bare percentile gives you.

That is the whole discipline. Unglamorous, and it settles every disagreement on this page. Both sides of an argument about lifetime value can see in one glance whether the company or the divisor is what they disagree about.

For the four metrics defined on this site the formula and the window sit in the entry. Send a colleague who disagrees with a number there. SaaS churn rate holds the retention arithmetic and its windows, and cost per lead holds the acquisition one.

For the five with no definition here, name them and say the definition sits elsewhere. A named gap costs you less than a definition recalled from memory, and the table above takes that advice.

Everything on this page is produced somewhere before it is reported, and the SaaS marketing funnel is where the acquisition side of it happens.