Every early-stage SaaS founder I have talked to tracks too many numbers and trusts too few of them. The problem is rarely the dashboard. It is that each metric is computed slightly differently by whoever built the sheet, so two people in the same company quote different churn figures in the same meeting.
This is the short list I would keep if I had to defend it: five numbers, each with one formula, each answering a question the others cannot.
1. Growth efficiency: the CAC payback period
Payback answers "how long until this customer has repaid what it cost to win them" and it is the metric that decides whether growth is self-funding or funded by the bank.
CAC payback (months) = CAC / (ARPU × gross margin)
Two things break it. Using revenue instead of gross profit understates payback by the margin: at 80% margin, $100 of ARPU is $80 of contribution, so payback is 25% longer than the naive version. And including only ad spend in CAC — not salaries, tools, or sales commissions — understates the denominator by a factor of two or three in a sales-led business.
Under 12 months is the usual target for SMB; 18–24 months is normal for mid-market. If it is over 24 and you are not venture-funded, the growth is being paid for out of the bank account.
The full method, including how to handle a blended versus paid-only CAC, is in the CAC payback period guide. The denominator is worth its own read: how CAC is defined changes more outcomes than the payback arithmetic does.
2. Unit economics: the LTV:CAC ratio
LTV = ARPU × gross margin × (1 / monthly churn)
ratio = LTV / CAC
Three times is the line, and it survives because below it every new customer is financed from cash. Above 5x, the usual explanation is underinvestment in acquisition rather than unusual efficiency — which is a problem, just a different one.
The single most common error is computing LTV on revenue instead of gross profit. At 50% margin that doubles the ratio; at 80% margin it inflates it by 25%. Two companies with identical economics can appear to have completely different unit economics purely because one skipped the margin term.
Benchmarks by segment, and the reasons the ratio moves, are laid out in the LTV:CAC ratio reference.
3. Retention: net revenue retention
Gross churn tells you what left. NRR tells you what the base did including expansion, contraction and cancellations:
NRR = (starting MRR + expansion − contraction − churn) / starting MRR × 100
Above 100% means the existing customer base grows with zero new sales — the strongest single signal in SaaS, and the reason enterprise businesses can grow while losing accounts. Below 90% means you are running to stay in place.
Report NRR and GRR (gross revenue retention) together. NRR alone hides how much retention is being propped up by upsells to a shrinking base. The distinction, and the formula in full, is in the net revenue retention breakdown.
4. Reporting hygiene: MRR versus ARR
MRR = monthly recurring revenue, normalized
ARR = MRR × 12
The trap is not the multiplication. It is what counts as recurring: implementation fees, one-off onboarding, overage, and annual contracts billed upfront all get misclassified, and each one changes both MRR and every ratio built on top of it. Annual contract values divided by twelve is the standard; anything else needs to be stated explicitly.
The normalization rules — including what to do with discounts, credits and mid-month upgrades — are covered in MRR vs ARR.
5. Solvency: the budget line
runway (months) = cash / net monthly burn
Where net burn is cash out minus cash in, and cash in means collections, not bookings. A company with strong bookings and slow collection can run out of money while reporting record revenue. Three months of runway is not a plan; eighteen is the point at which you can make multi-quarter decisions.
The startup runway guide covers gross versus net burn and how a single large annual prepayment distorts both.
If you want one composite number
The SaaS quick ratio collapses growth and retention into one figure:
quick ratio = (new MRR + expansion MRR) / (churned MRR + contraction MRR)
Above 4 is healthy, 2–4 is workable, below 2 means growth is being offset by losses. It is the fastest sanity check on a dashboard that otherwise looks fine.
The meta-problem: nobody agrees on the benchmarks
Most "healthy" thresholds circulating in pitch decks trace back to a handful of vendor surveys with different sample definitions. That is why a churn rate of 3% is excellent for one segment and alarming for another — and why comparing your number to a table without checking the segment is close to meaningless.
I keep a SaaS metrics benchmarks page that states the segment and source next to each band, precisely because the segments are what make the numbers comparable at all.
None of this is complicated arithmetic. The work is in the definitions, and in picking one definition per metric and writing it down where everyone can see it. Five formulas, agreed once, save a surprising number of meetings.
Top comments (0)