SaaS metrics investors ask for and how to define them

The SaaS metrics investors ask for are ARR and the waterfall behind it, net and gross revenue retention, gross margin, net burn, burn multiple, CAC payback and runway. Each has a standard formula, but a figure only means something when its definition is written down and applied the same way every month, because an investor will rebuild it from the underlying data before trusting it.

This guide is for founders and finance leads of seed to Series B software companies preparing board reporting or a raise. The deeper treatment of retention is in net revenue retention and GRR, and burn and runway are covered in burn rate, burn multiple and runway.

Which SaaS metrics do investors ask for?

Investors ask for metrics that answer four questions, and the core set maps onto them:

• How fast is the company growing? ARR, net new ARR and the ARR waterfall.
• Do customers stay and expand? Net revenue retention (NRR), gross revenue retention (GRR) and logo churn.
• What does growth cost? Gross margin, burn multiple and CAC payback.
• How long does the money last? Net burn and runway.

Annual recurring revenue (ARR) is the annualized value of the recurring subscription contracts in force at a point in time. Monthly recurring revenue (MRR) is the same measure on a monthly basis, so ARR is MRR multiplied by twelve. Most of the metrics below start from ARR, which is why its definition comes first.

SaaS metrics definitions: formulas and what investors check

The table gives the standard formula for each metric and the first thing a diligence team tests.

The burn multiple was introduced by David Sacks of Craft Ventures in his essay The Burn Multiple, as a measure of how much a startup burns to generate each incremental dollar of ARR. The higher the multiple, the more cash each unit of growth costs.

Write the definitions down before reporting the numbers

A metric definitions page is one document that states how every metric is calculated and the choices behind it. Without it, two people in the same company can report different ARR figures and both be right on their own terms. The page should settle:

1. What counts as ARR. How to treat one-time fees, implementation and services, usage above the contracted minimum, multi-year price steps and free months.
2. When a customer counts as churned. At contract end, at notice, or when ARR reaches zero. Pick one.
3. How returning customers are treated. New revenue or a reactivation, applied the same way every time.
4. The retention method. Trailing twelve months or monthly annualized, and how the starting cohort is selected.
5. What sits in cost of revenue. Hosting, customer support, onboarding and third-party software that ships with the product.
6. What burn includes. Whether debt service and capital purchases are inside net burn. Most investors expect everything that reduces cash except financing inflows.
7. The CAC payback choices. Whether expansion ARR is included and whether spend is lagged by a quarter.

When a definition changes, note the date and restate prior periods on the new basis. A trend that mixes two definitions is not a trend. Public companies are held to the same idea: SEC guidance on key performance metrics expects each metric to come with a clear definition of how it is calculated, and when the method changes, disclosure of the change and consideration of whether to recast prior periods (SEC Release No. 33-10751).

Build every metric from one data set

The metrics should reconcile to each other, which only happens when they come from the same sources:

• Customer-level ARR by month, one row per customer, built from contracts or billing. It produces the ARR waterfall, and the waterfall produces net new ARR, NRR, GRR and churn.
• The books. The income statement gives revenue, cost of revenue and sales and marketing spend for gross margin and CAC payback.
• The bank and the cash flow statement, which give net burn and the starting point for runway.

Tie them out every month. Beginning ARR plus the four movements must equal ending ARR, ending ARR should reconcile roughly to recognized subscription revenue, and net burn should match the change in cash once financing is removed.

A worked example

One fictional software company, measured over the trailing twelve months. It started the year with 50 customers and lost 6 of them. All figures are $K except customers, ratios and months.

Illustrative figures, not client data.

All expansion, contraction and churn come from customers in the beginning ARR, so retention can be read straight off the waterfall. The company also burned $1,800K over the year, spent $1,350K on sales and marketing, runs a 75% gross margin, and holds $3,600K of usable cash.

Illustrative figures, not client data.

What definition drift does to the same company

Now suppose the company counts $150K of one-time implementation fees as new ARR. New ARR becomes 1,050 and net new ARR 1,150.

Illustrative figures, not client data.

Every efficiency metric looks better and none of it is real. The second classic error is including new customers in retention: (2,000 + 900 + 300 - 80 - 120) / 2,000 gives 150%, which is ARR growth, not NRR. An investor who rebuilds ARR from contracts finds both, and then questions every other figure in the deck.

How an investor reads your metrics

Decipher sets up a company's finance the way an investor or acquirer will read it. On SaaS metrics, the same checks come up in almost every diligence process.

• Is there a definitions page, and has it changed? The first request is usually how each metric is defined. A changed definition without restated history reads as a red flag even when each figure is defensible.
• Can they recompute it? Investors ask for customer-level ARR by month and the cash flow statement, then rebuild ARR, retention and burn. Their numbers should match the company's.
• Does it tie to the books? Ending ARR should reconcile to contracts and roughly to subscription revenue, and net burn to the change in cash.
• Is NRR shown with GRR? A strong NRR alone invites the question of how much churn expansion is covering. Cohort tables answer the next question.
• Is growth efficient? Burn multiple and CAC payback sit next to retention. Efficient growth from a base that keeps its customers is worth more than the same growth bought with heavy spend.
• What is the trend? Several quarters on stable definitions say more than one strong period.
• Is runway honest? A downside case, usable cash rather than the bank balance (net of any minimum balance set by venture debt covenants), and no money counted before it arrives.

Common mistakes

Beyond definition drift and new customers in NRR, three errors come up often:

• Bookings in ARR. A signed contract that starts next quarter, or the full value of a multi-year deal, is not ARR today.
• Mismatched periods. Burn for a quarter against net new ARR for a year, or spend in one quarter against ARR won in another, distorts the ratio.
• Thin cost of revenue. Leaving hosting or support out of cost of revenue flatters gross margin, which then flatters CAC payback.

Decipher sets up and reports these metrics as part of its fractional CFO for SaaS startups and its FP&A services.