What Financial Metrics Should a Recurring-Revenue Business Track Monthly?

Five metrics, not fifteen. Monthly recurring revenue and how it moved. Net revenue retention. Gross margin on the recurring line. Cash collected against revenue recognized. CAC payback, in months. Net revenue retention above 100% is the one that decides whether growth compounds or you are just refilling the bucket.

Which five metrics actually earn a monthly look?

Dashboards multiply until nobody opens them. Thirty tiles, no decisions. These five stack on the three reports you already read every month, and each closes a hole the others leave open.

Every band below comes from the service books I clean up, not a published benchmark.

Most books this size are mixed. 60% retainer, 40% project. Run these on the recurring slice only, or one $600K build makes a quarter look broken. Under a third recurring? You own a project business with retainers attached.

Twenty minutes a month, once the books close clean.

Why MRR is not last month's revenue

MRR is money a customer pays you again next month, contract or not. A month-to-month retainer with no term counts. Normalize each to a monthly figure, add them up. Then strip what feels recurring and isn't: setup fees, implementation, one-time projects, hardware, an overage nobody repeats. Real revenue. Not base.

The level matters less than the movement.

Break the change into four buckets and read them separately: new, expansion, contraction, churn. A month where MRR grew $8,000 describes two different companies. One added $8,000 and lost nothing. The other added $30,000 and gave $22,000 back to downgrades and cancellations. Same headline. One compounds. The other runs to stand still.

What do churn and retention really tell you?

Churn is the leak, and retention is whether the bucket refills itself.

Customer churn is the share of accounts cancelling this month. In the retainer books I clean up, 1% to 2% is ordinary and invisible in the P&L, because new sales paper over it before anyone reads the trend. Push past 3% and you lose roughly 30% of your base a year, compounded, never multiplied by twelve. Track dollars beside accounts. Three small cancellations and one big one look identical in a headcount. That is not a sales team growing the company. It is one maintaining it.

Net revenue retention is sharper, and almost nobody under $5M calculates it. Take the recurring revenue from customers you had twelve months ago. Look at what those same customers pay today, counting upgrades to their rate, downgrades, and cancellations. Exclude everyone won since. Divide.

Worked, since prose trips people up. A year ago, forty customers paying $120,000 a month between them. Today those forty pay $126,000, after $18,000 of upgrades, $4,000 of downgrades, and $8,000 gone. 126 over 120 is 105%. The eleven accounts signed in March appear nowhere in it. Bill annually and nobody moves until renewal, so a book with renewals stacked in one quarter flatters you until it lands. Above 100%, the base grew on its own while you slept. Below 100%, every new sale patches a hole first.

Gross margin on the recurring line, and what counts as cost

Everyone lists this one, almost nobody defines it, and the definition is the whole metric.

Recurring gross margin is recurring revenue minus what it costs to deliver. The only real question is what counts as delivery. My line: if a cost scales with how many customers you serve, it belongs there. Account managers, delivery staff, contractors, per-seat software, hosting. Sales, marketing, admin, rent, and your time running the company stay below. Owner labor is where service books usually break. Deliver forty hours a week of client work and a market-rate slice of your pay is delivery cost, paid or not. Miss it and margin reads twenty points high, and payback lands several months too optimistic. Two firms with identical economics can report 45% and 70%. Only one is measuring anything.

In the books I clean up, this lands at 50% to 65% once owner time sits inside. Observed, not published. Holding the definition still matters more than the level, because the trend is the signal and drift erases it. Margin sliding while headcount climbs means you service accounts with people instead of a system.

Why doesn't the cash ever match the revenue?

Recurring models bend cash and profit apart, and the direction depends on how you bill. Collect a year up front and cash lands in one lump while revenue trickles onto the P&L a twelfth at a time. That gap sits on your balance sheet as deferred revenue. A liability, not a win. You already have the money. You still owe the work.

That assumes two things. First, accrual books: plenty of businesses this size run cash basis, where no deferred balance exists and the prepayment hides in the bank account. Second, tax does not wait. Collect twelve months up front on cash basis and it is taxable the year it lands. Plan on it carrying its own tax bill, and ask your preparer what your structure allows.

Bill monthly and it inverts. You pay to acquire and onboard now, then recover over the next year or two, so a fast-growing month drains the bank while the P&L looks healthy. Purest form of the profitable but always short on cash trap.

Watch the deferred balance. Rising means you collected ahead of delivery. Falling can mean you are living on last year's prepayments, or just that you moved clients to monthly billing. Find out which first.

Collected is not earned.

How long until a new customer pays for itself?

Take everything you spent winning new customers last month: sales payroll, commissions, marketing, agency fees. Strip out whatever aimed at existing accounts, since expansion spend belongs to customers you have. Divide by the new MRR you added, then by recurring gross margin as a decimal. Months. With numbers: spend $40,000 to win $5,000 of new MRR at 60% margin. $40,000 over $5,000 is 8, then 8 over 0.6 is 13.3. Thirteen months.

Two corrections people skip. Spend closes business one to three months later, so line this month's outlay against the MRR landing one to three months out. And a paid-back account is not free money, just contribution margin, minus the share who cancel first.

In the recurring books I see, inside 12 months is comfortable and past 18 means you finance growth out of working capital. Survivable with reserves. It is how a growing company dies without them. That spread traces back to which service lines actually make money.

One warning. Blended payback hides the same way blended margin does. If enterprise accounts pay back in six months and small ones take thirty, your blended eleven-month number describes a company that doesn't exist.

The honest answer

The real question underneath this one is never "which metrics." It is "is my growth real, or am I renting it?"

Two companies, same revenue chart. One compounds. The other sells hard every month to stand still, and when sales slow there, revenue follows on a lag equal to your contract length. You find out late.

All five depend on something the question never mentions: bookkeeping that can produce them. Separate income accounts for recurring and one-time work, so the split survives the close instead of being rebuilt by hand. A service item on every invoice line. Contract start, renewal, and cancellation dates in one place. CRM or spreadsheet, as long as it is one. And cancellations dated when they happened, not when somebody got to them.

None of that math is hard. The tagging is the work, and it is why most recurring-revenue businesses that can't produce these numbers don't have a metrics problem. They have a bookkeeping problem wearing a metrics costume.

Tag the contracts, close the month clean, then read the five.

A one-time build off a chart of accounts that separates recurring from one-time. Not a hire.