Monthly recurring revenue: calculating MRR so the number means something

Monthly recurring revenue is the predictable subscription revenue a business expects in a given month, normalised to a monthly figure. It matters because it is the one number that answers "what will we earn next month if nothing changes" — and because, unlike total revenue, it separates the part of the business that repeats from the part that has to be won again.

It is also the number most commonly overstated, usually without anyone intending to. One-off setup fees, hardware, consultancy days and expenses get included because they arrived in the same invoice, and the resulting figure predicts nothing.

What counts and what does not

• Include recurring subscription fees, normalised monthly — an annual contract counts as one twelfth per month, not as a spike in the month it was signed.
• Include committed recurring add-ons and extra seats.
• Exclude one-off fees: setup, implementation, training, hardware, professional services. These are real revenue and they are not recurring.
• Exclude usage that varies and is not committed, unless you track it separately as variable revenue.
• Exclude tax, and exclude anything invoiced but not yet in service.
• Handle discounts by counting what the customer actually pays, not list price. Contracted discounts are part of the price.
• Handle trials at zero until they convert. Counting them is the fastest way to a number nobody believes.
• Handle non-payers explicitly: a subscriber who has stopped paying is churn waiting to be recorded, not revenue.

Normalise annual contracts, and say so where the number is reported. Counting a year's contract as a single month's MRR produces a spike and a cliff that make every trend line useless. It also makes the same business look wildly different depending on when contracts happen to renew.

The four movements

1. Start with opening MRR — last month's closing figure.
2. Add new MRR from customers who were not paying last month.
3. Add expansion MRR: upgrades, extra seats and add-ons from existing customers.
4. Subtract contraction MRR: downgrades and reduced seats from customers who stayed.
5. Subtract churned MRR from customers who left entirely.
6. The result is closing MRR, and the four movements explain the whole change. A business growing because expansion offsets heavy churn is in a very different position from one growing on new sales, and the total alone hides which you are.
7. Track net revenue retention — expansion minus contraction and churn, as a percentage of opening MRR from existing customers. Above 100% means the existing base grows without any new sales at all.
8. Reconcile MRR against invoiced revenue quarterly. Drift between the two is normal; unexplained drift means a definition is being applied inconsistently.

Where MRR misleads

Two cases deserve care. A business with heavy one-off revenue can have modest MRR and be perfectly healthy — MRR is not a measure of size, only of predictability, and treating it as the headline number distorts decisions in a services business. And MRR says nothing about cash: annual contracts paid upfront generate cash long before the MRR is earned, while monthly billing does the opposite. Two companies with identical MRR can have completely different bank balances, which is why the cash-flow forecast stays a separate document.

Ettex Sheets is where the model belongs: import an existing XLSX with formulas intact, one row per month with columns for opening, new, expansion, contraction, churn and closing, formulas so closing is calculated rather than typed, conditional formatting to flag a month where churn exceeds new, charts for the twelve-month trend, pivots by plan or segment, and cell comments so an unusual movement carries its explanation. Version history means the figure as reported last quarter is recoverable. The invoices behind it live in Ettex Invoices with line items, rates and statuses, the ledger side in Ettex Books, and the customers whose upgrades and cancellations drive the movements sit in Ettex CRM.

Plainly: Ettex does not calculate MRR. There is no subscription management, no revenue recognition, no cohort or retention analytics, and no dashboard that updates as customers upgrade or leave. You export the invoice data and maintain the model. For a business with tens or low hundreds of subscriptions that is a monthly half-hour; at larger scale you want billing software that reports this natively.

Common errors

• One-off fees included, which makes the figure unable to predict anything.
• Annual contracts counted in the month signed, producing spikes and cliffs.
• Trials or unpaid signups counted as revenue.
• List price used instead of the discounted price actually paid.
• Only the total tracked, so nobody can see whether growth comes from new sales or from expansion masking churn.
• Non-payers left in the base for months before being recorded as churn.
• MRR treated as cash, leading to spending that the billing cycle has not funded.
• Definitions changed mid-year without restating history, which destroys the trend.

Monthly recurring revenue is only useful if it excludes what does not recur and is split into new, expansion, contraction and churn. Normalise annual contracts, count what customers actually pay, and never confuse it with cash.