Calculate Net Revenue Retention Without Hiding Losses

Fix the Starting Group Before Doing the Sum

Choose the opening and closing dates and keep the revenue definition consistent. For an annual view, compare the recurring amounts at those dates for the clients active at the start. Keep a stable identifier for each client so a name change doesn't create a fictional loss and win.

Use that boundary in the working sheet. A new sale can increase the business's total MRR, but it can't repair retention from a group it never belonged to.

Separate managed fees from recurring products, then show a combined view only if you define it clearly. Keep hardware and one-time projects out of the recurring-revenue calculation. If a classification changes, explain and reconcile it rather than leaving the reader to guess why the series moved.

Show Each Movement in the Recurring Fee

Start with the opening amount and reconcile it to the ending amount.

I'd split expansion into price changes and changes in what the client buys. The distinction lets a buyer see whether the movement came from more work, a different service, or the annual contractual step. Keep the supporting amendment or billing record beside the movement.

Record lost-client revenue separately from reduced scope at a continuing client. Keep both in the reconciliation. A report that groups them into an unexplained net adjustment makes it harder to understand the revenue that disappeared.

Work Through a Loss and a Price Increase

For a hypothetical year using that starting book, assume one model-sized client leaves. Assume the remaining clients all receive the model's contractual 5% annual increase by the closing date, with no other scope or price movements. These are scenario assumptions, not reported retention results.

NRR is $160,650 divided by $157,500, which equals 102%. The opening group produces more recurring revenue at the end despite losing a client. The increase doesn't prove the departed client was satisfied or that delivery improved.

If an increase hasn't taken effect by the closing date, don't include it in the ending amount. If seats or scope change, add the actual movement rather than preserving this tidy example. The calculation is only as useful as those inputs.

The ending MRR is a point-in-time recurring amount. It isn't the cash collected or total revenue earned during the year. The dates on which the client left and the increase took effect matter to those separate calculations.

Put Gross Retention and Client Count Beside NRR

Gross revenue retention, or GRR, measures the starting revenue retained after losses and contraction, excluding expansion. It shows the erosion that an expansion amount can otherwise offset.

In this example, GRR is $153,000 divided by $157,500, or about 97.14%. Client retention is the number of starting clients still present divided by the starting client count. Here, 34 divided by 35 also gives about 97.14%.

Those percentages match because every starting client has the same fee in this example. They needn't match in a real book with different client sizes. Losing a larger client changes revenue retention more than losing a smaller one, while each still counts as a client loss.

The model's 95% annual retention is a planning assumption. Don't label it an observed NRR or use it as a substitute for these calculations. State the measure and its basis before comparing your result with anything else.

A healthy-looking percentage isn't a reconciliation. Keep the lost-client line visible even when NRR exceeds the opening revenue. Then the conversation can address what actually changed.

Keep the Historical Method Consistent

For each completed year in the requested history, fix that year's opening client group and use the same method. A client won in one year can belong to the next year's starting group. Don't silently change whether it qualifies midway through a comparison.

Keep the client-by-client sheet and where each number came from. If you changed how you code revenue, or the currency, say so, or last year won't compare. Show the revised basis when a correction is needed.

Ryan Spillane describes answering financial-diligence questions with written procedures alongside the way the work appeared in the accounting system, then exporting requested records. I'd apply that approach here. Be able to demonstrate how the retention sheet was built rather than simply saying the data is documented.

Explain the Loss Without Inventing Its Cause

For a departed client, record the effective date and the reason you can support. Separate a stated client reason from your interpretation. If the evidence doesn't establish why the client left, leave that uncertainty visible and assign the investigation.

The model calls for quarterly review of client margin and fit. Use that record to explain the loss where it identifies a relevant service or commercial issue. Retaining low-margin work and retaining profitable work are different operating outcomes even when the revenue percentage looks the same.

Start by reconciling a completed period from its opening clients to its closing recurring fees. Show the losses before the expansion, then calculate the percentages. Give a buyer that working sheet and the source trail rather than asking them to trust a retention headline.