Answer first: Net revenue retention is useful only when the cohort, revenue definition, currency treatment and observation window stay fixed. Pair the headline with a bridge from opening revenue through churn, contraction, price and expansion; then examine gross retention, customer concentration, contribution and the customer events behind movement so one large account cannot disguise broad weakness.

Key takeaways

  • Freeze the starting cohort and document every inclusion rule.
  • Separate churn, contraction, price, usage and cross-sell in a reconciled bridge.
  • Show gross retention and concentration alongside NRR.
  • Trace changes to product adoption and customer outcomes, not sales labels alone.

Decision framework

Decision area Evidence to require Risk it controls
Opening cohort Fixed customer/account population and baseline revenue Denominator drift
Churn Revenue lost from exited cohort members Hidden logo loss
Contraction Lower price, quantity or scope Weak adoption
Expansion Price, usage, seats and product separated Misread value
Closing cohort Reconciles to billing/revenue evidence Metric without audit trail

Net revenue retention is useful when it shows how revenue from a defined existing-customer cohort changes over a fixed period. It becomes misleading when expansion conceals customer losses, price increases conceal declining use or the cohort definition changes without explanation.

For a London B2B fintech, the practical question is not simply whether the percentage is above a fashionable benchmark. It is whether existing customers produce durable, diversified and economically attractive revenue.

Freeze the cohort and the revenue definition

Choose the customers present at the start of the period and exclude genuinely new customers from the closing figure. State whether the measure uses recognised revenue, recurring contract value or another basis. Do not silently mix them.

Define the treatment of acquired customers, foreign exchange, refunds, one-off work and customers with multiple legal entities. A group opening another account may represent expansion rather than a new customer, depending on the documented methodology.

Public companies publish their own definitions. Snowflake's annual report provides one example of a consumption business explaining its retention measure. 1 Comparisons require reading the calculation, not copying the label.

Build the revenue bridge

Begin with opening cohort revenue. Subtract churn and contraction, then add expansion on the same measurement basis. Show price changes separately where possible.

In a hypothetical cohort starting at £1 million, £150,000 of churn and £100,000 of contraction leave £750,000. Expansion of £400,000 takes the closing amount to £1.15 million: 115% net revenue retention. The attractive final percentage coexists with 25% of opening revenue lost before expansion.

That does not make the business weak automatically. It identifies the next questions: which customers expanded, why others left and whether the pattern can continue.

Look at gross retention and concentration

Gross revenue retention removes the benefit of expansion, helping reveal the underlying leakage. Logo retention shows how many customers remain. Neither should replace net retention; together they explain it.

Calculate how much expansion came from the largest accounts. A company can report strong net retention while becoming dependent on one customer's budget and implementation choices. Show a sensitivity excluding the largest expansion event, clearly labelled rather than presented as the official metric.

Segment customers by product, acquisition channel, size and tenure. Avoid tiny groups that produce unstable percentages. Counts and absolute revenue should accompany the rates.

Distinguish price from value

A contractual increase can raise revenue while customer activity falls. Examine usage, adoption of important workflows, support issues and renewal discussions. Customers may tolerate a price increase once before deciding to leave.

For consumption products, lower usage can mean successful optimisation rather than dissatisfaction. Ask whether the customer achieves the same outcome more efficiently.

Charging less for a cheaper-to-deliver task can preserve contribution and improve the relationship.

Recognised revenue also depends on the applicable accounting framework and contract facts. FRC materials explain UK accounting standards, including FRS 102. 2 A retention dashboard is a management measure and should reconcile coherently to financial reporting without pretending the two are identical.

Add contribution and cash

Expansion can require more cloud processing, implementation or account management.

Show contribution by cohort where costs can be attributed reasonably. Explain allocation limits rather than creating false precision.

Track collection as well. A large invoice that remains unpaid is different from cash received.

Separate disputed balances, credits and amounts financed through unusual terms.

An illustrative account expanding from £50,000 to £80,000 of annual revenue may contribute less if service costs rise from £20,000 to £60,000. Revenue grew 60%; contribution fell from £30,000 to £20,000. The sales story and the economic story differ.

Use a constant-currency view without hiding the reported result

When customers pay in several currencies, exchange movements can change the retention figure without changing their underlying spend. A supplementary constant-currency calculation can help explain that effect.

State the method and present it beside the reported measure. Do not choose whichever conversion basis makes growth look strongest in each period. Keep the calculation reproducible and apply it consistently.

For a hypothetical dollar account with unchanged local-currency billing, a stronger dollar can raise sterling revenue. That is useful cash information, but it is different from winning more work within the account. The board needs both explanations.

Trace expansion to a customer event

Classify growth as more users, more activity, another product, a higher price or another documented change. Review the largest events with account teams and finance.

A one-off migration project may explain a large invoice without demonstrating recurring expansion. A customer buying a second product may be strategically valuable but also require another implementation. Connect the event to its expected duration and cost.

Review contraction with the same care. Reduced use after a successful efficiency improvement can differ from a customer withdrawing because the service failed. The data should help management ask the right question rather than impose one interpretation automatically.

This event-level view turns retention from a fundraising statistic into a commercial diagnostic.

It reveals which forms of growth are repeatable and which depend on unusual circumstances that the forecast should not extrapolate casually.

What belongs in the board pack?

Show the cohort definition, revenue bridge, gross retention, logo retention, concentration and contribution evidence. Add a brief explanation of the largest changes and what management will do about them.

Keep the retention percentage, but bring the revenue bridge to the meeting. Churn, concentration, pricing and contribution explain which customer relationships deserve investment and which assumptions the forecast should stop repeating.

A practical 90-day implementation plan

  1. Write the cohort, revenue, FX, merger and reactivation rules in a metric dictionary.
  2. Build an account-level bridge that reconciles opening and closing cohort revenue.
  3. Add gross retention, top-account concentration, contribution and cash collection views.
  4. Review the largest movers with customer-success and product evidence, then record actions and owners.

Frequently asked questions

What is a good NRR?

There is no universal threshold. Interpret the metric by business model, cohort maturity, pricing, concentration, gross retention and contribution, using a consistently calculated history.

Should price increases count as expansion?

They can appear in the bridge, but show price separately from usage, seats and cross-sell so readers can distinguish pricing power from deeper adoption.

Why show gross retention too?

Strong expansion from a few accounts can lift NRR while many customers shrink or leave. Gross retention exposes that underlying loss.

Reporting by Shoreditch Talk