What is Net Revenue Retention?

Most growth plans are written about clients the firm has not met yet. Meanwhile the clients it already serves are quietly deciding the year: renewing or drifting, expanding or trimming, recommending or merely tolerating. Net revenue retention is the number that watches this, and in a people business it is usually the most honest growth number in the building.

What net revenue retention measures

Take the clients you served at the start of a period, typically a year, and follow only them. Ignore every client won during the period. At the end, compare the revenue that same cohort now generates with the revenue it generated at the start. Expansion within the cohort pushes the figure up; churned clients and shrunken engagements pull it down. The result, expressed as a percentage of the starting revenue, is net revenue retention.

The arithmetic is simple, and every number here is invented for illustration. A cohort bills ten million at the start of a year. One client worth a million leaves. Another trims its scope by half a million. Expansion across the remaining accounts adds two million. The cohort ends the year at ten and a half million, and NRR is 105 percent. Above 100, the existing base grows before a single new logo is signed. Below 100, new business must fill a hole before it can add an inch of growth.

The compounding engine of a people business

In software, expansion often means more licences. In a people business it means something stronger: a client who has watched a team deliver chooses to hand it another problem. Expansion revenue in services is bought with delivered work, not with campaigns, which is why it arrives without the acquisition cost a new logo carries, and why it tends to be won at better rates than a competitive pursuit. The deliberate version of this motion is land and expand.

Net revenue retention is the compounding engine of a people business, because it measures whether the delivered work itself is selling the next engagement. A firm above 100 percent starts each year larger than the last before its pipeline contributes anything, and run over several years that compounds the way interest does. A firm below 100 percent is renting its revenue: sales must keep running just to hold the top line still. And because churn in services is rarely sudden, a decaying NRR usually announces a problem in delivery quality or cost-to-serve pricing long before it reaches the annual accounts.

Cohort honesty

NRR only tells the truth if the cohort is honest, and there are well-worn ways to flatter it:

• Blending in the new. Folding mid-year wins into the base inflates the figure with revenue that retention did nothing to earn. The cohort must be fixed at the start and closed to newcomers.
• Netting that hides the violence. A placid 105 percent can conceal a large slice of the base churning while two accounts expand furiously. Gross retention belongs beside the net figure, always.
• Bookings instead of revenue. Signed expansion is intent; recognised revenue is fact. Counting intent brings the number forward and the disappointment later.
• The survivor’s cohort. Quoting the retention of the accounts that stayed is a tautology. The clients who left are the point of the measurement.

The remedy is boring and absolute: one fixed cohort, one constant definition, recognised revenue only, gross and net reported side by side.

One concrete example

Clearly illustrative, with no customer implied. A consultancy of a few hundred people reports NRR of 104 percent and toasts a compounding year. Decomposed, the number tells two opposite stories at once. The expansion is concentrated in two accounts served by one strong practice, and the mid-tier of the base is churning faster than the year before, mostly accounts that were staffed thin during a busy spring. The single blended figure hid both facts. The two expanding accounts point at what the firm should sell more of; the churning mid-tier points at a staffing decision that is quietly taxing next year’s revenue. Neither appears in a pipeline review, because neither is a pipeline event.

The decisions inside the number

NRR is a lagging summary of many small choices: which accounts got the strongest team, which requests were absorbed to keep a relationship warm, which renewal was defended a quarter early and which was left to arrive as a surprise, which price was quietly conceded. By the time the number moves, those decisions are months old and unrecorded, which is why so many retention reviews end in theories rather than causes.

The decision-intelligence reading is that retention is decided upstream, in moments too small to feel like decisions at the time. A firm that records those choices (who decided, against what evidence, expecting what) can trace a moving NRR back to specific decisions, and learns which kinds of staffing, pricing and generosity actually compound. Revenue that disappears without a decision ever being taken has its own name: revenue leakage. Net revenue retention is where both stories are eventually netted together, whether or not anyone recorded them.