Can gross retention ever exceed its starting point?

No. Gross retention excludes expansion entirely, so it is bounded by the cohort's starting recurring revenue. Only net retention, which adds expansion, can exceed the starting figure.

What does a large gap between the two measures indicate?

The gap equals expansion on that cohort. A large gap alongside weak gross retention means growth is coming from a small number of accounts while the broader base is leaking, which is a concentration risk rather than a sign of health.

Which measure should a customer success team be held to?

Gross retention is the fairer measure when the team does not control pricing or expansion, because it reflects whether customers stay and keep what they bought. Net retention is appropriate when the team carries an expansion target.

Gross Retention vs Net Retention: Why Strong NRR Can Hide a Shrinking Customer Base

Comparison · customer-success · 4 min read · last verified 2026-07-21

Reviewed before publication Editorial board Independent commercial review

In shortGross retention excludes expansion and cannot exceed its starting point; net retention includes it and can exceed it. The gap between them shows whether growth is masking a shrinking base.

Gross retention measures how much recurring revenue an existing customer base keeps, excluding all expansion, so it can never exceed its starting point; net retention measures the same base including expansion, so it can exceed its starting point and can conceal a shrinking base underneath growth from a few large accounts. The gap between the two figures is the whole story, and either number reported alone is close to uninterpretable.

Gross retention vs net retention at a glance

What is counted. Gross retention counts churn and downgrades only. Net retention counts churn, downgrades, and expansion.
What is excluded. Both exclude revenue from customers acquired during the period, so both describe the existing base rather than total growth.
Ceiling. Gross retention has a hard ceiling at its starting value. Net retention has none.
What it answers. Gross retention answers whether customers stay and keep what they bought. Net retention answers whether the base is worth more over time.
Primary distortion. Gross retention is distorted by scope reductions being recorded as renewals. Net retention is distorted by expansion concentrated in a small number of accounts.
Who relies on it. Product and customer success read gross retention as a fit signal. Finance and investors read net retention as a capital-efficiency signal.

What gross retention is

Gross retention takes the recurring revenue of a defined cohort at the start of a period, subtracts revenue lost to cancellations and to downgrades within that cohort, and divides by the starting figure. Expansion is deliberately omitted, and new customers are excluded because the measure describes the durability of what already existed.

Because expansion cannot offset losses, gross retention is the closer proxy for product fit and delivery quality. It is also the harder number to improve, since the only routes are fewer cancellations and fewer reductions.

The main way gross retention is overstated is classification. A customer that renews at reduced scope is still a renewal, and if the reduction is recorded as a contract change rather than a downgrade, the loss disappears from the measure. Separating revenue loss from customer loss is the subject of dollar churn vs logo churn, and the same classification discipline is what makes gross retention trustworthy.

What net retention is

Net retention uses the same cohort and the same exclusions, then adds expansion — upsell, cross-sell, seat growth, usage growth, and price escalation on existing scope. A value above the starting point means the base grew without any new customers, which is why the measure is treated as a proxy for capital efficiency. The full definition and its variations are covered in net revenue retention.

Two structural properties make it easy to misread. First, expansion and contraction are netted, so equal amounts of each cancel out and the measure reports stability where two opposite forces are actually at work. Second, expansion is usually far more concentrated than churn: a handful of accounts can supply most of it while losses are spread thinly across many small ones.

How they relate

The arithmetic relationship is direct — net retention equals gross retention plus expansion, on the same cohort and period. The interpretive relationship is where the value sits.

Consider a hypothetical cohort that starts a period at 100 units of recurring revenue. If 15 units are lost to cancellation and downgrade while 20 units of expansion arrive from two accounts, net retention reads 105 and gross retention reads 85. The base is being kept alive by two customers, and the report says the base grew.

That structure produces several practical consequences:

A wide gap with low gross retention is a concentration warning. The healthy-looking number depends on accounts whose own departure would remove both the expansion and its source.
A narrow gap with high gross retention describes a durable but static base. Customers stay and do not grow, which limits efficiency but is not fragile.
Netting hides direction. Reporting expansion and contraction as separate lines alongside both retention figures restores the information that netting removes.
Cohort choice changes the answer. Segment-level gross retention frequently reveals that a company-wide net retention figure is the average of one strong segment and one failing one.

Which to use when

Use gross retention to judge product fit, implementation quality, and support performance. It is the right measure for questions about whether the product holds its position once bought, and the right measure to hold customer success accountable to when that team does not control pricing or expansion.
Use net retention to judge the economics of the installed base and the return on acquisition spend. A base that grows without new customers changes the payback profile of every acquired customer, which is why the figure is read alongside customer acquisition cost.
Use both, always, on the same cohort. Neither number is meaningful without the other. Diligence processes ask for both plus the expansion and contraction components precisely because the netted figure can be assembled from very different underlying realities.
Watch the composition when either is used as a target. Compensating a team on net retention alone rewards concentration in large accounts and makes small-account losses invisible, since those losses barely move a number that a single large expansion can lift.