Logo retention

Logo retention is the percentage of customers (accounts, “logos”) that are still customers at the end of a period, ignoring how much each one pays. It counts heads, not dollars. If you start the quarter with 200 accounts and 184 are still active at the end — before any new sales — your logo retention is 92%.

Logo retention is not revenue retention. GRR and NRR weight every account by its ARR; logo retention weights every account equally. A 5-seat startup churning counts exactly as much as your largest enterprise account churning. That difference is the entire point of tracking both — and the entire reason logo retention can lie if you read it alone.

The formula

Logo retention = (Accounts at start − Logos churned during period) / Accounts at start
Logo churn rate = Logos churned during period / Accounts at start

Count only logos you had at the start of the period. New accounts acquired mid-period are excluded from both numerator and denominator — otherwise new sales mask churn. A team that began the quarter with 200 accounts, lost 16 to cancellation, and signed 30 new ones reports:

• Logo retention = (200 − 16) / 200 = 92%
• Logo churn = 16 / 200 = 8% for the quarter

Do not net expansion or new logos into this number. Logo retention has no “expansion” analog the way NRR does — an account is retained (1) or churned (0). A “net logo” figure that adds new customers is a growth metric, not a retention metric; keep them separate.

Count-based vs revenue-based: when each tells the truth

The two views diverge most when your revenue is concentrated. Picture 100 accounts: 90 small ones at $5K ARR and 10 enterprise ones at $200K ARR. Total ARR is $2.45M, and 82% of it sits in 10 logos.

• If you lose 10 small accounts, logo retention drops to 90% but GRR barely moves (you lost $50K of $2.45M — under 98% retained on dollars). The headcount metric screams; the revenue metric shrugs.
• If you lose 1 enterprise account, logo retention is 99% — looks pristine — but GRR craters to roughly 92% on that one loss. The headcount metric shrugs; the revenue metric screams.

That is the rule: logo retention tells the truth in low-concentration, high-volume books (PLG, SMB, transactional) where every account is roughly the same size and you need an early signal that the product or onboarding is failing at scale. Revenue retention tells the truth in concentrated, enterprise books where a handful of logos carry the number and one churn reshapes the year. Read either one alone in the wrong book and you will optimize for the wrong thing.

The honest dashboard shows both side by side, and the gap between them is itself a signal. Logo retention materially above GRR means your churn is biased toward small accounts (often fine, sometimes a packaging problem at the low end). Logo retention materially below GRR means you are losing many small logos while keeping the big ones — a self-serve or onboarding problem that hasn’t hit revenue yet but will once it reaches a larger cohort.

What’s a good rate

There is no universal number because it tracks with segment and contract length, but rough B2B SaaS bands:

Monthly SMB tools live lower; multi-year enterprise contracts live higher simply because there are fewer renewal events to fail. Always state the period — quarterly logo churn of 3% is roughly 12% annually, and the two numbers describe very different health if you confuse them.

Common pitfalls

• Reading logo retention alone in a concentrated book. A 96% logo rate hides the enterprise loss that wrecked GRR. Guard: never report logo retention without GRR next to it; the pair is the metric, not either one.
• Mid-period acquisitions in the denominator. Folding new logos into the base inflates retention. Guard: freeze the cohort at period start; new accounts enter next period’s base, not this one.
• Ambiguous “churn” timing. Is a logo churned on the cancellation date, the contract-end date, or when the last seat deactivates? Guard: pick contract-end date, document it, and apply it consistently — most tools default to it.
• Mixing periods. Comparing this quarter’s logo retention to last year’s annual figure is meaningless. Guard: annualize or compare like-for-like periods only.

Related

• NRR vs GRR — the revenue-weighted view to read alongside logo retention
• Churn rate calculation — logo churn is the inverse of logo retention
• Expansion revenue — what logo retention deliberately ignores
• Customer health score — the leading indicator before a logo churns