Win rate: the four formulas and why they disagree

Win rate is the share of opportunities that end in a win. The numerator is uncontroversial — deals marked closed-won. The denominator is a choice, and four defensible choices are in common use. They produce different numbers from identical data, which is why two people in the same forecast meeting can quote win rates twelve points apart and both be right. Before comparing a win rate to anything — a target, last quarter, a benchmark — establish which denominator produced it. Related pipeline terms are in the glossary.

The four formulas

Worked example. One hundred opportunities were created in Q1. Measured today: 40 closed-won, 45 closed-lost, 15 still open. In the same period 30 deals created earlier also resolved, of which 12 were won. Marketing passed 600 leads that became those 100 opportunities.

Same data, four answers between 6.7% and 47%. Three points follow.

The resolved-only figure is the highest, always, because it ignores the 15 open deals. Some of those will lose. If ten of the fifteen eventually lose, the cohort's true final win rate is 45 ÷ 100 = 45%, not 47%.

The cohort figure is the right one for pipeline maths, because pipeline coverage targets assume every created opportunity sits in the denominator. Using a resolved-only rate to set coverage systematically under-builds pipeline.

The lead-to-win figure is not a sales win rate at all. It measures lead quality as much as selling. Quoting it as a win rate makes a competent team look broken.

One more distortion worth naming: any rate calculated on an unresolved cohort is provisional. Q1 win rate quoted in April is a partial number. Only quote a cohort rate once at least 90% of the cohort has resolved, which for most teams means one to two sales cycles after the cohort closed.

Stage-level win rates

Overall win rate tells you the process leaks. Stage-level conversion tells you where. Calculate the share of opportunities entering each stage that reach the next one, using a cohort old enough to have resolved.

Multiply the four: 0.78 × 0.65 × 0.86 × 0.91 = 0.396, or 40% — the cohort win rate, reconstructed. The weak transition is discovery to proposal at 65%, the largest single leak in the process at 27 lost deals. That is where coaching and, if you are scoring activity, where the points should go. Whatever behaviour separates the deals that clear that stage is the leading indicator worth weighting.

How win rate gets misused on leaderboards

Win rate is a poor ranking metric for three structural reasons, and it is one of the easiest metrics on a board to manipulate.

Small denominators. A rep with four resolved deals and three wins ranks at 75% and tops the board. A rep with 22 resolved deals and 13 wins ranks at 59% and produced four times the revenue. Any win-rate board needs a minimum-denominator floor — typically ten resolved deals in the period — or it ranks noise.

Denominator suppression. The fastest route to a high win rate is to create fewer opportunities. Rank on win rate alone and reps stop logging marginal deals, pipeline shrinks, and the board reports improvement while the business gets worse.

Loss avoidance. Deals that should be marked closed-lost get moved to "on hold" or pushed out indefinitely so they never enter the denominator. This is invisible on the board and corrosive to the forecast. The guardrail is an ageing rule: any opportunity untouched past a threshold auto-resolves as lost for scoring purposes.

Blueprint's gaming-risk audit scores exactly these — every metric is rated for how easily it can be padded, and the guardrail ships with the metric rather than being discovered three months in. Win rate belongs on a season-long secondary board with a denominator floor, alongside a primary live board scoring the activity that drives the weakest stage transition. The board design rules go further, and there is a 14-day trial, no card required.

FAQ

How is sales win rate calculated?

Wins divided by a denominator you have to choose. The four common denominators are resolved deals only (won ÷ won + lost), all opportunities in a created cohort, all deals resolved in a reporting period, and all leads. On 100 opportunities with 40 won, 45 lost and 15 open, those give 47%, 40%, a period-dependent figure and 6.7% respectively. State the denominator whenever you quote the number.

Why do two win rates from the same data disagree?

Because they use different denominators. A resolved-only rate excludes open deals, so it is always the highest and always provisional — some of those open deals will lose. A cohort rate includes everything created, which is what pipeline coverage maths assumes. A lead-to-win rate includes unqualified leads and measures lead quality more than selling. The gap between them can exceed ten points.

What is a stage-level win rate?

The share of opportunities entering a stage that advance to the next one, calculated on a cohort old enough to have resolved. Multiplying the stage conversions reconstructs the overall win rate: 0.78 × 0.65 × 0.86 × 0.91 gives 40%. Stage-level rates identify where the process leaks. Overall win rate only tells you that it does.

Should win rate be on a sales leaderboard?

Only as a season-long secondary board with a minimum-denominator floor of around ten resolved deals. Without a floor, a rep with three wins from four deals outranks a rep with thirteen from twenty-two. Win rate also invites denominator suppression — the fastest way to raise it is to log fewer opportunities, which shrinks pipeline while the board reports improvement.

How do reps game a win-rate metric?

Two ways. They stop creating marginal opportunities, which shrinks the denominator and raises the rate while reducing real pipeline. And they avoid marking losses, parking dead deals in "on hold" or pushing close dates out so those deals never resolve. The guardrail for the second is an ageing rule that auto-resolves untouched opportunities as lost for scoring purposes.