How to compute win rate from closed outcomes rather than open pipeline, why the team's single figure hides a 45 percent rate in one segment and a 12 percent rate in another, the three splits that matter, segment, source and deal size band, the minimum count that makes a split trustworthy, and how the split changes where the pipeline effort goes.
A sales team reports a win rate and it is one number. The closed outcomes say it is four numbers with a wide spread, and the spread is the finding. This guide sets out win rate from outcomes, the three splits, the minimum count, and what the split changes.
Per period:
Win rate = won ÷ (won + lost), among opportunities with an outcome in the period With no-decisions: won ÷ (won + lost + no decision)
Per cell: segment, source, deal size band, and any two of them together.
Opportunity identifiers only.
outcomes in the period = won + lost + no decision
Every outcome in one class. An opportunity closed with no outcome recorded fails it and is listed.
Team win rate: 28 percent, 410 outcomes.
| Segment | Outcomes | Won | Win rate | No-decision share |
|---|---|---|---|---|
| Mid manufacturers | 140 | 63 | 45% | 8% |
| Enterprise | 90 | 11 | 12% | 31% |
| Small services | 180 | 41 | 23% | 22% |
| Source | Outcomes | Win rate |
|---|---|---|
| Referral | 60 | 52% |
| Inbound | 190 | 29% |
| Outbound | 160 | 18% |
| Deal size band | Outcomes | Win rate |
|---|---|---|
| Under $20,000 | 210 | 34% |
| $20,000 to $100,000 | 150 | 24% |
| Over $100,000 | 50 | 14% |
The enterprise segment wins one in eight and a third of its outcomes are no decision, which is a qualification problem. Referrals win at twice the inbound rate and the team generates the fewest of them. None of that is in 28 percent.
| Segment × source | Outcomes | Win rate |
|---|---|---|
| Mid manufacturers × referral | 24 | 63% |
| Mid manufacturers × outbound | 48 | 38% |
| Enterprise × outbound | 41 | 7% |
| Enterprise × referral | 9 | greyed: under 20 |
Enterprise outbound wins one in fourteen. That is a cell the pipeline should stop being filled from, and the cell says so with a count behind it.
Win rate from open pipeline. The CRM's probabilities, averaged.
One number. Twenty-eight percent, describing no segment.
Small cells read as rates. Nine outcomes, four wins, "44 percent".
No-decisions in lost. A qualification problem read as a competitive one.
Mapped once, the outcomes and the bands produce win rate by segment, source and size, the pairs, and the counts behind each cell every quarter. Covirage builds this from the export as it is. The sales insights solution describes the setup, and the sales velocity guide covers the formula win rate sits inside.
They are reported separately: won, lost to a competitor, and no decision. A segment with a high no-decision rate has a qualification problem, not a competitive one, and blending it into lost hides that. Win rate is shown both ways, with and without no-decisions, and labelled.
Twenty outcomes per cell is a common floor; below it the rate is shown with its count and greyed, not used for decisions. A cell with eight outcomes at 50 percent is four wins, and the next four could be losses.
To decide where to qualify harder and where to invest. A segment at 12 percent with the same cycle length as one at 45 costs almost four times as much per win. The split turns that into a rule for what gets into the pipeline.