Win rate = deals won ÷ deals decided, over the same period. The formula takes five seconds. The reason this article exists is the denominator — because "deals decided" has three defensible definitions, teams quietly switch between them, and a win rate computed one way cannot be compared to one computed another. Pick a definition, write it down, never change it mid-year.
The three denominators (pick one)
- Won ÷ (Won + Lost). The classic. Ignores everything still open, counts only resolved outcomes. Best for: mature pipelines where deals actually get closed out. Weakness: flattered by zombie deals that never get marked lost.
- Won ÷ all opportunities created. The honest one. Every opportunity you opened eventually counts against you, including the ones that just faded. Lower number, better decisions. Best for: teams serious about diagnosing their funnel.
- Won ÷ deals that reached a given stage. Stage-specific win rate ("of deals that got a proposal, how many closed?"). Not a replacement for the headline number — a diagnostic layer on top of it (see pipeline stages for the stage definitions this depends on).
Our recommendation: report the first, manage by the second, diagnose with the third. If you only track one, use won ÷ created — it's the only version that can't be gamed by simply never declaring defeat.
The worked example
Q2: 40 opportunities created, 10 won, 18 lost, 12 still open at quarter end.
- Won ÷ decided: 10 ÷ 28 = 35.7%
- Won ÷ created: 10 ÷ 40 = 25%
Both are "your win rate." They tell different stories — the 12 open deals are the gap, and what eventually happens to them is usually less flattering than the 35.7% implies. Count by deal count and separately by dollar value: winning 40% of deals but 20% of pipeline dollars means you're losing the big ones, and that's a different problem than a low rate.
What's a good win rate?
The honest answer: against your own last two quarters, trending up. Published benchmarks (typically 15–30% for B2B, by the won-÷-decided math) vary so much by industry, deal size, and pipeline discipline that borrowing them mostly produces false comfort or false alarm. The comparisons that matter: your rate by lead source, by rep, by deal size band, and over time. Those four cuts locate real problems; a benchmark from someone else's business locates nothing.
The four ways teams fool themselves
- Zombie inflation. Deals that die without being marked lost never hit the denominator. This is the big one — pair your win rate with a deal-aging rule (anything past 2× your normal cycle gets closed) or the number drifts up while reality drifts down. The mechanics are in our pipeline coverage guide.
- Qualification creep. Tighten what counts as an "opportunity" and win rate rises with no change in selling. Fine if deliberate — but log the definition change, or future-you will misread the trend as improvement.
- Sandbagging. Reps who only enter deals that are nearly certain show 80% win rates and empty pipelines. Symptom: high rate, low volume, forecasts that swing wildly.
- Period mismatch. Counting Q2's wins against Q2's created deals mixes cohorts when cycles run long. For cycles over a month, cohort the calculation: of deals created in a window, what fraction eventually won.
Making it move
Win rate improves at two moments, and neither is the closing call: qualification (letting fewer bad-fit deals in — see qualification frameworks) and the silent middle (follow-through on deals that stall — the discipline in our follow-up playbook). Teams that chase win rate with harder closing usually just get faster losses.
The measurement side is automatable: in AutomateNexus CRM the denominators stay honest mechanically — activity is logged automatically, aging deals get flagged before they zombify, and win rate is computed consistently across reps, sources, and time instead of in a spreadsheet with a flexible memory.
Win rate FAQ
What's the difference between win rate and close rate?
Usage varies, which is the problem. Most teams treat them as synonyms; some use close rate for won ÷ all leads (not just opportunities). Define both in writing, or better, use one term with an explicit formula.
Should lost-to-no-decision count as a loss?
Yes. "No decision" is the most common way deals die, and excluding it is how teams hide their real problem — deals that never should have entered the pipeline, or follow-up that stopped too early.
How often should win rate be reviewed?
Computed monthly, judged quarterly. Month-to-month it's noisy at small deal counts; a single quarter of decided deals is usually the smallest honest sample.
Does win rate matter more than deal size or cycle length?
They multiply together — that's literally the sales velocity formula. A rising win rate bought with smaller deals or longer cycles is a wash; check all three before celebrating.