How to Calculate Win Rate by Sales Stage

A single blended win rate tells you how often you win. Stage win rates tell you where you lose, and they are what makes weighted pipeline worth anything. This guide covers the two stage metrics that matter, the cohort method that keeps them honest, and how to apply the output to open pipeline.

What are the two stage win rate formulas?

Stage-to-stage conversion measures the next step, stage-to-close win rate measures the final outcome.

``` Stage-to-Stage Conversion = Opportunities Advancing to Next Stage / Opportunities Entering the Stage x 100

Stage-to-Close Win Rate = Opportunities from Stage That Closed Won / Opportunities Entering the Stage x 100 ```

They answer different questions. Stage-to-stage conversion is a process diagnostic. If 62 percent of discovery opportunities reach demo, the 38 percent that do not are a qualification or discovery problem you can work on directly.

Stage-to-close win rate is a pricing input. If opportunities entering demo close won 31 percent of the time historically, an open demo-stage deal is worth 31 percent of its value in expectation. That is the number that should drive weighted pipeline, rather than the default probabilities most CRMs ship with.

How do you build the calculation without the snapshot error?

Use a resolved cohort of opportunities, never a snapshot of current pipeline.

The common mistake is counting the deals in each stage right now and dividing. That produces rates distorted by where open deals happen to be sitting, and stages with long dwell times look worse than stages that move fast for reasons that have nothing to do with conversion.

The cohort method:

1. Pick a creation window far enough back that most deals have reached a terminal outcome. At least two median sales cycles. 2. Pull every opportunity created in that window, including ones that never advanced past stage one. 3. Record the furthest stage each opportunity reached, not the stage it sits in now. 4. Record the terminal outcome: closed won, closed lost, or still open. 5. Exclude nothing, and report the still-open share so readers know how resolved the cohort is.

Read the stage-to-stage column first. The two 60 percent steps, discovery to demo and demo to proposal, are where this funnel leaks hardest. The stage-to-close column is what you apply to open deals sitting in each stage.

How do you handle deals that skip or move backward?

Record the furthest stage reached, and count backward moves as a signal rather than a correction.

Skipped stages are common and mostly harmless if you count entry into a stage as "reached at least this stage." A deal that jumps from discovery to proposal counts as having entered demo for the purpose of the funnel, otherwise your demo stage denominator shrinks and its win rate inflates.

Backward movement is different. A deal pushed from negotiation back to discovery is rarely recovering. Track those separately, because they contaminate the negotiation-stage win rate if you leave them in the numerator population without noting they left.

The strongest deterioration signal is close date movement. When a rep changes the close date, the deal becomes less likely to close, even if it is sitting in commit. The earliest signal is quieter: no stage change, no close date change, no amount change, no activity. Meaningful activity means a change in stage, close date, or amount, and its absence is data.

How should stage win rates change your coverage math?

They convert a raw coverage ratio into an expected value, which usually looks much worse.

Most teams run on a 3x to 5x coverage rule. Across ORM's customers that range holds as a standard, with some accounts as low as 1.4x, some at 5x, and most around 3.5x. The ratio by itself hides composition, and composition is what stage win rates expose.

Two numbers make the point. More than 10 percent of open pipeline is typically stale, untouched for twelve months. And across ORM's customer base, of the pipeline carrying a close date inside the quarter on the first day of that quarter, roughly 20 percent actually closes in the quarter, meaning about 80 percent of the value visibly sitting in the quarter will not be realized in it.

Apply stage-to-close rates to that same pipeline and the picture sharpens. A $12,000,000 pipeline against a $3,000,000 quota is 4x raw coverage. If most of that value sits in qualification at a 14 percent stage-to-close rate rather than in proposal at 60 percent, the expected value is nowhere near the target. Coverage is an input, and treating it as the answer is where forecasts go wrong.

How often should stage win rates be recalculated?

On a rolling basis, because the conditions that produced them keep changing.

Forecasts miss most often when the business or market shifts and the model is still running on old assumptions. A new competitor creates pricing pressure and average deal size falls. Rates rise, buyers slow down, and win rates drop. Uncertainty stretches the time from qualified to closed. A territory realignment leaves coverage intact while execution suffers.

Each of those changes moves stage win rates before it moves the annual number. Refresh the rates on a rolling window rather than annually, and watch the direction of movement as much as the level.

For metric definitions, see win rate and pipeline coverage. For why the coverage rule fails on its own, see the 3x pipeline coverage rule is wrong.