How to Calculate Stage Conversion Rates in Your Sales Pipeline

What is a stage conversion rate?

A stage conversion rate is the share of opportunities entering a stage that advance past it.

``` Stage conversion rate = Deals that advanced past the stage / Deals that entered the stage ```

If 240 opportunities entered proposal during your measurement window and 91 moved forward to negotiation, the proposal conversion rate is 38%. The remaining 149 were lost, closed out, or are still sitting in the stage unresolved.

That last group is where most calculations break. A deal still open in proposal has not converted and has not failed to convert, so counting it as a failure understates the rate and excluding it from the denominator overstates it. The clean method is to measure a cohort old enough that nearly all of it has resolved, then exclude the small remainder from both sides of the division.

How do you build the cohort correctly?

Fix an entry window, follow those specific deals forward, and measure their outcomes rather than counting what currently sits in each stage.

Pick a window that closed at least two sales cycles ago. If your median cycle is 70 days, a cohort of deals that entered the funnel five to eight months back will be nearly fully resolved. Tag every opportunity that entered your first stage inside that window, then trace each one to its furthest stage and final outcome.

The alternative approach, which is what most CRM reports produce by default, is a snapshot. It counts how many deals are in each stage right now and divides one stage by the previous one. That method is wrong in a specific direction. Deals that entered a stage and then died are no longer in the funnel at all, so they never appear in the denominator. Snapshot ratios are consistently better than reality, which is exactly the wrong kind of error to have in a forecast input.

What does the full stage conversion table look like?

Report entry count, advance count, single-stage conversion, and cumulative conversion to close in one view. The cumulative column is calculated backward from the final outcome. Of the 1,000 deals that entered discovery, 116 closed won, which is 11.6%. Of the 310 that reached proposal, 116 closed won, which is 37.4%. Those cumulative figures are the numbers you use to weight pipeline, because they account for every remaining step rather than a single hop.

Notice that discovery converts at 52% while the three later stages sit near 60%. That gap is normal and it is also where the largest absolute loss occurs. Improving discovery conversion by five points adds 50 deals to the funnel. Improving negotiation by five points adds nine.

How do you turn conversion rates into forecast weights?

Multiply the open pipeline in each stage by that stage's cumulative-to-close rate, then sum the results.

A pipeline with $2,000,000 in proposal at 37.4% contributes $748,000. The same $2,000,000 sitting in discovery at 11.6% contributes $232,000. Same dollars, different quarter.

Two cautions apply. First, use your own historical rates rather than the default probability percentages that ship with a CRM, since those are configuration defaults and not measurements of your business. Second, weighted pipeline is a directional estimate rather than a forecast, because it assumes every open deal behaves like the historical average and ignores deal-level signals such as close-date movement or an absence of buyer activity. The mechanics and the limits are covered in weighted pipeline.

Why do conversion rates differ so much by segment?

Because segment, source, and deal size each change the shape of the funnel, and a blended rate describes none of them.

Enterprise deals typically show lower discovery conversion and higher late-stage conversion, since more evaluation happens before a deal advances and fewer deals die once procurement is engaged. SMB motions often show the reverse pattern. Blending them produces a middle number that misprices both.

Run the cohort analysis separately for each segment, each primary lead source, and each deal size band. The volume requirement is real: a segment with 20 deals per quarter will not produce a stable rate, so widen the window for the smaller segments rather than accepting a noisy percentage. Conversion at the final step is the same calculation as your win rate, measured from the last stage instead of the first.

What does a sharp drop at one stage actually mean?

It usually indicates a stage definition problem rather than a selling problem.

When deals consistently die between two stages, the first thing to check is the entry criteria for the earlier stage. Loose criteria let unqualified deals in, and those deals have to die somewhere. The drop shows up at the next gate, which is why it gets misread as a coaching issue at that gate.

The second thing to check is whether the exit criteria are objective. A stage that advances on rep judgment produces inconsistent data, and the conversion rate calculated from that data will vary by manager rather than by market. Criteria tied to observable buyer actions, such as a completed technical validation or an economic buyer meeting, produce rates you can plan against.

Only after both checks come back clean is the drop a performance signal worth coaching.

How does stage conversion change over time?

Conversion rates move when market conditions move, and they move without any change in your process.

When a new competitor enters and creates pricing pressure, deals take longer and more of them stall in evaluation. When capital conditions tighten and buyers cut costs, win rates fall across every stage at once. Neither event produces a CRM field that says what happened. Both produce a conversion table that looks worse than last quarter's.

That is why a conversion rate calculated once and used for a year is a liability. A model built on assumptions that no longer hold will keep producing confident output while the business changes underneath it, which is the most common reason a forecast misses. Recalculate the cohort every quarter, watch which stage moved, and treat a sustained shift as information about the market rather than an execution failure. The broader practice is covered in sales forecasting best practices.