How to measure sales conversion by cohort

01

Define the cohort and stages

Choose a cohort by the date a distinct opportunity entered a named stage, such as qualification. Define what qualifies an opportunity for each stage and how to handle reopened opportunities, duplicates and records that do not qualify. Record stage entry dates and the first confirmed sale date.

Set an observation window that fits the sales cycle. Keep opportunities still open at the cutoff visible, and compare cohorts only after the same window has elapsed. A stage count at one date divided by a different month's earlier-stage count does not track the same opportunities.

02

Calculate stage and overall rates

Suppose 100 distinct opportunities qualify in January. Within 90 days, 60 receive a proposal and 15 of those become customers. Qualification-to-proposal conversion is 60 ÷ 100 = 60%; proposal-to-sale conversion is 15 ÷ 60 = 25%; qualification-to-sale conversion is 15 ÷ 100 = 15%.

Count each opportunity once per stage under the stated rule. Show open and lost outcomes and their reasons alongside rates. If a sale can occur without a proposal, record that path explicitly rather than forcing it into the proposal denominator.

03

Investigate differences

Break down a mature cohort by source, offer or team only when the counts support a useful comparison. Inspect stage definitions and missing timestamps before attributing a difference to performance. Different deal sizes, customer needs and sales cycles may explain part of the gap.

Review a sample of lost and stalled opportunities with the team. Choose a specific change, record its start date and assess a later comparable cohort with the same definitions and window. A comparison alone cannot prove the change caused the result.

04

Check the records before comparing

Take a small sample from each stage and inspect the underlying dates and outcomes. Look for opportunities created twice, sales recorded before a stage entry and stage changes entered weeks after they happened. Correct confirmed errors or mark the affected records as uncertain.

If one team qualifies an opportunity at the first call while another waits for a budget discussion, their stage rates are not directly comparable. Agree a common definition or report the groups separately. Show the count behind each percentage, since one additional sale can substantially change a small cohort's rate.