Sales management
Compare sales win rates without mixing closed periods and created cohorts
Define eligible deals, observation windows and unresolved outcomes before AI explains a win-rate change.
A closed-period win rate and a created-cohort outcome rate answer different questions.
Editorial workflow guide · AI-assisted preparation · Not a hands-on product test. How we work
Name the population and the observation date
Ask the manager which decision the rate supports. A closed-period view asks what share of completed decisions in a period were wins. A created-cohort view follows opportunities opened in a particular period, some of which may still be unresolved. Define the exact window, pipeline, eligible stages and snapshot date before asking AI for a comparison.
Keep a record ID list behind every numerator and denominator. Specify how reopened deals, duplicate opportunities, no-decision outcomes and migrated records are handled under your reporting policy. AI should expose mismatched definitions rather than choose whichever denominator makes the trend look strongest.
Sponsor
SPONSORED
AI FOR YOURSales team.Less busywork.More selling.
Your team knows sales.
We help put AI to work.
Work two legitimate but different rates
In a fictional month, six deals close won and four close lost. The closed-period win rate is 6 divided by 10, or 60%. Some may have been created in earlier months. It would be misleading to label that result as the eventual success rate of opportunities created this month.
Suppose a separate created cohort contains 20 deals: four won, six lost and ten still open at the review date. Four of twenty have won so far, or 20%; among the ten resolved deals, four of ten is 40%. Neither figure establishes the final cohort result. Show the ten open deals and observation date beside the rate.
Validate formulas against your own stage mapping
HubSpot documents dataset summary functions and provides a win-rate example using distinct deal IDs and closed-status conditions. Its illustrated stage label is an example; copying it without checking your own pipeline can misclassify outcomes. Dataset access and formula behavior also depend on the actual account configuration.
Compare a small approved ID-level sample with the report output. Check date filters, whether an association join duplicates rows and whether a reopened opportunity still qualifies as closed. A distinct count can address repeated IDs but cannot repair the wrong population or an incorrectly mapped winning stage.
Write a bounded interpretation
Return the definition, date window, numerator, denominator, unresolved count, exclusions and a short reconciliation of any difference from the prior report. Separate a real outcome change from a change in filtering or stage definitions. Include counts beside percentages so a single extra win does not disappear inside a dramatic headline.
Rehearse with an empty denominator, a newly created open cohort and a reopened deal. Do not convert an undefined rate into zero. This is an editorial reporting method, not a calibrated prediction of future wins or evidence that an AI tool improved selling. Investigate the underlying deals before assigning causes to a percentage change.
A prompt to adapt
Prepare closed-period and created-cohort win-rate definitions from these approved deals. Show eligible IDs, windows, wins, losses, open outcomes and exclusions. Calculate each denominator explicitly; flag zero denominators and reopened records. Do not mix populations, infer causes or present incomplete cohorts as final outcomes.
Use fictional or anonymized examples while testing. Add your approved business rules before using a draft.Source documents
Primary sources support the references above. These workflows are editorial suggestions; confirm current product availability, pricing, permissions, and integration details directly.
Questions or corrections?
See our editorial policy for how we distinguish workflow suggestions, vendor statements, and tested results.