Calculate win probability based on pipeline stage in HubSpot

Summary

Sales forecasts in HubSpot often rely on static deal stage probabilities that are set once and rarely revisited. These probabilities are typically based on intuition rather than historical performance, which leads to inaccurate forecasts and limited visibility into where deals break down across the pipeline.

In this use case, Tory Ferrall shows how her team used HubSpot and Databox to calculate real win probabilities by pipeline stage using historical deal data. By tying deals to the date they entered each stage and calculating win rates based on Closed-Won versus all Closed deals, the team gained a clearer view of pipeline health, rep performance, and stage-level drop-offs. The result is more accurate forecasting and more precise coaching conversations.

Tory’s tips and best practices

“Static probabilities are usually set once and never revisited.”

Default deal stage probabilities in HubSpot are often based on assumptions. Watching how actual win rates differ from those assumptions highlights where forecasts may be overstated and where pipeline expectations need to be reset.

“Tie deals to when they entered a stage, not when they closed.”

Using the Entered Deal Stage Date property ensures win rates are calculated based on when deals actually hit a stage. This makes comparisons over time more accurate and avoids skewing results toward recent closes.

“Stage-level win rates show exactly where coaching is needed.”

Breaking win probability down by stage and by rep reveals where deals consistently fall apart. In Tory’s case, this made it possible to identify late-stage drop-offs for a specific rep and focus coaching on that exact point in the funnel.