Pomio CRM keeps stage, probability, forecast category and notes on the same deal. Giving every open deal a probability is the baseline; the next question is where that number comes from. If each salesperson invents a percentage, the pipeline looks precise and is often wrong. Stage defaults fix the starting point; a short note covers the exception.
Why free-form probability drifts
Without a default, people tend to round up. A deal that just entered Discovery shows 60% because the call went well. Another sits in Proposal at 40% because the seller is cautious that week. Those numbers do not compare, and the weighted forecast becomes a mood board. Stage defaults put the same starting point on every deal in the same stage, so differences come from real overrides, not from habit.
What a stage default should be
Keep the set simple and shared. Typical SME defaults look like Lead 10%, Qualified 25%, Demo 40%, Proposal 60%, Negotiation 80% — adjust to your own win rates, not to a textbook. Write them into the shared stage definitions next to the exit criteria, so sales and management agree what “Proposal” means and what probability it carries. Revisit the defaults once or twice a year against closed deals; do not tweak them every week.
When the deal moves, the probability moves
When a deal advances or goes back a stage, set the probability to that stage’s default automatically. That habit stops stale percentages: a deal that slipped from Negotiation back to Proposal should not still show 80%. Combine this with days in current stage in the weekly pipeline review; a deal that aged in Proposal at 60% with no next step is a review item, not a forecast line you trust blindly.
When an override is allowed
Overrides are for clear exceptions, not for “I feel good about this one.” Good reasons include a signed letter of intent while still in Demo, a verbal yes pending legal, or a sudden budget freeze in Negotiation. Bad reasons include liking the contact or needing the quarter to look fuller. Require a short note on every override, the same discipline as a note on a stage change. Without the note, the override is just another free-form guess.
Probability and forecast category together
Probability is not a substitute for a forecast category. Category (pipeline, best case, commit) is a management call about what to count this period; probability weights the amount. A deal can sit in Proposal at the 60% default and still be “pipeline,” not “commit,” until there is a dated next step. Keep deals without a next step out of the committed view, and keep unqualified deals out of the forecast altogether.
How Pomio CRM helps
In Pomio CRM each deal shows its current stage, probability, forecast category and notes together. When the stage changes, the probability can follow the stage default; when someone overrides it, the note stays on the deal. In the weekly review you can sort by stage, spot overrides without notes, and check whether the weighted forecast still matches what the team actually expects to close.
A practical checklist
Set a default probability per stage from your own win rates; write defaults into the shared stage definitions; update probability to the stage default on every stage change; allow overrides only with a short note; review overrides in the weekly pipeline review; keep forecast category separate from probability; exclude deals without a next step from commit; keep unqualified deals out of the forecast; revisit stage defaults once or twice a year.
Further reading
Should every deal start at 0% until someone sets a number?
No. A stage default is clearer and faster. Starting at 0% just delays the same argument and leaves early-stage deals invisible in a weighted forecast.
Can two stages share the same default probability?
Yes, if their win rates are similar. Prefer fewer, honest defaults over a unique percentage for every label.
What if our win rates per stage are unknown?
Start with conservative industry-style defaults, then replace them with your closed-won and closed-lost history after a few months of clean stage use.
Should lost and won deals keep a probability?
Closed-won is 100% and closed-lost is 0%. Open-deal probabilities matter for the forecast; closed deals matter for learning the next set of defaults.