Why a stage-weighted forecast is only as good as the stage definitions, what an exit criterion is, five stages with criteria a rep can verify and a manager can audit, how to check from the CRM that deals in a stage actually meet its criteria, the stage probabilities derived from the team's own history rather than set in a workshop, and the versioning that keeps last year's pipeline comparable.
A stage-weighted forecast looks rigorous: values times probabilities, summed. Its rigour depends entirely on what a stage means, and at most teams a stage means whatever the rep clicked. This guide sets out exit criteria that can be verified, the CRM audit that checks them, probabilities from history, and the versioning.
A fact, not a judgement:
| Stage | Exit criterion to enter | Verifiable by |
|---|---|---|
| Qualified | Need, budget owner and timeframe recorded | Three fields populated |
| Discovery complete | Economic buyer identified as a contact with role; requirements documented | Contact role field; attached document |
| Proposal sent | Written proposal delivered, dated | Attached document with date, or logged activity of type proposal |
| Negotiation | Customer has responded to the proposal in writing | Logged activity of type customer response, dated after proposal |
| Closed won | Signed order form or contract | Document status signed |
"Customer is very interested" is not a criterion. "Proposal document attached, dated" is.
Weekly, per deal:
In stage S and criterion field for S is empty → fails audit
Per rep: share of deals failing, and the weighted value they carry. A deal that fails is not moved automatically; it is listed, and the rep either populates the field or moves the deal back.
Per stage, per quarter, from outcomes:
Probability = deals that reached closed won ÷ deals that were in the stage at the same point in past quarters
| Stage | Workshop probability | Historical conversion | Difference on $10m in stage |
|---|---|---|---|
| Qualified | 10% | 8% | −$200,000 |
| Discovery complete | 25% | 19% | −$600,000 |
| Proposal sent | 50% | 32% | −$1.8m |
| Negotiation | 75% | 61% | −$1.4m |
The workshop forecast was four million dollars high on the same pipeline.
Opportunity identifiers only.
When a criterion changes, the definition version changes, and the pipeline before the change is reported under the old version. A stage whose criterion tightened in April will show fewer deals from April, and that is not a pipeline decline; the version on the report says so.
Σ deals in stages = open pipeline, per snapshot every deal in exactly one stage with its criterion fields recorded
Criteria as judgements. Every deal meets them.
No audit. Criteria written, never checked, forgotten by the second quarter.
Workshop probabilities. Confident, round, wrong.
Criteria changed silently. The pipeline drops and nobody knows it was the definition.
Mapped once, the pipeline snapshots, the outcomes and the versioned definitions produce the stage audit per rep, the historical probabilities and the weighted forecast under both every week. Covirage builds this from the exports as they are. The metrics governance solution describes the setup, and the forecasting methods guide covers where the stage-weighted method sits among the others.
As few as have distinct exit criteria and distinct conversion rates. Five is common: qualified, discovery complete, proposal sent, negotiation, closed. Eight stages with three that convert at the same rate are three stages with extra clicks.
Each criterion maps to a field: economic buyer to a contact with a role; proposal sent to an attached document or a dated activity; order form to a document status. A deal in a stage whose criterion field is empty fails the audit and is listed. The audit runs weekly and is the measure of stage hygiene.
Because a workshop sets proposal at 50 percent and the team's history says 32. The weighted forecast at 50 is wrong by the difference on every proposal-stage deal. Historical conversion per stage, from outcomes, is what the probability actually is, and it is recomputed each quarter.