Deal stages are the discrete checkpoints a deal moves through from creation to close. Defined well, they make pipeline data reliable, forecasts honest, and coaching mechanical — every stage answers a specific question. Defined poorly, every rep stages on vibes, and every report is fiction. The framework here is exit-criteria-based: a deal moves to the next stage only when a specific buyer-side commitment is true.
The exit-criteria principle
Each stage has a single defining question with a verifiable answer. Reps cannot move a deal forward unless they can produce evidence the criteria are met. This is the difference between data you can forecast on and data you cannot.
A good stage definition includes:
- The exit criterion (what must be true)
- The required artifact (what must exist in the CRM)
- The expected duration (so stalled deals are flagged automatically)
- The stage probability (used for weighted pipeline)
A standard 6-stage B2B framework
| Stage | Exit criterion | Artifact | Probability |
|---|---|---|---|
| 1. Qualification | Confirmed pain, role, timing | Discovery notes, multi-thread plan | 10% |
| 2. Discovery | Champion identified, pain quantified | Champion-confirmed business case | 20% |
| 3. Solution fit | Demo delivered, technical fit confirmed | Mutual evaluation plan | 35% |
| 4. Proposal | Pricing presented, scope agreed | Signed mutual action plan | 60% |
| 5. Negotiation | Verbal yes, redlines from legal | Counter-signed term sheet or order form | 80% |
| 6. Closed-won | Signature received | Executed contract | 100% |
The probabilities should match your historical conversion data, not aspirational numbers. If your “Negotiation” stage closes at 60 percent, set the probability at 60, not 80.
The three pathologies of bad stages
- Activity-based stages. “Demo scheduled” is not a stage; it is a calendar event. The stage is what the demo proved.
- Rep-perception stages. “Strong interest” is not an exit criterion; it is a feeling. Define what the buyer did, not what the rep thinks.
- Too many stages. Eight or more stages means reps spend time on stage hygiene instead of selling, and most stages get used as parking lots.
Five to seven is the right count for most B2B motions.
How to roll this out
- Audit closed-won deals from the last 6 months. What was true at each stage?
- Write exit criteria as testable statements. “Champion confirmed pain in writing” beats “champion is engaged.”
- Set artifact requirements. Force CRM enforcement — no proposal stage without an uploaded mutual action plan.
- Pilot with one team for 30 days. Calibrate before company-wide rollout.
- Review monthly for the first quarter. Conversion rates between stages will tell you which definitions are too tight or too loose.
Common pitfalls
- No stage demotion process. Reps will move deals forward but never back. Build the demotion path and enforce it.
- Stage probabilities chosen by feel. They are stage-conditional historical close rates. Compute, don’t guess.
- Identical stages across segments. Enterprise needs different stages than transactional SMB. Two motions, two stage maps.
- Allowing skip-stages. A deal that jumps from Discovery to Proposal is either misstaged earlier or not real.
Related
- Pipeline velocity — the metric clean stages enable
- Forecast accuracy — stage probability is the input