Pipeline stages are checkpoints a deal passes on its way to closed — and the entire value of the concept rests on one discipline: each stage must have an exit criterion, a fact that's true or false. "Proposal sent" is a stage. "Feeling good about it" is a mood. Most broken pipelines are broken here, not in the software.
Here's the standard six-stage model, the exit criteria that make it honest, and when to deviate.
The standard six stages
1. Lead / New
Someone entered your world — form fill, list import, referral, reply. Nothing is known yet beyond contact info.
Exit criterion: qualified against your bar (fit + intent) → advance; otherwise nurture or disqualify. This is the MQL/SQL line, and speed matters most right here.
2. Qualified
A real potential buyer: right profile, plausible need. Worth a rep's calendar.
Exit criterion: a discovery conversation happened and confirmed a problem you solve (your BANT-style checklist, applied as judgment).
3. Discovery / Meeting
You've talked. You understand their problem, stakeholders, and rough timeline.
Exit criterion: a concrete solution discussion is scheduled or delivered — demo done, scope agreed, or the prospect asked for numbers.
4. Proposal
A specific offer — scope and price — is in their hands.
Exit criterion: verbal or written intent ("yes, pending X"). Deals rot here more than anywhere; this stage needs the tightest follow-up cadence and the strictest aging alarm.
5. Negotiation / Commit
They want it; you're closing terms, legal, security review, signatures.
Exit criterion: signed, or dead. Nothing lives here longer than your contracting process actually takes.
6. Closed (Won / Lost)
The outcome — with a required reason code either way. Lost isn't failure; unlabeled lost is. Those codes are your win/loss report, which is your roadmap.
When to deviate from six
- Add a stage only when a distinct exit criterion exists. A security-review or pilot stage earns its place if it's a real, dated checkpoint in your motion. "Strong interest" does not.
- Transactional sales: collapse to four. Lead → Qualified → Proposal → Closed fits short cycles; more stages than conversations is theater.
- Enterprise: split Negotiation. Legal/procurement often deserves its own stage because its dwell time and owners differ from commercial negotiation.
- Never model the customer journey as your pipeline. Stages track your selling checkpoints; awareness/consideration language belongs to marketing funnels, not deal management.
Stage design mistakes that make pipelines lie
- Mood stages. Any stage a rep can enter or leave based on feel will be gamed at forecast time. Facts only.
- No aging rules. Every stage needs a normal dwell time and an alarm past it — otherwise stage four becomes a hospice (the zombie math is in our coverage guide).
- Skipping backwards silently. Deals do move backwards — a champion leaves, scope reopens. Track regressions explicitly; they're your best early-warning signal.
- Percent-complete as probability. Stage 4 of 6 isn't 66% likely to close. Use historical per-stage close rates — or per-deal ML probabilities, which is the same idea with the guesswork removed.
Keeping stages honest, automatically
Stage discipline is another practice that decays under busyness: criteria blur, aging alarms get snoozed, reasons go uncoded. The systematic fix is making the pipeline maintain itself — in AutomateNexus CRM, Karrie nudges stage updates from actual activity (a sent proposal moves the deal; silence past dwell time flags it), keeps every stage's follow-up sequence running, and the daily brief surfaces exactly which deals sit past their welcome. The six stages stay six facts.
Pipeline stages FAQ
How many stages should a sales pipeline have?
As many as you have verifiable checkpoints — for most B2B teams that's five to seven. Fewer beats more: every stage must earn its exit criterion.
What's the difference between a pipeline and a funnel?
Same journey, different lens: the funnel is the aggregate conversion view (marketing's tool); the pipeline is the deal-by-deal management view (sales' tool). Stages belong to the pipeline; percentages belong to the funnel.
What should trigger a stage change?
An event, not a feeling: meeting completed, proposal delivered, verbal commit received. If you can't name the event, the deal doesn't move.
Should lost deals stay in the pipeline?
They leave the active pipeline but stay in the system with a reason code — feeding win/loss analysis and the quarterly re-engagement sequence that revives the "not now" cohort.