A sales pipeline stage should mark a commitment the buyer made, not a task a rep completed. Most B2B teams run five to seven stages, from first contact to a closed deal, and each one needs an exit criterion a rep can check against rather than judge on instinct.
Get the stages wrong and the pipeline report becomes decoration. A deal sitting in "proposal" for six weeks with no signed scope looks like a forecasting problem, but it usually started as a stage with no real exit test attached, so nobody could say honestly that the deal had actually moved.
What a pipeline stage actually marks
Most sales teams write their stages around what a rep did that week. Sent a proposal. Had a call. Followed up again. None of that tells you whether the deal moved, because a rep can do all three with a prospect who was never going to buy.
A stage should mark something the buyer did instead, a commitment that is harder to walk back than a seller action is to log. "Proposal sent" is a task. "Buyer confirmed budget and a decision date in writing" is a commitment. The second one tells you something real about the deal's odds. The first one only tells you what the rep did that day.
That distinction is what a sales pipeline is built to track in the first place, one deal at a time, by stage. Get the stage definitions wrong and the whole report turns into a list of rep activity dressed up as progress.
The sales pipeline stages most B2B teams use
Most B2B pipelines run on five to seven stages. Fewer than that oversimplifies a deal with several stakeholders and a multi-step approval. More than eight turns the board into overhead, since reps spend more time moving cards than selling. A look at pipeline conversion benchmarks puts the average opportunity-to-close rate at roughly 22 percent, a number that only means something once every deal counted in it crossed the same stage line as every other one.
These six hold up across most B2B motions, but the exact count matters less than whether every name on the list reflects something the buyer did, not a step the rep checked off alone.
How many stages your pipeline actually needs
Adding a stage feels like adding precision. Past seven or eight, a pipeline mostly ends up measuring how finely a team likes to slice its own paperwork. HubSpot's own guide to pipeline stages lands on a similar count, and recommends against splitting one buyer decision into two or three internal steps just because different people happen to touch it.
A useful test: if two stages never show a different close rate or a different average deal size in your reports, they are not two stages. They are one stage with an extra click in the middle. Collapse them, and the report that is left is one a sales manager can act on in a pipeline review instead of just scroll through.
The same test works in reverse. If deals in a single stage routinely sit there for months with wildly different odds of closing, that one stage is probably hiding two real buyer decisions, and splitting it will tell you more than adding a tenth stage somewhere else on the board.
How to write exit criteria that actually hold up
Every stage needs an exit criterion a rep can answer with a yes or a no, not a feeling. Write it as a single checkable sentence, not a paragraph of guidance a rep has to interpret under deadline pressure.
For a deal with a longer B2B cycle, add a stage for any step that genuinely blocks a signature, technical validation, procurement review, legal review, rather than folding it into "negotiation" where it goes untracked until it quietly costs you the quarter.
Why B2B stages need a buying group, not one contact
A pipeline stage can look perfectly healthy and still be wrong, because it was built around a single contact instead of the group that actually decides. Research on B2B buying puts the average committee at 13 stakeholders across two or more departments, and the same research found that 86 percent of B2B deals stall somewhere in the process, often because one of those 13 was never identified in the first place.
A deal sitting in "negotiation" with a confirmed champion and nobody else named is not actually in negotiation. It is a guess with a stage label attached. See why deals actually stall for the pattern this creates, a rep keeps a deal moving on paper while the real blocker, an economic buyer or a procurement contact nobody has spoken to, sits outside the pipeline entirely.
Buying group mapping exists for exactly this gap. Instead of a rep guessing who else needs to sign off, it surfaces the roles a purchase like this one typically involves, with a confidence score attached to each, so an exit criterion like "stakeholders confirmed" means something closer to actually confirmed.
Common mistakes when defining pipeline stages
Where to start this week
Pull your current stage list and test each one against a single question: what did the buyer do, not the rep, to earn that label. Any stage that fails the test needs a rewritten exit criterion before your next pipeline review, not a new field added to the CRM.
Then check how many named stakeholders sit on your ten largest open deals. If most show just one, that is the stall risk hiding behind a pipeline that otherwise looks healthy stage by stage. You can see how LeadLx supports mapping that group on the pricing page, starting with a free plan.
