Stages defined by buyer actions with explicit exit criteria, so two reps stage the same deal the same way.
Our forecast is never right. Rebuild our pipeline stage definitions.
Seven stages collapse to five: stages 3 and 4 convert within two points of each other, so the split carries no information. Stage 5 is where reps diverge most — two skip it entirely — and its new exit criterion requires a named economic buyer, which is the fact that best predicted closing in your own data.
A sales pipeline stages template is only useful if a stage means the same thing to every rep. This one defines each stage by an observable buyer action, writes the exit criteria that must be true to advance, and sets the hygiene rules that keep the forecast from filling up with deals nobody has touched.
A definition of each deal stage in terms of what the buyer has done, plus the exit criteria required to advance, the expected duration, and the rules for stalled deals. It exists so staging is a fact rather than an opinion.
The conditions that must be true before a deal moves forward — a stakeholder identified, a technical requirement confirmed, a timeline stated. Without them, "advanced to negotiation" reflects rep optimism rather than deal progress.
Four to six for most B2B motions. More stages produce finer-looking reports and worse data, because reps stop distinguishing between stages whose definitions overlap.
By defining what happens to deals that exceed the expected stage duration: flagged, pushed, or closed-lost. Pipelines without decay rules accumulate deals that inflate the forecast and stop anyone trusting it.
The agent reads historical deal data from HubSpot or Salesforce to derive real stage durations and conversion rates rather than aspirational ones, and can compare rep-level staging to find stages being interpreted differently. Definitions export as CRM stage descriptions, required fields, and hygiene automation rules.
Historical deals give real durations and conversion rates per stage.
The agent flags overlapping stages and rep-level inconsistency.
Each stage gets conditions that must be true to advance.
Update stage descriptions, required fields, and stalled-deal automation.
Stages anchored to buyer actions, not rep activity
Durations and conversion rates derived from your closed deals
Flags stages different reps interpret differently
Hygiene rules ship with the definitions
Commonly qualification, discovery, evaluation, proposal, and negotiation — but the labels matter far less than whether each has exit criteria. Borrowed stage names with no criteria produce the same unreliable forecast as invented ones.
As observable facts: a specific stakeholder engaged, a requirement confirmed, a timeline stated. If a criterion cannot be checked by someone other than the deal owner, it will not hold under forecast pressure.
Past six, most teams start using stages inconsistently. The tell is a stage whose deals convert at a similar rate to the one before it, which means the split is not carrying information.
Flag them at a multiple of expected stage duration and require a decision — push with a reason or close as lost. Deals that sit indefinitely are the main reason forecasts drift from reality.