Pipeline stages should reflect observable progress in a customer’s decision process. Give each stage an entry event, an exit condition and an owner. Separate a seller’s activity from a buyer’s commitment, and preserve stalled, paused and closed states without inventing progress to improve the report.
Name events that another person can verify
List the real decisions in your sales journey before choosing stage names. “Interested” can mean a polite reply, an accepted meeting or an active purchasing project. Replace the vague label with an event such as “discovery conversation accepted” or “scope confirmed for proposal”. Describe the evidence required and where it is stored. A colleague should be able to classify the record without asking what the salesperson feels.
Distinguish actions you control from commitments the buyer makes. Sending a proposal is a seller activity; the customer acknowledging that the scope is under review is a different event. Both may be worth recording, but they should not be treated as equivalent proof of progress. Keep the definitions appropriate to the offer. A standard self-service purchase may need fewer stages than an engagement involving technical assessment and procurement.
Write entry, exit and exception rules
For each stage, record the entry event, information needed, responsible owner and condition for moving forward. Explain what happens if a deal skips a normal step or returns for scope revision. Use a separate task list for reminders so unfinished internal work does not become a new customer stage. Stages describe the opportunity’s state; tasks describe what someone needs to do about it.
Add explicit paused, outside-scope and closed reasons. A record waiting for a buyer’s budget meeting should not look identical to one nobody has answered for months. Never interpret silence as acceptance. Define a review rule for records without a current next action, and preserve the reason when closing them for inactivity. Keep dates for stage entry so reporting can show elapsed time without relying on a memory of the last conversation.
Calibrate the definitions with real records
Ask two colleagues to classify a sample of recent opportunities independently. Compare disagreements and inspect the evidence. If one treats a requested demonstration as qualification while another requires confirmed problem fit, improve the stage rule. Trial the revised definitions before automating transitions or assigning probabilities. A numerical probability attached to an ambiguous stage gives a false appearance of precision.
Review the pipeline using counts, age, next actions and evidence quality. Forecast assumptions should be explicit and revisited, especially when only a few completed journeys exist. SBA marketing guidance supports a sales plan that describes customer steps; observable stages translate that idea into day-to-day records. Preserve previous definitions when the process changes so a larger pipeline is not mistaken for growth when it merely reflects relabelling.
Limit who can alter shared stage definitions, while allowing staff to flag exceptions. Log meaningful changes and tell affected teams what changed. Sales, delivery and finance may rely on the same stage for different actions; a quiet edit can start work or trigger expectations prematurely. Check those downstream uses before introducing a new automatic transition.
Fictional worked example
A fictional scenario to explain the method, not a customer result or a market benchmark.
In the fictional firm Fern Bridge, “proposal” previously included drafts never sent, documents emailed without discussion and scopes actively reviewed by buyers. The team splits the record into a proposal task and an opportunity stage entered only when a scoped proposal has been delivered through the agreed route. A separate field notes whether the buyer has confirmed review.
Two colleagues test ten existing records and disagree about three. They inspect the event history and refine the rule for revised proposals rather than assigning whichever stage produces the largest forecast. One case becomes paused because the buyer has no current review date. Fern Bridge retains the earlier stage dates and definition version. These invented records show calibration and do not establish a standard stage probability.
TAKE THIS INTO YOUR NEXT CONVERSATION
A practical next-step checklist.
- Define each stage with a verifiable entry event.
- Separate seller tasks from buyer commitments.
- Specify exit, pause and closure conditions.
- Calibrate reviewer agreement before automating transitions.
Common questions.
Should every business use the same pipeline stages?
No. Use the decisions required by your offer and customers. Extra stages that do not change a real action add maintenance without improving understanding.
Can we attach probabilities to stages?
Yes, with suitable historical evidence and disclosed assumptions. Avoid treating an arbitrary percentage as a fact about an individual opportunity.
Further reading & source context.
These primary references provide background. They are not evidence for the fictional example above and do not endorse AccelerateLink.
Found an error or a missing qualification? Send an editorial correction.
Find my next move.
Use a practical tool to turn this idea into a clearer starting point.
Find my next move