How to Build a Sales Pipeline That Actually Works
How to define pipeline stages and write transition rules: pipeline management, bottlenecks and practical guidance for sales forecasting.
A sales pipeline is not reporting decoration; it is a diagnostic tool. Built properly, it answers “will we hit the number this month?” with evidence rather than a guess.
What a pipeline is and why it has stages
A sales pipeline is the ordered set of stages a potential sale passes through from first contact to close. The point of stages is not classification but seeing where deals are lost. Without stages you cannot know where the loss happens.
Three rules for defining stages
- A stage should describe the customer’s behaviour, not your intention. “Quote sent” is a good stage. “Following up” is not, because it cannot be verified.
- Every stage needs a written exit criterion. Under what condition does a record move to the next stage? Without this, stages mean different things to different reps.
- Keep the number of stages low. Five to seven works for most businesses. A twelve-stage pipeline does not get filled in, and if it does nobody reads it.
A typical pipeline
1. Qualified
Need, budget and authority confirmed. Exit: a conversation with the decision-maker has taken place.
2. Discovery
Their process is understood and the scope is clear. Exit: agreement on the scope of the solution.
3. Quote
A written quote has been sent. Exit: the customer has confirmed receipt.
4. Evaluation
The customer is comparing and asking questions. Exit: objections have been answered.
5. Negotiation
Price, scope and terms are being discussed. Exit: terms agreed.
6. Won / Lost
The outcome is recorded. A lost deal must always be tagged with a reason.
Habits that break a pipeline
- Carrying dead records. An opportunity with no movement for three months inflates the pipeline and makes the forecast lie. Add an automatic “stale” rule.
- Updating stages in bulk at month end. A pipeline that is not current is a table describing the past.
- Marking everything 90%. If probability is not tied to the stage, the forecast is useless.
- Leaving the loss reason blank. “Price” is easy to write; the real reason is usually timing or scope.
Four numbers to take from the pipeline
- Stage-to-stage conversion: which transition loses the most? That is your bottleneck.
- Time in stage: how long do deals wait on average? A lengthening stage usually signals missing information.
- Average deal size: how many opportunities do you need to make the number?
- Pipeline coverage: total open value as a multiple of target. The healthy range depends on the business model, but below three is usually risky.
How to forecast
The simple method: assign each stage a win probability derived from your own history and weight opportunity values by it. What matters is that the probability comes from past data, not the rep’s feeling. If you do not have three months of data yet, do not publish a forecast — collect data first.
Frequently asked questions
Should I build separate pipelines for different products?
If the sales process differs materially, yes. A project-based implementation and a subscription sale do not fit one pipeline; mixed together, the conversion rates become meaningless.
What should happen to lost deals?
They should not be deleted. Tag them with a loss reason and archive them; six months later that is the best list to reach out to.
Who updates the pipeline?
The owner of the opportunity. A pipeline maintained by managers reflects what management sees, not what the field sees.
Next step: to define your own stages and see the pipeline take shape, request a demo.