Insights — CRM & Pipeline — 3 min read
How to Build a Sales Pipeline That Actually Predicts Revenue
A pipeline that never shrinks is not a pipeline. It is a list of every conversation the business has ever had.
Most SME pipelines fail the only test that matters: could you plan the next quarter's production, cash and recruitment from this? Usually not, because the pipeline contains everything anyone has ever discussed, dated optimistically.
A pipeline is a planning instrument. Building a useful one takes four things: honest stages, entry and exit rules, dates set by the customer, and coverage arithmetic.
Design stages around evidence, not feelings
Every stage should describe something verifiable that has happened, not how confident the salesperson feels. "Interested" and "warm" are moods. "Requirement confirmed", "decision-maker engaged", "proposal presented" are facts.
Four or five stages is enough for most SMEs. A useful default:
- 01Qualified opportunity — a real requirement, identified budget holder, and a reason to act.
- 02Solution agreed — you know what you are proposing and the customer agrees it fits.
- 03Proposal presented — the quote has been talked through, not just emailed.
- 04Decision — commercial terms under discussion with a named signatory.
- 05Won or lost — closed with a recorded reason.
Write entry and exit rules down
Without rules, stages become a matter of taste and two salespeople will report the same deal differently. Define what must be true to enter each stage and, just as importantly, what causes a deal to leave the pipeline.
Date deals from the customer's calendar
The close date should come from something happening in the customer's business — a project start, a contract end, a shutdown, a budget year. A date invented to fit your quarter is the main reason forecasts slip repeatedly by exactly one month.
If nobody can name the event that forces a decision, the deal is not in the current period. That single rule fixes most forecasting in one meeting.
Work out the coverage you actually need
Coverage is the value of live opportunities against the target for the period. The multiple you need is set by your own conversion rate: at a 25% win rate you need four times the target in genuinely live opportunities, not the vague "3x" that gets repeated everywhere.
Then account for the cycle. If deals take four months, coverage for Q4 has to exist by the start of Q3. Reviewing coverage inside the current quarter is reviewing a result you can no longer change.
Clean it, and accept that the number gets smaller
The first honest clean-up is uncomfortable — a £2m pipeline can become £700k. That is not a loss; it is the discovery that you were planning against fiction. The smaller number tells you precisely how much new opportunity has to be created this month, which is actionable.
A small honest pipeline is a plan. A large dishonest one is a comfort blanket.
Review it weekly, deal by deal
A pipeline review is not a reading of the total. It is a short conversation about specific opportunities: what has changed, what has to happen next, who does it, and by when. If a deal has had no next action for three weeks, it is not progressing regardless of its stage.
- Every live opportunity has a next action with a date and an owner.
- Every stage change is justified by evidence, not optimism.
- Every loss has a recorded reason — that data is the cheapest market research you will ever get.
Keep the CRM light enough to be true
Pipelines fail in software when the admin cost exceeds the perceived benefit. Fewer fields, honestly filled, beat a comprehensive system nobody trusts. If a field does not change a decision somebody makes, remove it.
Get those things right and the pipeline stops being a reporting exercise and becomes the tool you run the commercial side of the business from.
Think your sales operation could be performing better?
A Sales Growth Assessment finds where revenue is being lost before anything gets changed.
