Why your pipeline misleads you
The stages describe what you do, not what the customer does.
“Proposal sent” means nothing. Sending is your activity, and a proposal that was sent may have been read, ignored or deleted. On the other hand, “proposal reviewed with the customer, follow-up meeting booked” is a fact you can check.
Rewriting your stages around what the customer does is the change that improves a forecast most, and it costs nothing.
A set of stages that works
| Stage | Criterion you can check | Starting probability |
|---|---|---|
| Qualified contact | We have talked and there is a fit | 10% |
| Need confirmed | They told us their problem and there is a budget | 25% |
| Proposal reviewed | They have seen it and we have discussed it | 45% |
| Negotiation | We are discussing terms, not yes or no | 70% |
| Verbal commitment | They said yes; the signature is pending | 90% |
The probability belongs to the stage, not to the opportunity. If each rep enters whatever percentage they feel like, the forecast is the sum of the team’s optimism.
The next-step rule
Every open opportunity has a next step with a date. No exceptions.
If nobody knows what happens next, the opportunity is not alive: it is in the pipeline out of inertia. A pipeline full of dead opportunities makes any forecast impossible. And it demoralizes the team, who see a list that never moves.
Review opportunities with no next step every month. The first time, the cleanup is usually big.
How to calibrate probabilities with your own data
Do not make them up: take them from your history.
Stage probability = opportunities that reached the stage and were won ÷ opportunities that reached the stage
One year of data is enough for a reasonable number. Review it every six months.
Example
Last year, 40 opportunities reached “proposal reviewed” and 14 were won. The real probability for that stage is 14 ÷ 40 = 35%, not the starting 45%. With five $20,000 opportunities in that stage, the weighted forecast is 5 × $20,000 × 35% = $35,000.
How to know the forecast works
By measuring the gap between forecast and closed deals, month after month. If it is large, the probabilities are poorly calibrated.
And watch out for bias. If the forecast always comes in above what closes, it is not bad luck: it is the probabilities or the stages.
Close reasons
It is the part nobody fills in and the one worth the most. Record why deals are won and lost, with a reason from a fixed list and the competitor when there is one. That turns your history into something useful.
After a year you know who you lose to, which industries you win in and which objection keeps coming up. That changes the whole sales pitch, and it only costs a dropdown at close.
This quarter, rewrite your stages around what the customer does, require a next step with a date and calibrate probabilities with your history. With that, the forecast stops being a feeling and becomes a number you can defend.
Last updated: