A forecast can become increasingly precise without the underlying opportunity becoming any more predictable.

An opportunity enters the CRM at 40 percent probability. A productive technical meeting raises it to 60. A quotation goes out, purchasing becomes involved, and the number reaches 80, with closing set for the end of the quarter.

A month later, nothing has happened. The customer is still interested, but internal approval is unsettled, no purchase date has been agreed, and the budget has not been released.

The CRM still shows 80.

The number mixes two different kinds of information

Some forecast inputs are observations: a quotation was sent, a meeting took place, a technical requirement was confirmed. Others are interpretations: the customer seems enthusiastic, purchasing appears ready, approval is expected before quarter-end.

Both are useful. The problem starts when they are handled as if they carried the same certainty — when a number backed by confirmed commitments and a number backed mainly by a salesperson's confidence are both displayed as "80 percent" in the same pipeline view.

None of which means the 80 percent was necessarily wrong. Some organizations attach historically calibrated probabilities to sales stages, which gives such a number a defensible basis. The problem is not that these numbers exist. It is that on screen, a calibrated estimate and a subjective judgment wear the same format — and readers cannot see which is which.

How each handoff adds certainty without adding information

The deeper mechanism is what happens as the number travels.

A salesperson estimates the deal will close in November, based on positive conversations and the customer's stated intention. The sales manager incorporates the estimate into the monthly forecast — and in doing so creates a real risk: the estimate starts being handled as a fact rather than a judgment. In the management meeting, "the November order" becomes part of the revenue outlook. By the time it informs a resource or inventory decision, it is spoken about as a plan, not a prediction.

At no point did anyone add information about the customer's actual decision process. But at each step, the uncertainty became less visible — because each new audience received the number already dressed as established fact.

Confidence compounds as information travels. Evidence does not.

What the number is standing on

Commercial organizations need forecasts, and estimates will always involve judgment. The useful discipline is not demanding more precise numbers — precision is cheap — but knowing what each number stands on:

• Is the expected timing based on an agreed customer milestone, or on the salesperson's own target?

• Has the budget been confirmed by the customer, or by hope?

• Which parts of the probability rest on observations, and which on expectation?

Those questions do not automatically produce a better percentage. They reveal whether the percentage has a defensible basis — which is what the forecast's readers, several layers up, are actually relying on.

From the book

Forecasts, probabilities and the certainty they can project are explored in more depth in Chapters 2 and 10 of Sales Between the Lines, alongside the wider role of observation and interpretation in understanding how much a number actually says about the opportunity behind it.