What does it take to create a sales forecast we can trust?

The forecasting feature in Dynamics 365 Sales totals your sales opportunities. If they are poorly maintained, you get a precise figure based on inaccurate data—which is worse than having no forecast at all, because it looks credible.

The closing date must mean something

The most common source of error is sales opportunities with close dates in the past, or dates that are pushed forward by a month every month. Make reviewing deals with past due dates a regular part of the sales meeting. It takes ten minutes and eliminates most of the noise.

Define what the stages actually mean

The standard stages are Qualify, Develop, Propose, and Close. If eight out of ten deals are in Develop, the stage is meaningless. Decide what specifically must have happened before moving forward—for example, that the need has been confirmed in writing or that the proposal has been sent—and stick to it.

Distinguish between hope and commitment

A calculated figure based on probability and a salesperson’s own assessment are two different metrics, and both are useful. Problems arise when they are mixed. Show them separately, and be clear about which figure is used for what.

Measure accuracy

Compare the forecast with the actual result each period. If you are consistently 30 percent too optimistic, that is not a disaster—it is a known bias you can correct for. Without measuring it, you will not know whether the forecast is improving or worsening over time.