At Monday's sales meeting a large deal is still pencilled in for Friday. The quote is with the customer, but their purchasing committee meets in two weeks. The sale may well happen. The delivery manager has to decide whether to reserve people for this month.
A useful CRM sales forecast separates orders already won, the value open deals could bring in by period end and the extent to which the number depends on one customer.
Forecast the period you need to staff
Suppose you are forecasting November orders, excluding VAT. Take won orders at full value. For each open opportunity, multiply the expected order value by its estimated probability of closing before month end. That probability depends on how far the sale has progressed and how much time is left.
| Item | Order value | November outlook | Forecast contribution |
|---|---|---|---|
| Orders already won | SEK 300,000 | Won | SEK 300,000 |
| Deal A, quote submitted | SEK 400,000 | 25% | SEK 100,000 |
| Deal B, contract with buyer | SEK 200,000 | 70% | SEK 140,000 |
| Total | SEK 900,000 | SEK 600,000 still open | SEK 540,000 |
Save what the sales team knew on the day
When a salesperson moves the close date to December, they overwrite information you need to assess the November forecast. Capture the pipeline at the same time each week. A first snapshot can hold deal ID, value, stage, owner and expected close date. Keep the forecast period and any manual adjustments alongside it.
Estimate the weights from deals in comparable circumstances. Of proposals open with two weeks left, how many were won by month end? Include deals that slipped or were still open at the end of the month. Display the number of observations beside each percentage so the team can tell when it rests on only a handful of opportunities.
Reopen the forecast when the month closes
If actual orders in the example came to SEK 500,000, the forecast was SEK 40,000 too high. Compare meetings held at similar distances from month end. Track the average signed error as well as the size of the misses regardless of direction. A high forecast can otherwise cancel out an equally poor low one.
Inspect the deals behind the difference. Did close dates slip, did order values change or did new opportunities arrive and close after the meeting? A forecast based on open opportunities misses that last group entirely. If short-cycle deals recur, show a separate estimate for them alongside the pipeline forecast.
Common questions
What if our CRM does not retain deal history?
Start keeping a weekly snapshot and state your assumptions until you can test them. Today's close date cannot tell you what the salesperson expected at last month's meeting.
Can finance use the forecast as a revenue budget?
That requires a further step. Allocate orders to delivery periods using your finance team's definitions. An annual contract can create a large November order while the related revenue spans several months.