CRM sales forecasting estimates future revenue from open opportunities. A common method is pipeline-based: each open deal has an amount, a close date, and a probability, and the weighted amount is amount times probability. Some CRMs also let managers apply a weight by stage instead of asking every salesperson to type a probability.
The forecast is not a promise. It is a calculation on top of pipeline stages. If stages are vague, the forecast is vague.
The fields that drive the number
Amount. The value you are willing to forecast. Decide whether that is the quotation total, the likely first invoice, or recurring revenue for a defined period. Mixing a three-year contract total with a one-month fee makes the chart incomparable.
Close date. The date you expect the deal to reach Won, using your definition of Won. A date in the past on an open deal should stand out, not hide inside the month.
Probability or stage weight. Either a percentage on the opportunity or a default percentage on the stage. Stage weights are more consistent. Per-deal percentages are useful only if managers review them. Using both without a rule means two forecasts.
Forecast category. Labels such as pipeline, best case, commit, and closed. Commit should mean “we would be surprised if this slipped,” not “the salesperson is hopeful.” Write the definitions next to the field.
Excluded deals. Lost and disqualified records do not belong in the open forecast. Won deals belong in actuals, or in a “closed this period” line, not double-counted as commit.
A small numerical example
Three open opportunities:
| Deal | Amount | Stage weight | Weighted |
|---|---|---|---|
| Repeat order, quotation accepted verbally | 20,000 | 70% | 14,000 |
| New site, quotation sent | 80,000 | 30% | 24,000 |
| Early discussion | 50,000 | 10% | 5,000 |
The unweighted pipeline is RM 150,000. The weighted forecast is RM 43,000. Reporting only the unweighted number will disappoint anyone who expected RM 150,000 to arrive this month. Say which number you are looking at.
Building the month’s forecast
- Include only open opportunities in the period
- Drop records with a blank amount or close date
- Apply the agreed stage weight or probability
- Separate commit from upside
- Compare next month with what actually won
Common forecasting errors
- Happy close dates. Everything is dated the last day of the quarter, then slips.
- Stale amounts. The quotation was revised and the opportunity was not.
- One giant deal. A single unreviewed opportunity dominates the total. Inspect the top deals, not only the sum.
- Stage skipped. People jump to a high-weight stage without the evidence you defined.
- No comparison. If you never look back at last month’s forecast versus won revenue, the weights never improve.
When a figure changes, the audit trail should show who moved the amount or the date.
Reading the history with the team
A monthly review is easier if you agree what “good enough” evidence looks like. For a disputed discount, the trail should show the old amount, the new amount, the user or integration, and the time. A note added afterwards that says “we think it was the salesperson” is not a substitute. If the product only stores the latest value, write that limitation into the admin guide so managers do not promise a history the screen cannot show.
Train people to log the outcome of a call in the activity, and to leave the amount field to the audit. Mixing a price change into a note (“updated to 10k”) creates a second, unofficial ledger.
A practical example
A manager’s meeting uses two lines: commit (stage weight 70% and above, close date this month) and upside (everything else open this quarter). Deals with a close date before today are listed as overdue and are not silently included. After the month, the team notes which commit deals slipped and whether the stage definition was wrong or the date was wishful.
Key takeaways
- Forecast from amount, timing, and an agreed weight.
- Publish the weighted figure and the category definitions.
- Review the largest deals and the overdue dates.
- Compare forecast to won revenue so the model can be corrected.
FAQ
Is a CRM forecast the same as a finance budget?
No. The budget is a target. The forecast is a view of current deals. They should be compared, not treated as the same field.
Should we forecast from leads?
Only if you have a measured conversion rate from lead to won revenue, kept separate from the opportunity forecast. Mixing raw leads into commit inflates the month.
How often should weights change?
Rarely. Change them when several months of results show a stage is consistently over- or under-trusted. Changing weights every week makes the history unreadable.