How to build a sales forecast your CEO will believe
A forecast becomes credible when you can show how wrong you have been before, state the range rather than a point, and name in advance the things that would move it. Accuracy helps, but it is not what buys trust — plenty of accurate forecasters are disbelieved because they produce a number with no visible reasoning, and one bad quarter destroys them.
1. Show your error history before you show the number
The most effective thing you can put in front of a CEO is twelve rows: what you forecast each month, what closed, and the gap. It does three things at once. It proves you measure yourself. It tells them how much to discount this month's number. And it moves the conversation from whether you are right to how wrong you usually are, which is a question with an answer.
If the history is bad, show it anyway. "I have run 20% high for six months, so I am taking 20% off this number" is the single most credible sentence available to a sales leader. It converts a weakness into a method.
A forecaster who knows their own bias is more useful than one who is occasionally right. The first can be corrected for; the second cannot be planned around.
2. Give a range and the odds, not a point
A single number invites exactly one question — is it right — and you cannot answer it. A range invites a better one: what would have to happen to land at the top of it.
Say it plainly: "We land between X and Y four quarters in five, the middle is Z, and we are at 60% to make target." That is a sentence a CEO can plan around, because it tells them whether to hold hiring or commit to the board. A point estimate hides precisely the information they need, which is how confident you are.
3. Name the deals that move it
Every forecast has two or three deals that decide it. Name them, with their value and their odds, and say what each one hinges on. This is the difference between a number and an argument.
It also protects you. When a named deal slips, you reported that risk in advance and the conversation is about the deal. When an unnamed deal slips, the conversation is about your forecasting.
4. Use the same method every month
Changing how you build the number destroys comparability and reads as motivated. If you weighted by rep judgement last month, do it again this month, even in a month where the answer is unflattering. Credibility is built by a method that survives a bad quarter.
A corollary worth holding to: do not adjust the forecast upward because the number is uncomfortable. Everyone in the room can tell, and the next three forecasts are discounted whether or not they deserve to be.
5. Make revisions boring
Forecasts move. What damages trust is a forecast that moves without explanation. Keep a snapshot of what you committed and diff it: "Down 140k because Dunmore went from 70 to 40 and Fenwick slipped to next quarter" is a defensible sentence that takes ten seconds.
Without the snapshot you are reconstructing from memory, which is where "I think it was always around that" comes from — and that is the sentence that costs you the room.
What to stop doing
- Stop presenting coverage as a forecast. 3× coverage is a pipeline measure. A CEO hearing it as reassurance is being misled, usually unintentionally.
- Stop giving best case first. Lead with the conservative end. Upside stated after a credible floor sounds like upside; stated first it sounds like the plan.
- Stop hedging verbally instead of numerically. "I feel good about it" carries the risk in your tone of voice, which nobody can plan against. Put the uncertainty in the number where it belongs.
I ran this on my own numbers for a year: recorded every monthly call, found I was high by a similar margin nearly every time, and started subtracting it out loud. My error fell from roughly 50% to about 14%, but the part that changed the room was showing the history, not the improvement.