The value sits in the spread. A quarter forecast at $10M with a $9.4M floor and a $10.7M ceiling is a different business from a quarter forecast at $10M with a $6M floor and a $14M ceiling, and a single number reports them identically. The first quarter is a distribution problem. The second is a concentration problem, and it needs pipeline built now rather than a sharper estimate in week 10.
Setting the three numbers
Each point needs a definition that holds across quarters and across reps.
The worst case is the revenue that lands if nothing new goes right. Build it from deals with signed commercial terms in motion and a confirmed buying process, and nothing else.
The most likely case is the honest expectation, built on current conversion rates rather than target ones. It includes deals that need one more step to close and assumes those steps happen at the rate they historically happen.
The best case names the additional deals plus the specific condition each one requires. Without the condition written down, the best case becomes a wish and the whole exercise collapses.
Collapsing three numbers into one
When a single figure is required, the PERT weighting handles it: expected value equals (worst + 4 x likely + best) / 6. The four-times weight on the middle point keeps the estimate anchored to the realistic case while still letting an asymmetric spread pull it. If the best case sits far above the likely case and the worst case sits close beneath it, the PERT estimate lands above the middle number, which is the correct behavior.
Report the interval alongside the collapsed number. The single figure is for planning. The interval is for deciding whether the plan is safe.
Where three-point forecasts get gamed
Two failure modes recur. Reps set the worst case low enough to guarantee a beat and the best case high enough to look ambitious, which makes every number unfalsifiable. And close dates get moved rather than estimates revised, which hides a shrinking quarter inside an unchanged forecast.
ORM identifies the close-date change as the strongest available signal that a deal is slipping. A deal that moves from one quarter to the next is less likely to close at all, even when it sits in commit. So a three-point forecast where the middle number never moves while close dates keep sliding is reporting stability that does not exist. Reconcile the three points against the deal list every week and the gaming stops, because each point becomes a set of named opportunities anyone can check after the quarter closes.
Pair this with a running measure of forecast accuracy so the record shows which reps set honest floors, and read the spread against pipeline coverage to see whether the width is a pipeline problem or an estimating problem. Deals that keep sliding belong in a deal slippage review rather than in next quarter's best case.
Frequently Asked Questions
What is the PERT formula for a three-point estimate?
Expected value equals worst case plus four times the most likely case plus best case, all divided by six. The weighting pulls the estimate toward the middle scenario while still letting the outer cases move it. On a $7M worst case, $10M likely case and $13M best case, the PERT estimate is $10M.
How is this different from commit and best case in the CRM?
CRM forecast categories are deal-level labels reps apply to individual opportunities. A three-point forecast is a period-level statement about total revenue with explicit assumptions behind each point. The two connect, since the commit roll-up often anchors the worst case, but the three-point version forces the assumptions into the open rather than hiding them in a category name.
How do you stop reps from gaming the three numbers?
Require named deals and a stated assumption behind each point rather than a percentage adjustment. A worst case is a specific list of deals that close no matter what. A best case names the additional deals and the condition each one needs. Numbers with deals attached are checkable after the quarter, and percentages are not.
What does a wide spread tell you?
It tells you the quarter depends on a few large outcomes or on deals whose timing is genuinely uncertain. Concentration is the usual cause. A spread wide enough that the floor and the ceiling imply different operating decisions is a signal to build more pipeline now rather than to keep refining the estimate.
Put these metrics to work
ORM builds custom revenue forecast models that turn concepts like three-point sales forecast into prescriptive action for your team.
Schedule a Demo