A Sales Forecast Needs a Confidence Explanation, Not Only a Percentage
A percentage can make a forecast look precise while leaving its basis unclear. When reviewing an opportunity, ask what the confidence statement means, which evidence supports it and what could change it. A number without that context is difficult to challenge constructively.
Define the kind of estimate being discussed
Distinguish a seller’s judgment, a team convention and a model-produced output. They may use the same percentage while meaning different things. Record the time horizon and the event being estimated. Do not present a personal confidence rating as a calibrated probability or imply that a model’s output guarantees a purchase.
Identify evidence and missing inputs
List the few facts that materially support the assessment and the uncertainties that could alter it. Separate customer-confirmed information from internal assumptions. NIST’s AI risk-management work is voluntary guidance, not certification of a particular sales product. A team using automated assistance still needs to understand the limitations of the actual system and the information supplied.
Set a review trigger
Name the change that would justify revisiting the forecast, such as a confirmed process step or an unresolved dependency being answered. Preserve the earlier estimate and its date rather than silently replacing the history. The aim is a better review conversation and a clearer record, not more pressure on the buyer or a claim that every opportunity should follow the same path.
Copyable worksheet
Use one row for each field. Mark unknown information clearly and retain the original records.
| Field | Your note |
|---|---|
| Estimated event and time horizon | ____________________ |
| Judgment, convention or model output | ____________________ |
| Material supporting evidence | ____________________ |
| Unknowns affecting confidence | ____________________ |
| Review trigger and current date | ____________________ |
Worked example
Illustrative example: a fictional team gives an opportunity a high confidence percentage but cannot state the event or time horizon. During review it records those definitions and an unresolved procurement dependency. The percentage becomes easier to interpret without turning uncertainty into a guaranteed sales outcome.