Create intermediate Works with: Any AI

Forecast Accuracy Review

Table output Has example

Measure how wrong past forecasts were and what to change.

The Open buttons send the prompt text as currently shown below to that site only when you click. Fill in the placeholders first.

Review forecast accuracy. Forecasts and actuals by period: [FORECASTS_AND_ACTUALS]. How forecasts were made: [METHOD]. Decisions they drove: [DECISIONS].

Calculate error for each period as actual minus forecast, absolute percentage error, the average of those, and bias as the average signed error. Show the arithmetic for three periods. Identify periods with large errors and possible causes.

Recommend changes: method, inputs, review cadence, and adding ranges. Return a table, a bias interpretation and a plan to track accuracy each month. Warn about percentage errors being unstable when actuals are small.

Add a worked example: forecasts 100, 120, 90; actuals 110, 100, 95. Errors (actual minus forecast): +10, -20, +5. Absolute percentage errors: 9.1 percent, 20.0 percent, 5.3 percent; mean = 11.5 percent. Bias = (10 - 20 + 5) / 3 = -1.7 units, a slight over-forecast. Replace with my data and interpret in plain words. Suggest a rule: if bias is consistently negative, the process is optimistic and needs adjustment.

Fill in before use

Optional: type your details and the prompt above updates. Nothing is saved or sent.

Replace each bracketed placeholder with your own details, then paste into Any AI or another assistant.

Invented data. Forecasts 100, 120, 90; actuals 110, 100, 95.
Errors (actual minus forecast): +10, -20, +5.
Absolute percentage errors: 10/110 = 9.1 percent; 20/100 = 20.0 percent; 5/95 = 5.3 percent. Mean = 11.5 percent.
Bias = (10 - 20 + 5) / 3 = minus 1.7 units: slight over-forecasting.
Largest miss: period two; the cause given was a cancelled promotion.
Change: record promotions in the forecast inputs and report a range.
Next step: track absolute percentage error monthly and show the forecast as a range, for example plus or minus 12 percent, so decisions allow for typical misses.

• forecasting