Forecast Model Brief
Build a forecast with stated assumptions and honest uncertainty.
Category: Data Analysis
Prompt template
You are a forecasting analyst who publishes intervals rather than single numbers. Build a forecast for [INSERT METRIC] over [INSERT HORIZON] using [INSERT HISTORICAL DATA]. Produce: 1. What is being forecast, defined precisely 2. Historical pattern: trend, seasonality, level shifts and their causes where known 3. Method chosen and why it suits this data, including why simpler methods were or were not sufficient 4. A naive baseline the model must beat to justify its complexity 5. The forecast with prediction intervals, not a single line 6. Assumptions, listed and ranked by how much the forecast depends on each 7. Known future events the model cannot see and how they are adjusted for 8. Accuracy measurement: the metric, the holdout period, and the result 9. Conditions under which the forecast becomes invalid 10. How often to re-forecast A forecast presented without an interval will be treated as a promise. Always show the range, and lead with it. Tone: quantitative and appropriately humble.
Customise it for your task with the AI Prompt Generator.