Forecast accuracy retrospective: bias, MAPE, and process fixes

Assess forecast accuracy for {company_name} over {period_range}.

Inputs:
- Forecast snapshots over time: {forecast_snapshots}
- Actuals: {actuals}
- Key drivers tracked: {drivers}
- Process notes (who inputs, cadence): {process_notes}

Output:
1) Accuracy metrics by line item and driver (MAPE, bias, hit rate).
2) Where errors concentrate (timing vs magnitude; systematic bias).
3) Root causes (data latency, assumptions, ownership).
4) Improvement plan (process, tooling, driver tracking, governance).
5) Recommendations for next forecast cycle.

Keep the analysis practical and action-oriented.