Fill in these first
- Business and channel: [e.g. sweet shop, online fashion store]
- Last year's daily sales for the festive window: [paste date, sales value, orders]
- Last year's festival dates: [e.g. Diwali on 1 November 2025]
- This year's festival dates: [e.g. Diwali on 8 November 2026]
- Growth over the last three months versus last year: [percent]
- Known changes this year: [new store, bigger ad budget, price rise, none]
Task
Forecast daily sales for this year's festive window, lining up days by their distance from the festival rather than by calendar date.
Method
- Re-index last year's days as festival day minus N and plus N.
- Baseline = last year's matching day x (1 + growth).
- Adjust for the known changes, stating the assumption for each.
- Give a low, expected and high figure using a range that widens on the peak days.
Output format
| Date | Day vs festival | Last year matching day | Baseline | Adjustment | Low | Expected | High |
Then:
- Total expected sales for the window and the three busiest days.
- Staffing and stock notes for the peak days.
- Two warning signs that the forecast is running too high or too low.
Rules
- Show the working for one row.
- Never invent missing days; mark them and fill from the nearest days, saying so.
- Treat the output as an estimate to check against live sales each week.






