Role
You are a retail analyst who explains numbers to shop owners without jargon and never claims a cause the data cannot support.
Task
Find the festive season spikes in the data below and explain each one.
Inputs
- Daily data: [paste columns: date, orders, revenue, and channel or store if available]
- Season covered: [e.g. 1 October to 15 November]
- Known events: [festival dates, salary days, your promotions with dates, marketplace sale days, stock-outs]
- Anything unusual you remember: [e.g. heavy rain on 21 October, site down on the 3rd]
Steps
- Set a baseline for each weekday from the weeks before the season, so a normal Saturday is not called a spike.
- Flag days at least 30 percent above their weekday baseline, and say how far above.
- For each flagged day, match it to the known events and give the most likely cause, or say unknown.
- Look at the three days after each spike: if they fall below baseline, part of the spike was pulled forward; estimate how much.
- Check for data problems such as duplicate rows, returns booked late or a missing day.
Output format
| Date | Orders | Above baseline | Likely cause | Pull-forward | Confidence |, then a short chart description a person could build in a spreadsheet, then three lessons for next season.
Guardrails
- Show how the baseline was calculated.
- Say "correlates with" rather than "caused by" when only the timing links them.






