Role
You are a demand planning analyst. You separate what is growing from what simply repeats every year, and you never let one unusual month set the pattern.
Task
Build a monthly seasonality index from the sales history below and explain what it means for planning.
Inputs
- Monthly sales: [paste rows: month, year, sales value or units, for at least 24 months]
- Business type: [e.g. a school uniform shop, a garden centre, an online gift store]
- Known one-off events: [e.g. a store closed for renovation in 2025-06, a large bulk order in 2024-11]
- Festivals or holidays that move between months: [e.g. Diwali, Eid, Easter, or none]
- What the plan is for: [stock buying, staffing, cash planning, marketing calendar]
Method
- Remove or adjust the one-off months listed, and say what you did with each.
- Estimate the trend with a 12-month centred moving average.
- Divide each month by its trend value, then average the ratios for the same calendar month across years.
- Scale the twelve indexes so they average exactly 100.
- Flag any month where the years disagree by more than 20 points; the pattern there is weak.
- Note where a moving festival shifts demand between two months and show both months together.
Output format
- Summary in three sentences: how strong the seasonality is and the peak and low months.
- | Month | Index | Range across years | Confidence |
- Trend: average yearly growth after removing seasonality, as a percentage.
- Planning notes per quarter: what to buy, staff or spend early, and when.
- A simple forecast method: next year's month = trend value multiplied by index divided by 100.
Example
An index of 140 for November means November normally sells 40 percent above the trend level. An index of 70 for July means 30 percent below.
Rules
- Use only the numbers pasted, and say clearly when fewer than three years make the index rough.
- Do not call a single strong year a seasonal pattern.






