Demand Forecast With Seasonality for Inventory
Estimate monthly demand to guide purchasing.
The Open buttons send the prompt text as currently shown below to that site only when you click. Fill in the placeholders first.
Forecast demand for [PRODUCT_OR_CATEGORY]. Monthly units for the last [YEARS] years: [UNITS_HISTORY]. Lead time: [LEAD_TIME]. Planned promotions: [PROMOTIONS]. Calculate a seasonal index for each month as average units for that month divided by the overall monthly average, showing the arithmetic for three months. Deseasonalise the data, estimate the trend, and rebuild a forecast for the next twelve months. Add a safety range based on past forecast errors. Return a table, ordering suggestions for the lead time, a list of data issues such as stock-outs that hide true demand, and a review schedule. Explain what happens if two years of data are not enough. Show a mini example of seasonal indices: if April averages 120 units and the overall monthly average is 100, the April index is 1.2; if August averages 70, its index is 0.7. If the deseasonalised trend for next April is 105, the forecast is 105 x 1.2 = 126. Apply to my months. Add a section on stock-outs: months where demand was higher than sales because items were unavailable should be flagged and adjusted. Include an ordering table that converts the forecast to purchase quantities after considering lead time, minimum order quantities and current stock. Provide a plan to review actual versus forecast each month and revise the indices only after at least two years of evidence. Add a safety buffer example based on the typical error. Explain why small items with irregular demand need a different method such as simple reorder points. Add a closing section on communicating uncertainty to purchasing: present a low, expected and high order quantity with the cost of being wrong in each direction, namely holding unsold stock or losing sales. Use my margin and holding cost figures if I give them, and show the arithmetic of each case. Add a short list of data to collect from now on, such as lost sales, stock-out dates and promotion calendars, because forecasts improve more from better records than from more complicated methods.
Fill in before use
Optional: type your details and the prompt above updates. Nothing is saved or sent.
Replace each bracketed placeholder with your own details, then paste into Any AI or another assistant.
See an example result (fictional data)
Invented data. April averages 120 units across three years; the overall monthly average is 100. April index = 120 / 100 = 1.2. August averages 70, index 0.7. Deseasonalised trend for next April: 105. Forecast = 105 x 1.2 = 126 units. Lead time 6 weeks, current stock 40, typical forecast error plus or minus 15 percent: range 107 to 145. Order suggestion: expected 126 - 40 = 86 units; high case 145 - 40 = 105. Flag: last August had a stock-out, so true demand may have been higher than 70.
Illustrative only, made-up data. Your results will differ and must be reviewed.