As of 13 August 2026, AI can only partly forecast demand for your product in the UK.

What to actually do

1. Hand it to a person

   The route this page recommends

   A person who owns the outcome does this end to end, worth it when the failure is dear.

2. Use a tool built for this

   Second choice

3. Do it yourself

   The distant third

   A chat interface, power-user skill, and roughly 2 hours until you can act on the result.

   How to actually do it

   1. Open your sales, orders and stock records and export at least the relevant history with dates, units or revenue, product, channel, price, promotions, returns and stockout periods.
   2. Add a separate list of UK-specific events and operational changes, including campaigns, price changes, distribution changes, bank holidays and periods when the product could not be bought.
   3. Choose the forecast horizon, time granularity and decision it will support, then paste those details and the cleaned table into the prompt.
   4. Paste the prompt into a chatbot or upload the data to Akkio, and ask it to flag missing fields, inconsistent dates and stock-constrained sales before producing a forecast.
   5. Check the backtest period and error measure against the source data, and ask for a revised forecast if the test excludes a major promotion, stockout or seasonal period.
   6. Compare the central, low and high forecasts with your current stock, supplier lead times, capacity and cash limits, then have a colleague challenge the assumptions before using the result.

   Prompt

   You are a demand-planning analyst. Forecast demand for [PRODUCT] in [UK MARKET OR CUSTOMER SEGMENT] using the data below.
   
   Forecast target: [UNITS OR REVENUE]
   Forecast horizon: [NUMBER OF WEEKS OR MONTHS]
   Time granularity: [DAILY OR WEEKLY OR MONTHLY]
   Decision this supports: [STOCK, STAFFING, PRODUCTION OR CASH PLANNING]
   
   Data:
   [PASTE A TABLE OR CSV WITH DATE, SALES, PRICE, PROMOTIONS, STOCKOUTS, RETURNS AND CHANNEL. INCLUDE ANY AVAILABLE REGION, CUSTOMER SEGMENT, LEAD TIME AND DELIVERY DATA.]
   
   Known events and changes:
   [PASTE UK HOLIDAYS, CAMPAIGNS, PRICE CHANGES, DISTRIBUTION CHANGES, COMPETITOR EVENTS, WEATHER-SENSITIVE FACTORS OR OTHER RELEVANT DATES.]
   
   Do not invent missing values, market facts or external events. First identify missing fields, inconsistent dates, stockout periods and possible data leakage. Separate observed demand from sales lost because stock was unavailable. Then:
   1. Describe the trend, seasonality and major breaks in the data.
   2. Create a baseline forecast and at least one stronger alternative only if the data supports it.
   3. Backtest the approach on earlier periods using a time-based split, and report the error measure, test period and limitations.
   4. Produce a forecast table for the requested horizon with a central estimate and a clearly labelled plausible range, explaining how the range was made.
   5. Give separate results for important channels or regions only where there is enough data.
   6. State which assumptions would change the forecast most and provide low, central and high scenarios for those assumptions.
   7. End with a short list of decisions I can make from this forecast and checks a colleague should perform before using it.
   
   Show calculations or code where practical. Do not present the forecast as a certainty, and do not use current UK market information unless I provide a source and date.

   Open it prefilled in ChatGPT or Claude, or copy it into Gemini, which takes no prefill link.