Demand Forecast

How much stock to order in PrestaShop: sales forecast, safety stock and reorder point

Do your products sell out right at Christmas while others sit in the warehouse for months? Without complicated formulas, we explain how to forecast your PrestaShop shop's sales, how much safety stock you need and on which date to order from each supplier so you never run out.

The problem: ordering by gut feeling

Most online shops decide what to order by looking at last month's sales and the stock left. It works until a season arrives: the product that sold 10 units a week sells 40 in December, the supplier takes three weeks and the stockout comes just when sales peak. Meanwhile, other products pile up months of stock and tied-up money.

To order well you need three answers per product: how much will I sell, how many days will my stock last and by which date must I order.

1. Forecast sales, don't repeat last month's

A good sales forecast starts from PrestaShop's order history, but cleans and interprets it:

  • Remove out-of-stock days. If a product was sold out for two weeks, zero sales those days doesn't mean nobody wanted it. Counting them underestimates demand and causes the next stockout.
  • Separate trend and season. A product can grow month after month and also sell three times more in summer. Both need measuring separately.
  • Treat irregular sellers differently. A spare part sold once a month needs a different method from a daily best seller.
  • Learn from your campaigns. Black Friday, sales or Mother's Day move some products and not others. Previous years tell you how much.
  • Give new products a reference. Without history, start from their category or a similar product and correct with the first sales.

2. Cover: how many days your stock lasts

Cover is available stock (plus what's already on the way) divided by expected daily sales. With 60 units and 4 a day expected, you have 15 days of cover. Careful: with a seasonal forecast those 15 days can become 8 once December starts.

3. Safety stock: the cushion for the unexpected

No forecast is exact. Safety stock covers the weeks you sell more than expected while waiting for the supplier. It depends on two things:

  • How variable the demand is: the more irregular, the bigger the cushion.
  • The service level you want: covering 95 % of situations needs much less stock than covering 99 %.

A common rule is safety stock = z × demand deviation × √(supplier lead time), where z is 1.65 for a 95 % service level.

4. Reorder point and order-by date

The reorder point is when the remaining stock only covers the lead time plus the safety stock. In dates: order-by date = run-out date − supplier lead time − safety days. If that date has passed, you are late; if it falls in the next few days, it's time to order.

StatusMeaningWhat to do
Too lateAn order today arrives after the stockoutOrder now and find an urgent alternative
Buy nowThe order-by date is this weekAdd it to the next order
WatchThe order-by date is getting closeCheck again in a few days
OverstockStock for many monthsDon't order; consider a promotion

5. How much to order

The quantity is the expected demand during the supplier's lead time and the cover you want, plus safety stock, minus current stock and stock on the way, rounded up to the pack size. Grouped by supplier, that's your purchase list.

6. Repeat customers count too

In consumables shops (coffee, pet food, cosmetics, spare parts) part of the sales come from customers who buy the same thing every few weeks. Knowing that rhythm lets you add those expected purchases to demand and remind customers to come back just before they run out, instead of losing them to another shop.

How to do it in PrestaShop without spreadsheets

PrestaShop stores orders and stock, but it doesn't calculate forecasts or order dates. You can export to a spreadsheet, but keeping it up to date per product and combination takes hours every week. Demand Forecast, our module, does it all inside your shop: forecast with seasons and campaigns, cover, safety stock, order-by date per supplier and a PDF purchase list, plus reminders and scheduled purchases for your regular customers. No external services or fees.

Frequently asked questions

How much history do I need to forecast sales?

A few months give you the trend; to measure seasonality you need at least a year, and two for it to be reliable.

Which service level should I choose?

95 % is a good starting point. For star or very profitable products it can pay to go up to 98 %; for low-margin products, down to 90 %.

What if a product has many combinations?

Forecast each combination separately: size M and XL don't sell the same. If each one sells very little, group them and split afterwards.