The reorder point formula, worked through on a market stall
A reorder point is the count at which you place the order. Get it right and you never run out and never sit on more cash than you need to. Here is the arithmetic, in full, with the numbers written out — and the three situations where it gives the wrong answer.
The formula
reorder point = average weekly usage × (lead time in weeks + buffer weeks) Three numbers, all of which you can find out today: Average weekly usage is how many leave your shelf in a normal week. Take the last eight weeks if you have them, fewer if the business is newer. Lead time is the days between placing the order and it arriving, divided by seven. Use what actually happens, not what the supplier's website says. Buffer weeks is how much cushion you want on top. One week is a good default. When your count reaches the reorder point, you order.
A worked example
A stallholder sells silver clasps. Over the last eight weeks they have sold 80, so average weekly usage is 10. The supplier takes 14 days, which is 2 weeks. They keep 1 week of buffer. reorder point = 10 × (2 + 1) = 30 So when the count hits 30, the order goes in. Those 30 cover the two weeks of waiting (20) with 10 spare for a busier-than-usual fortnight. How many to order is a second question, and the usual answer is: enough to cover the lead time plus the weeks after it that you want covered. Cover four weeks and you want to end up at 30 + (10 × 4) = 70, so from 30 you order 40.
Why the average has to come from movements
The formula is only as good as 'average weekly usage', and that number cannot be estimated reliably from memory. People systematically overestimate their fast lines and forget their slow ones. It has to come from a record of stock leaving — date, item, how many. Once you have that, the average is arithmetic rather than a feeling, and it updates itself as the business changes. This is the real reason to record movements: not bookkeeping, but so that the reorder arithmetic has something true to work with. One caution: make sure corrections do not count as usage. If you miscount and fix it, that fix should not read as a sale, or every mistake inflates what you think you are selling.
Three cases it gets wrong
Seasonality. An eight-week average taken in January says nothing useful about December. For a line with a strong season, either shorten the window so it reacts faster, or set the minimum by hand ahead of the season and let the formula raise it, never lower it. A supplier with a minimum order. If your supplier only ships in boxes of fifty, a reorder point of 30 and a suggested order of 40 is academic. Set the hand minimum to what a sensible order actually looks like. A brand-new line. With no history, average weekly usage is zero and the formula will never flag it. New lines need a hand-set minimum until they have a few weeks behind them.
Doing it without a calculator every week
The formula is easy. Applying it to every line every week is the boring part, and boring parts get skipped. If you are in a spreadsheet, put the reorder point in its own column and conditional-format the count to flag when it drops below. If you use a stock app, it should compute this from your own movement history and show you its working, line by line — so you can see why something is on the list and disagree with it when you know better. Either way the arithmetic is the same. There is no cleverer version of this that a one-person business needs.
Frequently asked questions
What is the difference between reorder point and safety stock?
Safety stock is the cushion; the reorder point is the cushion plus whatever you will use while waiting for delivery. In the formula here, buffer weeks is the safety stock, expressed as time rather than units so it scales as your sales change.
Should I use weeks or days?
Weeks, for a small business. Daily figures for a line that sells ten a week are noise, and thinking in weeks makes the buffer easier to reason about.
How many weeks of history should the average use?
Eight is a reasonable default: long enough to smooth out a quiet fortnight, short enough to notice a real change. Shorten it for fast-moving or seasonal lines.
What if a line has no sales at all?
Then its reorder point is zero and it will never appear on a buy-this list, which is correct. A line that has not moved in months does not need reordering — it needs deciding about.
More guides
Other things we made
Also in English, made by the same people.
Also in English, made by the same people.
Also in English, made by the same people.
Also in English, made by the same people.