Retail Reorder Points: The Formula and the Setup Checklist

Two problems, one cause

Most retailers have both of these complaints at once and treat them as separate:

• We keep running out of the things that sell.
• We have too much money tied up in stock that does not move.

They are the same mistake seen from two ends. Both come from a single low-stock number applied to every product in the shop.

Set the shop-wide alert at five and your fast movers are gone days before anyone notices, while your slow movers alert constantly and get reordered for no reason. Set it at fifty and the reverse happens. There is no single value that works, because the correct number depends on how fast an item sells and how long your supplier takes — and those vary enormously across your catalogue.

The fix is arithmetic, not software, and it takes an afternoon.

---

The base formula

Reorder point = average daily sales × supplier lead time in days

This is the amount you will sell while waiting for the delivery. Order at that level and, if everything goes to plan, the new stock lands as the last unit sells.

Worked: a drink selling 10 a day from a supplier who delivers in 3 days.

10 × 3 = 30

Order when stock hits 30. Compare that to the shop-wide alert of 5 that most systems ship with, and you can see why the fast movers are always the ones that run out.

---

The refinement that actually prevents stockouts

The base formula assumes everything goes to plan. Stockouts happen specifically when it does not.

Almost all of the risk sits in one variable, and it is not demand — it is lead time. Demand varies a bit day to day and averages out. A delivery that was supposed to arrive Tuesday and comes the following Monday empties your shelf regardless of how well you estimated daily sales.

So buffer the thing that actually breaks:

Safety stock = (worst lead time − usual lead time) × average daily sales

Which collapses into something you can hold in your head:

Reorder point = average daily sales × worst realistic lead time

Back to the drink. The supplier usually takes 3 days but has taken 6 more than once.

10 × 6 = 60

Sixty, not thirty. The number doubled — not because demand changed, but because the supplier is unreliable. That is the honest cost of that relationship, and seeing it as a number is more useful than a vague sense that they are sometimes late. It also gives you something concrete to raise with them: reliable delivery in three days would free up thirty units of capital on this line alone.

---

Getting the two inputs right

Average daily sales

Units sold over the last 60 days, divided by 60.

Sixty days smooths out a quiet week without going stale. Two cautions:

• Do not use a twelve-month average for seasonal goods. It is wrong in both directions — too low exactly when you need stock, too high afterwards.
• Do not use a period containing a stockout. If you were out of stock for a week, your recorded sales understate real demand, and the reorder point you calculate from it will keep you understocked. Adjust upward or use a clean period.

Lead time

Not what the supplier says. What you have actually observed, at its worst.

Write down the last five orders for each significant supplier: date ordered, date it actually arrived. That short exercise usually produces a surprise, because the promised figure and the observed range are often far apart. Include the parts people forget: the day it sat in the delivery bay before being received into stock, and the weekend it spent in transit.

---

Where slow movers break the rule

Everything above assumes holding stock is cheaper than running out. For expensive, slow-moving items that is often false.

An appliance selling twice a month, with a 10-day worst lead time:

0.07 × 10 = 0.7, so a reorder point of 1.

Reorder when the last one sells. Some of these should not be stocked at all — order on demand, quote the customer a realistic wait, and keep the capital. The failure mode here is applying fast-mover instincts to slow-moving stock and holding two months of an expensive item for a service benefit nobody asked for.

The general principle: as unit cost rises and turnover falls, the correct reorder point trends toward zero. Our post on dead stock covers how to find the lines where this has already gone wrong.

---

The setup checklist

Once, per product group:

• Pull units sold over the last 60 days for each product
• Divide by 60 for average daily sales
• List your suppliers and the observed worst lead time for each
• Multiply: daily sales × worst lead time
• Round to something practical — a case, a pack, a sensible number
• Set the value on the product, not globally
• Sanity-check the extremes: does the fastest mover look high and the expensive slow mover look near one? If not, recheck your inputs

Do the top 20% of products by sales value first. They cause most of your stockouts and most of your tied-up cash, and you can finish them in an afternoon. The long tail can inherit a rough default until you get to it — an imperfect number on a slow mover costs very little, which is precisely why it is not urgent.

Ongoing:

• Review quarterly, and at every seasonal turn
• Review immediately after any stockout — it is free evidence the number was wrong
• Review immediately after changing supplier
• Recheck lead times once a year; they drift, usually upward

---

Reorder point is not order quantity

These get conflated constantly and they answer different questions.

	Reorder point	Order quantity
Question	When do I order?	How much do I order?
Driven by	Demand rate and lead time	Supplier minimums, price breaks, cash, shelf space
Wrong answer causes	Stockouts	Overstock and dead capital
Review trigger	Stockout, supplier change, season	Cash position, price change, storage

A perfect reorder point with a careless order quantity still produces the overstock problem. Decide them separately, and be especially sceptical of the supplier discount that requires tripling your order — a price break that ties up three months of cash on a line selling steadily is rarely the saving it appears to be. Our cash flow guide covers that trade-off in more detail.

---

When the formula gives an unaffordable answer

Sometimes the arithmetic says hold 120 units and you cannot afford 120 units. The formula has not failed; it has told you your cash cannot support that supplier's lead time.

Three real options:

1. Shorten or stabilise the lead time. Negotiate, or change supplier. This attacks the actual cause and reduces the requirement rather than the safety margin.
2. Find a local emergency supplier. More expensive per unit, used rarely, and it lets you plan around the cheap slow supplier for the bulk. Effectively you are buying a shorter worst-case lead time only when you need it.
3. Accept planned stockouts on that line. Decide it consciously, tell staff what to say to customers, and stop treating each occurrence as a crisis.

What does not work is quietly setting a lower number and hoping. That turns a known constraint into a recurring emergency, and emergency restocking almost always costs more per unit than the stock you could not afford to hold.

---

Setting these up in Zeneva

The threshold is a field on each product rather than a single global setting, which is what makes per-product numbers possible. Set it when you add the product and revise it at your quarterly review.

Two things worth knowing if you use Zen AI: it can surface which products are below their threshold and which are trending toward a stockout, and it can suggest a threshold value for a product from its actual sales history. Suggestions arrive as a proposal you approve or reject rather than a change it makes on your behalf — you remain the one deciding, which is the correct arrangement for a number this consequential. Our post on what Zen AI does and does not do is candid about where that boundary sits.

For the forecasting layer on top of this — which lines are accelerating, and what to buy ahead of a season — our guide to demand forecasting picks up where reorder points leave off.

Per-product thresholds, rather than one number for the whole shop, are available on every plan including the free one — compare what each includes.