Setting Reorder Points: A Simple Framework to Avoid Stockouts

"Reorder when it looks low" isn't a reorder point, it's a guess that works until the one week it doesn't. Here's the actual math, and how to keep it honest over time.

The basic formula

A reorder point answers one question: at what quantity on hand should you place a new order, so the new stock arrives before you run out? The standard formula:

Reorder point = (average daily usage × lead time in days) + safety stock

Each piece matters:

• Average daily usage — how many units you typically use or sell per day. This should come from actual historical data, not a guess, and should be revisited periodically since demand shifts over time.
• Lead time — how many days between placing an order and having it available to use, including your supplier's processing time and any receiving/putaway delay on your end, not just shipping transit time.
• Safety stock — a buffer on top of the bare minimum, to absorb the fact that both usage and lead time vary in the real world instead of being perfectly constant.

Why safety stock isn't optional padding

The first two terms of the formula describe the average case. But averages hide variation — some weeks you use more than average, some supplier deliveries run late. Without safety stock, a reorder point calculated purely from averages will stock out roughly half the time, by definition, since "average" usage is exceeded about as often as it isn't. Safety stock is what accounts for that variability instead of assuming every week is a perfectly typical one.

How much safety stock to hold depends on how variable your demand and lead times actually are, and how costly a stockout would be for that specific item — a critical component with unreliable suppliers justifies more buffer than a stable, easily-resupplied commodity item.

A worked example

Say an item sells an average of 10 units a day, your supplier takes 7 days from order to delivery, and you want a 3-day safety buffer on top of that:

Reorder point = (10 × 7) + (10 × 3) = 100 units

In plain terms: when stock on hand drops to 100 units, place the next order. That covers expected usage during the 7-day lead time (70 units) plus a 3-day cushion (30 units) in case usage runs a bit high or the order arrives a bit late.

Reorder points aren't set-and-forget

The most common mistake isn't getting the initial calculation wrong — it's calculating it once and never revisiting it. Usage patterns shift with seasonality, growth, or a product's changing popularity. Supplier lead times change too, sometimes without much warning. A reorder point that was correct six months ago can quietly become wrong in either direction: too low (stockouts start happening) or too high (you're carrying more inventory, and more carrying cost, than you actually need).

A reasonable habit: review reorder points on a regular schedule (quarterly is common), and always recalculate immediately after a known lead-time change from a supplier, rather than waiting for the scheduled review to catch it.

This only works if the usage number is real

Everything above depends on accurate historical usage data. If your inventory records have gaps, or usage is tracked inconsistently across locations, the "average daily usage" you're calculating from isn't describing reality — and a reorder point built on bad input data will confidently give you the wrong number.