How to Calculate Safety Stock: Formulas and a Worked Example

Run out of one component and the whole batch waits, however much of everything else is on the shelf. Safety stock is the buffer that stops that: extra inventory held to cover a late delivery or a busier-than-usual month.

There are three common ways to calculate it, from a one-line formula to the statistical method. For a small manufacturer, we’d start with the material that would stop a batch, and size its buffer from your own delivery history rather than a textbook service level.

Here are the formulas, a worked example with the arithmetic shown, and how to decide which materials deserve a buffer at all.

What is safety stock?

Safety stock is the extra inventory you hold to cover late deliveries or unexpected demand, so a delay doesn’t stop production or sales.

It sits on top of the stock you expect to use. If a supplier ships a week late or an order doubles overnight, you draw on the buffer instead of stopping work.

If you make what you sell, it applies to three kinds of stock: raw materials, packaging and finished goods. Raw materials and packaging matter most, because a missing component stops a batch and everything you’d have made from it.

The buffer isn’t free. Stock you hold is cash you can’t spend and, for some materials, shelf life you’re using up. The aim is the right amount, not the most.

How do you calculate safety stock?

Multiply maximum daily usage by your longest lead time, then subtract average daily usage times average lead time. The result is your safety stock.

That’s the maximum-minus-average method, and it’s the one we’d use for most small-batch manufacturers. Two other formulas are common. Which one fits depends on how much usage and delivery history you have.

Method 1: the basic formula

Pick the number of days you want covered and multiply it by what you use on an average day. It’s fast, and it works when demand is steady and your supplier is reliable. For a stable supplier, a few days to a week is a common starting point.

The weakness is that the buffer days are a guess. Nothing in the formula checks them against how late your deliveries really run.

Method 2: maximum minus average

This sizes the buffer to the gap between your worst realistic case and your normal one. It needs only numbers you can read off your own records, which is why it suits small manufacturers. You’ll also see it called the min-max or average-max formula.

Use your busiest week for maximum usage, not your busiest single day: the average daily usage across that week. For lead time, use the longest delivery in your last six to twelve orders of that material.

If your lead time never varies, the formula collapses to (maximum daily usage − average daily usage) × lead time, which is the version some guides print.

Method 3: the statistical method

Standard deviation measures how far your daily usage, or your delivery times, typically stray from their average, and variance is that figure squared. Z comes from the service level you choose, which is the chance of not running out while you wait for a delivery.

This is the textbook answer, and it’s the better fit when demand is steady, volumes are high and you have a year or more of clean usage and delivery records. If that’s you, use it.

Most small-batch manufacturers don’t have that history for every material. A formula that needs a standard deviation you’ve never measured gives you false precision, which is why we’d start from the delivery times you can see in your own purchase records.

Which formula should you use?

A worked example: glass bottles from an overseas supplier

Say you run a small-batch food business and fill about 40 glass bottles a day. They come from an overseas supplier and arrive in 14 days on average. In your busiest week you fill 60 a day, and the slowest of your last twelve deliveries took 21 days. These figures are illustrative:

Here’s the maximum-minus-average calculation:

That’s 17.5 days of average usage. At an illustrative $0.60 a bottle, it’s $420 of cash on the shelf.

Add it to your reorder point and the numbers line up. Reorder point = (40 × 14) + 700 = 560 + 700 = 1,260 bottles. That equals peak usage over your longest lead time (60 × 21), so in the worst case on both counts the delivery arrives just as the last bottle goes.

Now run the other two methods on the same material.

Basic formula. A week’s buffer is 40 × 7 = 280 bottles, which makes the reorder point 560 + 280 = 840. In the worst case you’d use 1,260 bottles before the delivery lands, so you’d be 420 short.

Statistical method. Say daily usage varies by a standard deviation of 8 bottles and delivery time by 3 days (also illustrative). Squared, those are variances of 64 and 9. Then (14 × 64) + (40² × 9) = 896 + 14,400 = 15,296, and the square root of that is about 123.7. At a 95% service level (Z = 1.65) the buffer is about 204 bottles. At 99% (Z = 2.33) it’s about 288.

Those two figures sit well below 700 because the maximum-minus-average method assumes your busiest week and your slowest delivery land together. That’s the cautious assumption, and it’s what you want for a bottle that stops every batch.

It also means the number is your choice. At $0.60 a bottle, 204 bottles is about $122 of stock and 700 is $420. Pick a point between them on purpose, instead of finding out what you chose when a delivery slips. For long-lead or single-source materials, our reorder point guide suggests 10 to 14 days of consumption, which here is 400 to 560 bottles.

Which materials need safety stock first?

Start with materials that would stop a batch if they ran out: single-source, imported, long lead time or high minimum order quantities.

One missing bottle holds up the whole batch that needed it, so a buffer on a part with no fallback is cheap insurance. A buffer earns its place on:

• Single-source materials with no substitute you’ve approved.
• Imported materials, where delivery times stretch and swing.
• Printed or made-to-order packaging, which can’t be topped up in a day.
• Anything a late delivery has already held up once.

When not to hold safety stock

Skip the buffer, or keep it small, when:

• You can restock it quickly. If a local supplier can deliver tomorrow, the cost of running out is a short delay, not a stopped line.
• It has a short shelf life. A buffer of perishable ingredients can expire before you use it. A second supplier does more for you than extra stock.
• You can substitute it without changing the product.
• You make to order. If customers expect a wait, a buffer of finished goods is stock you didn’t need.

For finished goods, hold a buffer only on the lines customers expect to find in stock and that you can’t remake quickly.

Safety stock and your reorder point

Safety stock and the reorder point work together. Your reorder point is the stock level that triggers a new order: average daily usage × average lead time, plus safety stock. The safety stock is the part that covers the unexpected.

Our guide to the reorder point formula walks through the rest.

Where your inputs come from

The formulas are short. Getting trustworthy inputs is the hard part.

• Usage comes from your production records: how much of the material you used in a period, divided by the days in it.
• Lead time comes from your purchasing records: the days between placing an order and having the material on the shelf.

Pull the last six to twelve orders for each material. The average and the longest are your two lead time inputs. Recalculate when a supplier changes or a delivery pattern shifts, rather than on a calendar. For the cost side of running out, the stockout cost calculator puts a number on it.

Keeping the numbers current in Stocksmith

Stocksmith records consumption at the batch level. Each time you record a batch, the materials in its recipe come off your stock, so your usage comes from your own production records instead of a rebuilt spreadsheet.

Stocksmith lets you set a low-stock level on each material and product. Set it at your reorder point, safety stock included, and you’ll see a material approaching it before a batch is queued, instead of finding out when it has already stopped.

Frequently Asked Questions

Start with the part that would stop a batch

Pick the one material that would stop your next batch if it didn’t arrive. Pull its last six to twelve orders, find the average and longest lead time, and run the maximum-minus-average formula once. You’ll have a number to hold, and you’ll know what it costs.

Do the same for the next material when you reorder it. Start a free 14-day trial and keep the usage and the reorder level in one place.