Weeks of Cover: The Simplest Inventory Metric You're Probably Using Wrong

Weeks of cover is one of the first metrics anyone learns in inventory planning, and one of the most commonly misused. The calculation is simple. The judgement around it is where most people go wrong.

The calculation

Weeks of Cover = Closing Inventory ÷ Average Weekly Forward Demand

If you're holding 800 units and expect to sell 200 a week, you have 4 weeks of cover. Straightforward.

The first mistake: using historical demand instead of forward demand

A lot of weeks of cover calculations quietly use average demand from the past few weeks, because that's the number that's easiest to grab. That's fine if demand is stable. It's actively misleading if demand is about to change.

If a SKU sold well historically but is heading into a slow season, using historical average demand will overstate how many weeks of cover you actually have — the inventory will run out faster than the metric suggests. Always use the best available forward-looking demand number, not a rearview mirror.

The second mistake: treating one target as right for everything

"We want 4 weeks of cover" sounds like a sensible, simple policy. It usually isn't, once you look SKU by SKU.

A fast-moving, high-volume, low-margin staple can often run comfortably on much less cover, because replenishment is frequent and demand is predictable. A slow-moving, high-margin, long-lead-time specialty item often needs considerably more cover, because you can't quickly top it up if it starts running low.

A flat target across your whole range means you're simultaneously over-stocking some SKUs and under-protecting others. Segmenting cover targets — by demand volatility, lead time, and commercial importance — is a small amount of extra work for a meaningfully better outcome than one number applied everywhere.

What weeks of cover doesn't tell you

It tells you how long your current stock will last at expected demand. It doesn't tell you:

Whether that's the right amount. High cover isn't automatically bad (it might be appropriate for a long-lead-time item) and low cover isn't automatically risky (it might be fine for something you replenish weekly). Cover needs a target to be meaningful, not just a number on its own.

Anything about demand shape. Two SKUs can both show "4 weeks of cover" while one has steady, predictable demand and the other has occasional huge spikes. The second one is at genuinely higher risk of a stockout despite an identical headline number.

Freshness or expiry risk. For anything perishable, a technically "correct" weeks of cover figure can still represent stock that's going to expire before it sells, particularly if demand is lumpy rather than steady.

Inventory turns: the flip side of the same coin

Inventory turns (how many times your inventory is sold and replaced over a year) is mathematically related to weeks of cover — roughly, turns ≈ 52 ÷ weeks of cover. They're two views of the same underlying efficiency, and businesses tend to favour whichever one their leadership is more used to talking in. Neither is more "correct"; they're just different framings of the same number.

A more useful way to use weeks of cover

Rather than a single flat target, use it as a triage tool: set a low-cover threshold (risk of stockout, worth investigating) and a high-cover threshold (risk of excess or expiry, also worth investigating), segmented by SKU category. Then let weeks of cover do what it's actually good at — quickly surfacing which SKUs, out of hundreds, deserve a planner's attention this week — rather than treating the number itself as the goal.

Next: How to set inventory targets by SKU — or see weeks of cover calculated across a real product range in the Nerd Foods dataset.