Weeks of Cover Explained: How to Calculate It and What Good Looks Like
Every inventory question a founder actually cares about reduces to one of two fears. Are we about to run out of the thing that sells? Or are we sitting on a pile of cash dressed as stock? Weeks of cover is the single metric that answers both, and most scaling brands either do not compute it at all or compute it in a way that hides the truth.
This is the full explainer: the formula, the one comparison that makes the number mean something, the mistakes that make it lie, and how to set a target that fits your business rather than a textbook.
What weeks of cover is
Weeks of cover is how long your current stock will last at the rate you are actually selling. The formula is as simple as inventory maths gets:
Weeks of cover = units on hand divided by average weekly units sold.
If you hold 1,200 units of a SKU and it sells 150 units a week, you have 8 weeks of cover. That is the whole calculation. Some tools call it weeks of supply, some express it as days of cover (multiply weeks by seven), and finance people will recognise its cousin, days sales of inventory. They are all the same idea at different resolutions.
Metric | What it measures | Convert |
|---|---|---|
Weeks of cover / weeks of supply | Stock in weeks at current run rate | The base metric |
Days of cover / days of inventory | Same, in days | Weeks x 7 |
Days sales of inventory (DSI) | Finance version, valued at cost against COGS | Directionally similar, computed from the P&L |
Inventory turns | How many times stock sells through per year | 52 divided by weeks of cover |
The arithmetic is trivial. The value is entirely in what you compare the number against, and that is where most brands stop short.
The rule that makes the number useful
A weeks of cover figure on its own tells you almost nothing. Eight weeks of cover sounds healthy. Whether it actually is depends on one thing: your true lead time.
True lead time is the full journey, production time plus transit plus receiving, not the number printed on the supplier quote. If your factory takes four weeks to make the goods and sea freight takes six weeks door to door, your true lead time is ten weeks or more. Hold that against the 8 weeks of cover and the picture inverts: the SKU that looked healthy is already out of stock, the system just does not know it yet. Unless a purchase order went in at least two weeks ago, there is a gap coming that no amount of urgency will close, which is precisely how emergency air freight gets booked.
So the working rule is this. Cover below true lead time means the stockout is already in motion. Cover comfortably above lead time plus a buffer means capital is idling. The metric is not a health score. It is a countdown clock, and lead time is the deadline it counts toward.
This comparison is the origin of a real reorder trigger, and its absence is the root cause behind most chronic stockout patterns. We covered how it fits into a full stock review in our guide to running an inventory audit, where cover versus lead time is the single most revealing cut in the whole exercise.
By SKU, never blended
The second discipline: compute it per SKU, never as one blended figure for the business. A blended average is where inventory problems go to hide. A brand holding its bestseller at two weeks of cover and a dead colourway at two years of cover can blend to a perfectly reasonable looking twelve weeks, and that single number conceals both a stockout in motion and a pile of trapped cash. The blended figure is fine for a board slide. It is useless for a buying decision.
Run the calculation across the catalogue and sort it twice. Sorted ascending, the top of the list is your restock queue, everything below lead time needs an order or an explanation. Sorted descending, the top is your cash conversation, everything holding a year or more of cover is money that already left the bank and is now paying storage for the privilege.
Setting the target: there is no universal number
Asking what a good weeks of cover figure is has no honest universal answer, because the right target is built from your own variables. It is your true lead time, plus a safety buffer sized to how volatile the SKU's demand is, bounded by how much cash you can afford to hold as stock. A stable core product on a reliable lane needs a thinner buffer than a trend item whose weekly rate swings by half.
As a practical orientation, most DTC brands land somewhere between 60 and 120 days of cover depending on lead times, with core replenishment lines toward the top of that band and fast-moving trend lines deliberately below it. But treat that as a shape, not a rule. The target that matters is the one derived from your lead times.
One consequence worth sitting with: the target moves when your freight mode moves. Shifting from air to sea stretches true lead time by weeks, which mechanically raises the cover you need to hold. That is the real trade inside the air to sea decision, cheaper freight bought with more working capital in stock, and it is why the freight saving and the planning discipline are one decision, not two. The numbers on that trade are in our air versus sea freight cost breakdown.
The mistakes that make the number lie
Five errors account for almost every misleading cover figure we see.
Stale run rates. Dividing by an average that includes weeks the SKU was out of stock understates true demand and flatters the cover figure. Average over in-stock weeks only, typically the last eight to twelve.
Ignoring in-transit stock. Goods on the water are real and belong in the picture, but label them separately from units on the shelf. Cover that only counts warehouse stock triggers panic orders for inventory that is already coming; cover that silently includes in-transit hides how exposed the next six weeks actually are.
Counting unsellable units. Damaged, mislabelled or quarantined stock in the on-hand figure is cover that cannot be sold. Count sellable units only.
Seasonal blindness. A flat average across a seasonal profile overstates cover going into peak and understates it coming out. Use a rate that reflects the demand you are heading into, not the demand you just left.
Monthly resolution. A monthly cover figure moves too slowly to catch a fast seller breaking loose. Weekly is the right cadence for a DTC catalogue, and it is no more work once the sheet is built.
Building it in a spreadsheet
You do not need software for this. One sheet, one row per SKU, seven columns: SKU, sellable units on hand, units in transit with arrival dates, average weekly units sold over recent in-stock weeks, weeks of cover, true lead time, and the gap between the two. Sort by the gap column and the sheet becomes a working reorder trigger, the exact discipline our inventory forecasting guide builds on. Refresh it weekly. The whole thing is an hour to build and ten minutes a week to run, which is a remarkable price for the two most expensive questions in the business.
Frequently asked questions
Are weeks of cover and weeks of supply the same thing?
Yes. Different tools use different labels for the same calculation. Days of inventory is the same figure multiplied by seven.
What is a good weeks of cover number?
There is no universal one. Good means comfortably above your true lead time plus a demand buffer, and not so far above it that cash is idling. The same figure can be dangerous for one SKU and wasteful for another, which is why the lead time comparison, per SKU, is the whole game.
How often should we recalculate it?
Weekly. Monthly is too slow to catch a bestseller accelerating, and daily adds noise without adding decisions.
Does stock on the water count as cover?
Count it, but show it separately with arrival dates. The question cover answers is whether you will run out before replenishment lands, and that needs both numbers visible, not merged.