Weeks of cover is not one number, and treating it like one costs you

Almost every brand runs a weeks-of-cover target, and almost every brand runs the wrong one, because they run one. A single number, say eight weeks of cover, gets set as the goal and applied across the assortment. Buyers chase it, the reorder engine defends it, and everyone treats the number like a law of physics instead of the crude average it actually is.

Weeks of cover is a ratio: how long your current stock lasts at the current rate of sale. The moment you apply one target across categories that sell at wildly different speeds and replenish on wildly different lead times, you guarantee two failures at once. The fast movers on long lead times stock out, because eight weeks of cover was never enough runway to reorder in time. The slow movers on short lead times drown in inventory, because eight weeks of cover was months of stock they will never turn.

A good weeks-of-cover target is not a number. It is a function of how fast the category sells and how long it takes to replace, and the brands with tight inventory and no dead stock are the ones who set it that way instead of picking one figure and defending it everywhere.

Think about what weeks of cover actually has to absorb. Its whole job is to keep you in stock from now until the next delivery lands. So the right amount of cover depends on exactly two things: how fast you are selling through it, and how long the replacement takes to arrive. A blanket target ignores both, which is like setting the same fuel gauge warning for a car and a plane.

Take the fast side. A hero category selling briskly on a twelve-week overseas lead time needs enough cover to survive the entire reorder cycle plus a safety buffer, or it stocks out before the replenishment ships. Eight weeks of cover on a twelve-week lead time is a stockout with a countdown timer attached. You will run dry four weeks before the new stock arrives, every single time, and blame the buyer.

Now the slow side. A steady basics category on a two-week domestic lead time can be reordered almost on demand, so it needs very little cover. Force eight weeks of cover onto it and you are holding four times the stock the lead time requires, tying up cash in inventory that turns slowly and will be there long after you needed it. The blanket target does not just misjudge these categories; it misjudges them in opposite directions simultaneously.

That capital is not freed by taking risk on the fast movers. It is freed by right-sizing the cover on the slow ones, where a blanket target had you holding months of stock you could have reordered in a fortnight. The fast movers actually get more cover under this approach, not less, because now the target matches their lead time.

It is worth being precise about why the blanket target feels safe when it is not. A single high cover number looks conservative, because for any given category it usually errs toward more stock rather than less, and more stock feels like less risk. But that intuition only holds category by category. Across the whole assortment, a blanket target that is generous enough to protect the long-lead-time heroes is wildly over-generous for everything short-lead-time, so the aggregate effect is not caution, it is a mountain of dead cash sitting in categories that never needed it. What looks prudent one line at a time is the opposite in total.

Here is how you set it properly. The right cover target for any category is roughly its lead time plus a safety buffer sized to how variable its demand is. Fast, predictable categories on short lead times need little. Fast, volatile categories on long lead times need a lot. Work three cases and it becomes obvious.

Case one: core basics, two-week domestic lead time, steady demand. You need enough to cover the two-week reorder cycle plus a small buffer for demand wobble, so call it three to four weeks of cover. Anything above that is dead cash. Case two: a hero seasonal category, twelve-week import lead time, brisk and slightly volatile demand. You need to cover the full twelve-week cycle plus a real safety buffer, so fourteen to sixteen weeks of cover is correct, twice the blanket target, because the lead time demands it. Case three: a long-tail category, six-week lead time, slow and lumpy demand. The lead time says six weeks but the lumpiness says add buffer, so eight to ten weeks, and you watch it closely because lumpy demand breaks simple cover math.

Three categories, three completely different correct answers: four weeks, fifteen weeks, nine weeks. The blanket eight-week target is wrong for all three. It over-covers the basics by double, under-covers the hero by half, and roughly lands on the long-tail by accident. This is not a rounding error. It is the reason you simultaneously stock out of the things that sell and drown in the things that do not.

Lay them side by side and the blanket target has nowhere sensible to sit. Pick eight and you are wrong three ways. The only correct target is the one derived per category from its own lead time and its own demand, which is precisely the calculation nobody has time to redo by hand every week.

The move is to stop treating weeks of cover as a target you set and start treating it as an output the plan computes. For each category, the platform reads the live sell-through rate, the actual supplier lead time, and the demand variability, and derives the cover target that keeps you in stock through the reorder cycle without holding a week more than that. Fast categories get the runway their lead times demand; slow ones stop hoarding cash.

Because it is computed from the live signal, it moves when the business moves. When a category accelerates, its required cover rises and the reorder fires earlier, before the faster sell-through outruns the old target. When a category cools, the required cover falls and the buy pulls back before the excess piles up. The agent stages the reorders against the per-category cover, and the buyer approves the set instead of defending a single number that was wrong for most of the range from the day it was set.

This is also where lead time stops being a static assumption and becomes a live input. Supplier lead times drift: a factory slips, a lane congests, a mode changes from sea to air and back. If your cover target is pinned to a lead time you typed in a year ago, it goes stale the moment reality moves, and you find out at the stockout. Computing cover from the actual, current lead time means the target tracks the supply chain you have, not the one you assumed, so when a lane slows down the required cover rises and the reorder pulls forward on its own, before the slower replenishment turns into an empty shelf.

Tight inventory and full shelves are not opposites. They are what you get when cover follows lead time and demand, instead of one number following everything.

Set cover per category from demand and lead time and the two chronic failures resolve at once: the fast movers on long lead times stop stocking out, and the slow movers on short lead times stop turning into dead stock. Same discipline, better math, and the working capital that was trapped in over-cover comes back to fund the next buy.