Weeks of Supply

Managers often know stock in units but need to know how long it might last, and weeks of supply converts inventory into a time measure that depends heavily on the demand rate and which stock is counted. The numerator should be clear, since on-hand, available-to-promise or on-hand plus confirmed inbound stock are different choices and quarantined, reserved and damaged items may not be usable, so label the basis.

Accelerated Analytics describes weeks of supply as relating inventory to sales rate and Oracle planning documentation shows that average demand depends on the selected time horizon and data, which explains why two valid reports can differ. The denominator might be average historical sales, forecast demand or planned consumption, and it should use units per week for the same item and location as the stock, because a mismatch produces a meaningless result.

A fictional planner uses recent sales for a steady product and a forecast for a seasonal launch, labelling each method, since a basic average works best when demand is fairly stable while seasonal peaks, promotions and new products need a forward-looking view and a six-week historical average may understate tomorrow's surge. When weekly demand is zero the simple ratio is undefined and does not mean stock will last forever, so investigate whether the product is inactive, new or missing sales data; a fictional new item with inventory but no sales history uses a supported forecast, not division by zero.

Lead time matters: four weeks of stock is not safe if replenishment takes eight weeks, while two weeks may be fine with reliable next-day supply, so compare cover to the full replenishment process. A fictional importer with five weeks of stock and a seven-week supplier lead time starts a shortage plan despite a seemingly healthy number.

Safety stock protects against uncertainty, and a ratio that counts all inventory may hide that some units are reserved as a buffer, so state whether the buffer is included. Confirmed inbound shipments can extend cover, but arrival timing matters, since adding goods due in week five cannot prevent a shortage in week three, so use a week-by-week projection for material decisions.

A fictional supplier promises 300 units next month, and the buyer checks whether current stock lasts until then. Returns and backorders can distort the rate, because historical sales may be limited by previous stockouts while unfilled orders represent demand not captured as sales, so clean the denominator; a fictional shop that sold only 50 units last week because shelves were empty does not assume true demand halved.

Measure by item and location, since a national total may look adequate while one branch is empty and transfers have lead time and cost too; a fictional retailer with ten weeks of stock overall but only one week at a busy branch plans a transfer. Weeks of supply is not a target by itself, because perishables, fashion and spare parts have different costs of holding and running out, so set thresholds by service goals and risk, as a fictional food business prefers lower cover for short-life goods than for a hard-to-source replacement part.

Check inventory accuracy, since an incorrect count can make the ratio precise-looking but wrong, so use cycle counts and reconcile discrepancies. Report the date, stock basis and demand horizon beside the number, because trends are most useful when the method stays consistent and a change from historical sales to forecast demand should be disclosed.

A more detailed forecast subtracts expected demand and adds arrivals week by week, which reveals the first possible gap and guides reorder timing, and a fictional planner sees stock lasts six weeks on average but dips below zero in a promotional week, while the simple ratio remains a quick screening measure. Financial teams may use weeks of supply to spot slow-moving inventory, but high cover alone does not prove obsolescence, so check shelf life, demand and strategic reserve; it is a useful question of how long usable stock would cover us at this demand rate, and the answer is an estimate with stated assumptions.