Weeks of supply (WOS) estimates how many weeks current inventory can cover at a stated demand rate. For seasonal apparel, it is a decision aid—not an automatic reorder signal. A style can show six weeks of supply and still be overstocked if only four selling weeks remain, or be at risk if the next receipt arrives in eight weeks.
The calculation becomes useful only after the team defines inventory status, demand horizon, product level, season phase, and open purchase-order timing.
Basic weeks of supply formula
Basic WOS = inventory available for the decision ÷ average weekly demand
Decide whether the numerator means on-hand or available inventory:
- On-hand inventory is physically present, including units that may be held, damaged, allocated, or otherwise unavailable.
- Available inventory removes units that cannot support the decision, according to a documented rule.
For a sellable-stock decision, available inventory is often more useful. Do not subtract allocated wholesale units twice, and do not include returns awaiting inspection as sellable until their status supports it.
The demand-rate horizon should match the decision. A 12-week backward average can be misleading during a product launch, peak gift week, or markdown phase. State the weeks included, any stockout periods, and whether sales were full price or promoted.
The U.S. Census Bureau’s retail definitions show why the inputs must be labeled. Its public inventory estimate is the value of merchandise available for sale at cost on the last day of the period, while its inventory-to-sales ratio compares that end-of-month value with one month of sales. Census explains that a 2.5 ratio indicates two and a half months of inventory at that month’s sales rate. Apparel WOS instead uses a declared unit pool and weekly unit demand, so a team should not copy a value-based monthly ratio into this calculation.
Forward weeks of supply
Forward WOS compares inventory with a week-by-week demand forecast rather than one backward average. It answers: through which future week can available and scheduled inventory cover expected demand?
Start with available inventory, add only open purchase orders expected to become sellable by their dated arrival, then subtract the forecast for each week in sequence. The first week with a negative projected balance is the coverage limit. This approach handles launch, peak, and markdown phases that have different demand rates.
| Method | Demand Input | Best Use | Main Limitation |
|---|---|---|---|
| Basic WOS | One declared average weekly rate | Quick, stable-item check | Hides demand and receipt timing changes |
| Forward WOS | Week-by-week forecast | Seasonal and phased coverage | Depends on forecast and receipt assumptions |
Seasonal apparel example with open purchase orders
A brand has 480 available units of a quilted vest at the start of launch week. An open purchase order (PO) for 300 units is due at the start of week 4. Expected demand changes by phase:
| Week | Season Phase | Beginning Units | PO Receipt | Forecast Demand | Ending Units |
|---|---|---|---|---|---|
| 1 | Launch | 480 | 0 | 90 | 390 |
| 2 | Launch | 390 | 0 | 110 | 280 |
| 3 | Peak | 280 | 0 | 160 | 120 |
| 4 | Peak | 120 | 300 | 180 | 240 |
| 5 | Peak | 240 | 0 | 150 | 90 |
| 6 | Markdown Transition | 90 | 0 | 80 | 10 |
| 7 | Markdown | 10 | 0 | 50 | -40 |
Basic WOS using the first three weeks’ average demand is:
Average weekly demand = (90 + 110 + 160) ÷ 3 = 120 units
Basic WOS = 480 ÷ 120 = 4.0 weeks
That result ignores the 300-unit PO and the phase change. The forward schedule shows inventory lasting through week 6, with a 40-unit shortfall in week 7 if every input occurs as planned. If the PO slips past week 4, the brand would run short during peak week instead.
The question is not simply “Do we have four weeks?” It is whether stock covers the next decision window and supplier lead time without leaving excess after the useful selling season.
Blastramp’s demand forecasting guide explains how to build the future demand inputs. Its wholesale inventory guide adds context for separate channel commitments.
Check input quality before setting a target
- Confirm style-color-size identifiers and parent relationships.
- Reconcile on-hand, holds, allocations, damaged stock, and pending returns.
- Date each open PO by expected sellable availability, not only vendor ship date.
- Separate confirmed supply from optional or unapproved buys.
- Mark stockout weeks; observed sales during those weeks may understate demand.
- Check cancelled orders, lost sales, waitlists, and wholesale requests for unmet demand signals.
- Separate full-price, promotion, launch, peak, and markdown demand.
- Use the channel and location scope relevant to the decision.
- Record lead time as a range when supplier or transit timing varies.
- Review variant results before relying on a parent-style average.
Stockout-censored demand is especially important. If medium sold 40 units before going out of stock on Tuesday, its weekly sales are not evidence that demand was only 40. Lost sales are rarely observed perfectly, so document the proxy and test a range rather than inserting a false precise value.
A 2024 peer-reviewed study in the International Journal of Forecasting reports that sales understate demand when stockouts occur and that forecasts built from those observed sales are biased downward relative to forecasts based on the unobserved demand. The paper also treats the lost-sales inventory policy as part of model selection. For an operator, the practical data point is the stockout timestamp: without it, a low sales week can be mistaken for weak demand.
Set a target from the decision context
| Situation | Coverage Question | Practical Response |
|---|---|---|
| Reorder lead time extends beyond current coverage | Will useful stock last until the receipt becomes sellable? | Test expedite, transfer, allocation, or a smaller fast receipt |
| Coverage runs past the season end | Will units remain after demand falls? | Reduce or cancel uncommitted supply; avoid solving with a universal target |
| Parent style looks covered but core variants do not | Can customers buy the sizes and colors they want? | Review and act at variant level |
| Demand range is wide for a new style | How does coverage change in downside, base, and upside cases? | Use scenario WOS and set a review trigger |
| Wholesale units are committed | Which inventory is truly available to other channels? | Remove protected commitments under the declared allocation rule |
| PO timing is uncertain | Does coverage survive a late-arrival case? | Run an arrival range and name the decision deadline |
An “optimal” WOS target cannot be universal. The target should reflect supplier lead time, review cadence, demand uncertainty, remaining full-price weeks, open-to-buy budget, minimum orders, and the cost of leftover stock.
Teams can use a purchase-order and warehouse operating view to frame where receipt timing enters the decision. Review available integrations before assuming any source can supply current fields automatically.
Do not mistake backward sales for a forecast
A backward average describes what sold under past prices, availability, weather, marketing, and channel conditions. It becomes a forecast only after the team judges whether those conditions apply to the future period.
For a new style, use an analogous product with explicit similarities and differences, add collection and attribute context, and calculate downside, base, and upside demand. Reforecast from early sales once enough comparable exposure exists. For a replenished style, account for current price, stockouts, planned events, channel commitments, and the remaining selling window.
Aggregation also changes the answer. Parent-style demand may hide that small and medium need replenishment while fringe sizes will carry beyond season end. Calculate at the level where action happens, then roll up using units rather than averaging variant WOS percentages.
WOS has a narrow, useful job
Sell-through measures sales against a declared inventory pool. Safety stock is an uncertainty buffer. Open-to-buy governs spending capacity. Turnover describes inventory efficiency over a financial period. WOS estimates near-term coverage. Use the measures together, but do not treat one as a substitute for another.
If a reorder is financially possible, open-to-buy planning still decides whether the budget supports it. Brands comparing software costs can review Blastramp pricing. To discuss inventory, order, and integration fit for current channels, request a demo.
