A stockout at a retail partner does not announce itself. Your sell-in data shows inventory shipped and received. What it cannot show you is an empty shelf. Between the moment a product sells out at a boutique in Stockholm and the moment your operations team hears about it, weeks can pass — weeks during which demand is lost to competitors or simply unmet.
For fragrance brands with 14-week production lead times, a stockout that goes undetected for four weeks is not a four-week problem. It is a 14-plus-week problem, because by the time you know, you are starting a production cycle that will not resolve the gap until the next season.
Why stockouts are invisible in wholesale
In direct-to-consumer channels — your own website, your own store — stockouts are immediate and obvious. An out-of-stock flag appears. Sales drop to zero. Your system tells you.
In wholesale, you have no access to a partner's inventory systems. You receive periodic sell-out reports, often weekly or monthly. What those reports cannot tell you — unless you calculate it — is whether a product is still on the shelf. A product that sold zero units last week might be out of stock, or it might just not have sold. The data looks identical.
The only way to distinguish between "sold nothing" and "nothing left to sell" is to track closing inventory alongside sell-out, and to watch for the pattern that precedes a stockout: accelerating sell-out velocity followed by a sudden drop to zero.
The warning signals in sell-out data
Sell-out data contains stockout signals if you know what to look for. The most reliable indicators:
- Weeks of Supply below threshold: If closing inventory divided by weekly sell-out velocity drops below four to six weeks, replenishment is urgent. Below two weeks, you are in stockout territory unless a reorder is already in transit.
- Velocity spike followed by zero: A product selling 30 units a week for three weeks, then zero, is almost certainly stocked out — not suddenly unpopular.
- Closing inventory reporting as zero: When a partner includes closing inventory in their report and it hits zero, the signal is direct.
- Store-level anomalies vs. network average: One store selling zero while the same SKU sells normally at five other stores points to a location-level stockout rather than a demand problem.
Most brands miss these signals not because they are subtle, but because the data is not consolidated in one place. When your sell-out reports live in fifteen different Excel files with different formats, calculating weeks of supply across all locations for all SKUs is not a quick task.
Setting weeks of supply thresholds
The WOS threshold that triggers an alert depends on your production lead time and your minimum order quantities. A rough framework:
- Lead time under 6 weeks: Alert when WOS drops below 3 weeks at any retail location
- Lead time 6-14 weeks: Alert when WOS drops below 8 weeks — enough time to place an order before stock runs out
- Lead time over 14 weeks (custom components, specialty glass): Alert when WOS drops below 16 weeks — which means you are essentially managing reorder timing continuously, not reactively
For fragrance brands with specialty components, the upper end of this range is common. You are not monitoring for stockouts after they happen — you are monitoring for the trajectory that leads to a stockout four months from now.
Automated alerts vs. manual monitoring
Manual monitoring — reviewing sell-out data each week and checking WOS for each SKU at each partner — is theoretically possible. At three retail partners and ten SKUs, a person can do it in an hour. At fifteen partners and fifty SKUs, you are looking at a day of work each week to monitor adequately.
Automated alerts invert the model. Instead of checking everything routinely, you receive a notification when something crosses a threshold. Your operations team spends time responding to signals, not searching for them. The coverage is also complete — automation does not forget to check the Stockholm boutique because it was busy with Paris this week.
Frequently asked questions
How do I calculate weeks of supply from sell-out data?
Weeks of Supply = Closing Inventory ÷ Average Weekly Units Sold. Use a rolling average of 4 to 8 weeks for the denominator to smooth out seasonal spikes. If a partner does not report closing inventory, you can estimate it using cumulative sell-in minus cumulative sell-out — though this estimate degrades in accuracy over time without a physical count to recalibrate.
What if my retail partners do not report closing inventory?
Many partners do not. In that case, ask for it specifically — it is often available in their system even if not included in standard reports. As a fallback, track sell-out velocity trends. An abrupt drop to zero units sold at a location that was previously active is a reasonable proxy for a stockout, even without confirmed inventory data.
How do I distinguish a genuine stockout from a slow period?
Look at the pattern, not just the point. If a SKU was selling 20 units a week for six weeks and then dropped to zero suddenly, that is a stockout signal. If it was selling 20 units a week and has gradually slowed to 5 over two months, that is a demand trend. Context from other stores selling the same SKU also helps — if peers are still selling normally, a single-location zero is almost certainly a stock issue.
Can I set different WOS thresholds for different retail partners?
Yes, and you should. A flagship partner doing high volume needs a longer lead time buffer than a small boutique that moves slower. Partners in markets with complex logistics — cross-border, customs, specific delivery windows — may also need wider buffers. Threshold management is one of the operational levers that experienced distribution teams use but rarely formalise.
Know before the shelf is empty
TaskifAI calculates Weeks of Supply at the SKU and store level across all your retail partners — and alerts you when any product crosses your reorder threshold. No more finding out about stockouts from a buyer email.
Book a demo to see WOS tracking for your own distribution network.