How to Reduce Stockouts With Data, Not Guesswork
Stockouts are one of the most expensive and most invisible problems in retail. They don't show up at month-end as a "loss" — they show up as a sale that simply never happened, which makes them easy to ignore until they become a recurring pattern.
On average, retailers lose 4–8% of annual revenue to stockouts on their top-selling items alone — most of it never shows up as a line item anywhere.
Origins
The scale of the stockout problem wasn't well documented until a landmark study by Thomas Gruen and Daniel Corsten, presented to the Grocery Manufacturers of America in the early 2000s, found that the average out-of-stock rate across retail was around 8% at any given time — and that roughly a third of customers facing one simply bought the item elsewhere. That study is a big part of why "on-shelf availability" became its own tracked discipline rather than something assumed to just work itself out.
The symptom is always the same
A high-turnover product runs out mid-week. Replenishment takes too long, the customer buys from a competitor, and no one on the team notices in time because the alert came late — or never came at all. Multiply that across dozens of SKUs over a month and the real impact hides inside "revenue that could have happened."
Fill rate
Fill rate = Units shipped complete and on time ÷ Units ordered
Fill rate and stockout rate measure closely related things from opposite ends — fill rate from the order's perspective, stockout rate from the SKU's. Tracking both catches gaps either one alone can miss.
Two kinds of stockout
Not every "out of stock" is the same problem. A genuine, "hard" stockout means the item truly isn't in the warehouse. A phantom stockout is different and arguably sneakier: the system says stock is available, but it isn't really there — misplaced, damaged and unlogged, or already gone through unrecorded shrinkage. Fixing a phantom stockout isn't a purchasing problem at all; it's a data-accuracy problem, closely tied to how good your stock tracking and shrinkage controls actually are.
Warning signs you're already losing sales
- The same handful of SKUs run out every single month, always around the same week
- Reorder decisions happen after someone notices an empty shelf or a customer complaint
- Purchasing relies on one person's memory of "what usually sells"
- You can't say, right now, which products will run out in the next 7 days
Why manual replenishment fails
Replenishment based on gut feeling works fine until the operation grows. Past a certain number of products and stores, it's humanly impossible to track stock coverage item by item — and that's exactly when stockouts stop being occasional and start being structural.
The real cost of a stockout isn't the missed sale — it's the customer who quietly switches to a competitor and never mentions why.
What changes with data analysis
Cross-referencing sales history, seasonality, and supplier lead time makes it possible to predict a stockout before it happens, not after. Instead of reacting to an empty shelf, the team gets the alert while there's still time to act — usually days before the product actually runs out.
Not every SKU deserves the same tolerance
A near-zero stockout tolerance makes sense for a high-revenue A-tier item, where a stockout is genuinely expensive. A C-tier item can usually tolerate an occasional gap without meaningfully hurting the business — spending the same monitoring effort on both isn't prioritization, it's just busywork spread evenly.
A 3-step process that actually works
Know how many days of stock are left for every SKU, updated daily.
Coverage alone isn't enough — you need to reorder before coverage runs out, not when it hits zero.
The team should see a warning days in advance, not an empty shelf report after the fact.
A simple example
Say a product sells 12 units a day, you have 40 in stock, and your supplier's lead time is 5 days. That's roughly 3 days of coverage left — which means the reorder should have gone out two days ago, not today. This is the kind of gap that's invisible in a spreadsheet but obvious once coverage is tracked automatically.
Limitations to keep in mind
- Only as good as the underlying stock data — coverage tracking assumes the recorded quantity is real, which breaks down under phantom stockouts or heavy shrinkage
- Sudden demand spikes can still outrun a good system — a viral moment or unplanned bulk order can exhaust coverage faster than any alert can react to
- Doesn't fix supplier-side constraints — perfect internal tracking still can't conjure stock a supplier physically can't produce in time
In practice
Teams that monitor stock turnover and coverage in real time can act days in advance, not hours. The gain isn't just avoiding one isolated stockout — it's no longer treating as normal a problem that, once you look at the numbers, is entirely predictable.
Key takeaways
Track coverage (days of stock left), not just current quantity. A product with "50 units" can be one bad week away from a stockout — coverage tells you that, raw quantity doesn't.
See also
Want to see this with your own company's data?
Upload your spreadsheet and get the analysis in seconds, for free.
