When Steady Sales Hide Shifting Customer Intent

The Global Signal

Retail's out-of-stock problem is larger, and stranger, than most inventory dashboards suggest. IHL Group, an independent retail and hospitality technology research firm, published an analysis estimating that retailers lose over $634 billion annually to out-of-stock situations worldwide, a figure widely cited across retail trade press including Chain Store Age. The number is not driven by empty shelves alone. IHL's breakdown attributes roughly $238 billion of that loss to genuinely empty shelves, but separately identifies large losses tied to customers who could not find sales help and merchandise that staff could not locate even though it was technically in the store.

That second category is the more interesting one for this argument. A product can be present in a store's inventory system, counted, in stock, and still functionally unavailable to the customer standing in front of an empty peg or an unhelpful search. From the register's point of view, nothing happened: no sale, no signal, no line item. From the customer's point of view, something happened that will shape whether they come back.

The Hidden Signal

Inventory velocity, the metric most retailers actually watch, measures what moved. It has no way of seeing what a customer wanted but did not get. Consider a hypothetical scenario, smaller and more specific than IHL's industry-wide figures but useful for tracing the mechanism: a skincare retailer sells a specific mineral sunscreen that a loyal customer buys every month. One month it is out of stock. She buys a cheaper alternative instead, then returns it because the formula does not suit her. The following month, she does not come back at all. The category's overall velocity number never moves. The stocktake shows nothing unusual. But the customer who mattered most to that category is gone, and the sale that replaced her purchase was a consolation, not a preference.

Multiply that pattern across a chain, and the category can look perfectly healthy by volume while its actual customer base is quietly turning over toward less loyal, more price-sensitive shoppers who are more likely to return what they buy.

Why the Visible Metric Misleads

If a category moves steadily, the intuitive conclusion is that customers want what is on the shelf. That logic holds until substitution enters the picture. A customer who came in for one product and left with another because the first was unavailable generates a sale in both categories: a data point in the product they left and a data point in the product they took instead. Both look like demand. Neither necessarily is.

The more useful signals sit beside the sales number rather than inside it: whether visit frequency to a category is declining while its raw velocity holds flat, whether a stocked-out item generates a lasting lift in a substitute that never fades once the original is back, and whether returns cluster around items that were bought as a second choice rather than a first one. None of these show up in a standard inventory report, because inventory systems were built to answer whether something moved, not why it moved or whether the customer would have preferred something else entirely.

Implication for Operators

Retailers that separate velocity from intent gain something a stable sales number cannot show them: an early warning that their most loyal customers are drifting toward substitution, well before it shows up as a category-wide decline. The practical shift is treating restocking as a preference decision, not a volume decision, and being honest that a flat sales line can still be hiding a customer base that has quietly changed underneath it.