SKU velocity: reading the speed of the shelf

How fast does your product move where it's actually stocked? That's the question SKU velocity answers. A SKU, or stock-keeping unit, is a specific sellable item, down to the flavor and size. Velocity is that item's rate of sale where it's available, most often expressed as units per store per week. It separates how well an item sells from how widely it's distributed, and that separation is the whole value of the measure.

Total sales can flatter a mediocre product with wide distribution. Velocity can't be flattered that way.

How is SKU velocity calculated?

Take the units sold, divide by the number of stores that actually sold the item, divide again by the number of weeks in the window. As a hypothetical illustration: an item selling 500 units across 50 stores in one week runs at 10 units per store per week. Another item selling 800 units across 200 stores runs at 4. The second item sells more in total; the first one sells faster where it sits.

The denominator deserves attention. "Stores stocking the item" and "all stores in the dataset" give different velocities, and both are legitimate measures with different meanings. The first tells you how the item performs on shelves it occupies. The second blends performance with distribution. Know which one a report is using before you compare anything.

Why not just use total sales?

Because total sales answers a different question. Volume tells you how big an item's business is. Velocity tells you how much demand exists per point of distribution, which is what predicts what happens next. An item with modest volume but strong velocity in a narrow footprint is a candidate for expansion. An item with big volume but weak velocity is coasting on distribution and may be vulnerable at the next shelf review.

In the hypothetical above, the 10-per-week item is the more interesting story despite the smaller total, because each store it enters can be expected to add sales at a rate the slower item has never demonstrated.

What makes a velocity comparison fair?

Same kinds of stores, same window, same denominator. Comparing an item's velocity in urban corner stores against a competitor's velocity in suburban supermarkets tells you about the formats, not the items. Comparing a holiday week against an ordinary one tells you about the calendar. The discipline is unglamorous: match the store set, match the weeks, state the denominator, and only then read the difference as real.

Small-format retail rewards this care, because store-to-store variation in the independent channel is wide. Aggregating across thousands of independent stores, as the data behind NRS Insights does, smooths single-store noise while preserving the channel's actual behavior.

How does velocity guide real decisions?

Three uses come up constantly. Defending shelf space, because a distributor or owner deciding what stays is really asking which items earn their facing, and velocity is that answer. Building an expansion case, because demonstrated speed in existing stores is the evidence that new stores will follow. And reading a launch, because early velocity in the stores that stocked first is the cleanest signal a new item sends, long before total volume means anything.

One hypothetical makes the launch case concrete: a new item enters a modest set of stores and, within weeks, runs at twice the velocity of the category's median item in those same shops. Distribution is still small, so total volume looks unimpressive on a summary page. The velocity line is the one telling the truth about the product's future.