Most inventory reports answer the wrong question. They tell you what you have: 4,200 units in the warehouse, 800 at the 3PL, 60 days of stock at last month's rate. What they rarely tell you is what is moving, how fast, and whether that speed is accelerating or falling off a cliff. Two SKUs with identical on-hand counts can be in opposite situations: one is three weeks from a stockout , the other is a year of carrying costs waiting to happen.
Sales velocity is the number that separates them. It measures how many units of a SKU you sell per unit of time, and it turns a static stock count into a statement about the future: at this pace, this product runs out on this date. Operators who track sales velocity per SKU, per channel, and per location make better calls on reordering, allocation, and markdowns than operators staring at quantity on hand.
This guide covers how to calculate sales velocity, how to use it to drive replenishment and safety stock, how to rank your catalog with it, and why the number is only as useful as the data feeding it.
What Sales Velocity Tells You That Stock Counts Don't
A stock count is a snapshot. Sales velocity is a trajectory. The distinction matters because every inventory decision you make, from cutting a purchase order to approving a promotion, is a bet on what happens next, not on what the warehouse looked like on Tuesday.
Velocity converts inventory into time. Divide on-hand stock by weekly velocity and you get weeks of cover, the single most decision-ready number in inventory management. A SKU with 4,200 units and a velocity of 1,400 a week has three weeks of cover. If the supplier lead time is four weeks, that comfortable-looking pile of stock is already a problem. The count said you were fine. The velocity says you are late.
Velocity also exposes the difference between products that look similar on a revenue report. Two SKUs each doing $30,000 a month are not the same business if one does it in 500 units through a single grocery chain and the other in 6,000 units across DTC, Amazon, and wholesale. They need different buffer stock, different reorder cadences, and different levels of attention when their numbers move.
How to Calculate Sales Velocity
The base calculation is simple: units sold divided by the number of days or weeks in the period. A SKU that sold 1,800 units over the last 90 days has a velocity of 20 units a day, or 140 a week. Most brands calculate weekly velocity from a trailing 90-day window, long enough to smooth out noise, short enough to reflect the current business.
The choices around the window matter more than the arithmetic. Use sell-through, not shipments into a channel, or a big retail order will masquerade as consumer demand. Exclude stockout days from the denominator, because a SKU that sold nothing for three weeks while it was out of stock did not have zero demand; it had zero availability, and averaging over those days understates true velocity right when you most need it accurate. Flag promotion periods, since a BOGO week will inflate the trailing average and trigger overbuying for a demand level that ended with the promo.
Then segment. Company-wide velocity is a vanity number. The useful versions are per SKU per channel, because Amazon, DTC, and retail move at different speeds and drain different stock pools, and per SKU per location, because replenishment happens at a warehouse, not at a company. A brand shipping from two warehouses and one 3PL needs velocity for each SKU-location pair to make any real decision about where stock should sit.
Finally, watch the trend, not just the level. A SKU at 140 units a week that was at 90 six weeks ago is a different animal than one that was at 200. Comparing a short window against a longer one, four weeks against twelve, is the quickest way to spot acceleration and decay before they show up as an emergency.
Using Sales Velocity to Set Reorder Points and Safety Stock
Velocity is the demand input in every replenishment formula that works. A reorder point is velocity multiplied by lead time, plus a buffer: sell 20 units a day against a 21-day lead time and you reorder at 420 units plus safety stock . When velocity shifts and the reorder point does not, you get the two classic failures. Velocity rises and you stock out during the lead time. Velocity falls and you keep reordering for a demand level that no longer exists.
Velocity variability should size the safety stock itself. A SKU that reliably sells 100 units a week needs a modest buffer. A SKU that averages 100 but swings between 40 and 220 needs a much larger one, even though the two averages are identical. Brands that set safety stock as a flat percentage across the catalog are overprotecting their steady sellers and underprotecting their volatile ones with the same policy.
Purchase quantities follow the same logic. Buying 12 weeks of cover means 12 weeks at current velocity, not at the velocity assumed in last year's plan. Before every significant purchase order , the two-minute check is simple: what is this SKU's velocity over the last four weeks, how does it compare to the window the order was planned against, and does the quantity still make sense? That one habit prevents most of the excess inventory a growing brand accumulates.
Ranking Your Catalog by Velocity
Line up every SKU by velocity and the catalog sorts itself into three groups that deserve three different playbooks. The top group, usually a fifth of the SKUs driving most of the volume, gets tight monitoring, generous safety stock, and priority when supply is constrained. Stockouts here are the expensive ones. The middle group runs well on simple rules and periodic review. The bottom group, the slow tail, is where cash goes to hibernate, and every quarter it should have to argue for its shelf space.
Velocity also settles allocation arguments with data instead of volume of opinion. When three channels want the same constrained stock, per-channel velocity paired with per-channel margin tells you where units convert fastest and earn most. When a multi-location brand decides where to position inventory, location-level velocity says which warehouse actually needs the next container.
The ranking earns its keep on the way down, too. A fast mover decelerating for six straight weeks is the earliest honest signal to slow purchasing, plan a markdown, or investigate what changed, months before the yearly review finds a pallet of it. Slow-and-declining SKUs flagged by velocity feed directly into discontinuation decisions, and steady velocity history per SKU is the foundation any real demand planning practice gets built on.
Why Velocity Numbers Go Stale in Spreadsheets
Every operator who has tried to maintain velocity numbers by hand knows the cycle. Someone exports order data from Shopify, order data from the EDI portal, and inventory from the 3PL dashboard, joins them in a spreadsheet, and produces a beautiful velocity report. It is accurate for about a week. Then a channel gets added, a SKU gets renamed, the export format changes, and the report quietly drifts from the truth until a stockout or a write-off exposes the gap.
The failure is structural, not personal. Velocity is a derived number: it is only as current as the sales data underneath it, and that data is born in four or five different systems. Any process that depends on a human re-assembling those systems weekly will skip a week, and the decisions keyed off the number, reorder points, buy quantities, allocation, will silently run on old information.
This is why velocity belongs in the system that already sees your orders, not in a file next to it. When order management and inventory management run on the same platform, every order that ships updates velocity per SKU, channel, and location automatically. Weeks of cover recalculates continuously. Reorder points reference live velocity instead of last quarter's export, and a decelerating SKU shows up on a dashboard the week it slows, not the quarter after.
Make Velocity the Number Everyone Watches
Quantity on hand tells you what your cash is currently sitting in. Sales velocity tells you what to do about it: what to reorder, what to redistribute, what to mark down, and what to cut. Brands that run replenishment, allocation, and portfolio reviews off live velocity carry less stock and miss fewer sales than brands managing from counts, and the gap widens every time demand shifts.
DOSS Operations Cloud makes that the default. It connects orders, inventory, and procurement in one place, so velocity is computed from every channel automatically and flows straight into reorder points and draft purchase orders. It integrates with the tools you already run, and implementation takes months, not years. Book a demo to see the velocity of every SKU you own, live, without a single export.
