Rate of saleRate of sale (ROS)

Rate of sale (ROS) is the number of units an option or SKU sells per week. It's the merchandiser's core velocity measure — the answer to "how fast is this line actually moving?" — and almost every downstream decision in trading depends on it.

If a line sells 40 units a week and you're holding 200 units, you have five weeks of stock. That single calculation — rate of sale into stock on hand — is the basis of weeks of supply, days of cover, and every replenishment trigger that follows.

Rate of sale is not sell-through

These two get conflated constantly, and they answer different questions.

• Rate of sale is a velocity: units per week. It's a rate, not a percentage. A line selling 40 units a week has a rate of sale of 40, whether you bought 200 units or 2,000.
• Sell-through rate is a proportion: the percentage of a specific buy sold within a window. 650 of 1,000 units sold in 8 weeks is 65% sell-through.

The relationship between them is the buy quantity and the window. Two lines can have identical rate of sale and wildly different sell-through — one was over-bought, the other was bought tight. And two lines can hit the same sell-through with very different velocities, because one was given more weeks to get there.

Rate of sale tells you how fast. Sell-through tells you how much of the buy is gone. You need both, and you need to keep them separate in your head.

The formula

Rate of sale = Units sold / Number of weeks on sale

A line that sold 320 units over 8 weeks on sale has a rate of sale of 320 ÷ 8 = 40 units per week.

Two refinements matter in practice:

• Only count weeks the line was actually available. If a line was stocked out for two of those eight weeks, dividing by eight understates the true rate. Divide by the six weeks it was genuinely on sale to get the demand rate of sale rather than the constrained one.
• Use a recent window, not lifetime. A trailing 4-week rate of sale reflects current demand. A lifetime average drags in launch-week spikes and end-of-life decay, and is rarely the number you want for a forward decision.

What ROS feeds into

Rate of sale is the input, not the output. Once you have it, it drives:

• Cover. Stock on hand ÷ rate of sale = weeks of cover. This is the everyday read on whether a line is healthy, light, or drowning in stock.
• Replenishment. Rate of sale × lead time in weeks tells you how much will sell before a reorder lands — the floor for your reorder quantity.
• Markdown timing. A falling rate of sale on a seasonal line is the early signal that it needs price action before the residual stock turns into dead stock.
• Buying. Last season's rate of sale on a comparable line is the most honest input to next season's buy quantity.

Trade at the level you actually decide at

Rate of sale is only useful at the level you make decisions. A brand-level rate of sale is a vanity number; it hides the fast lines propping up the slow ones.

• Option-level ROS for ranging and markdown decisions (the option is the unit a merchandiser trades).
• SKU-level ROS for size and colour curves, and for replenishment of individual variants.
• Channel-level ROS because the same option often sells at very different velocities on D2C versus wholesale versus marketplace — and a blended rate masks both.

The right granularity is whatever matches the decision in front of you.

Common pitfalls

• Dividing by calendar weeks instead of weeks in stock. A line that was out of stock for half the period looks slow when it was actually fast. Always rate against availability.
• Using a lifetime average for a forward decision. Launch spikes and tail decay both distort it. A trailing 4-week window is usually the right read for replenishment.
• Reading rate of sale at the wrong level. A blended brand or category number hides the spread; the decision lives at option, SKU, or channel.
• Confusing it with sell-through. ROS is units per week; sell-through is percentage of a buy. Reporting one when you mean the other leads to wrong reorder calls.