Your Amazon Sales Velocity Is Probably Wrong

A brand reviews a 30-day Amazon sales report and sees that its hero ASIN sold 1,140 units. The calculation appears straightforward — 1,140 units divided by 30 days is 38 units per day. The planning team uses 38/day for the next purchase order. The arithmetic is correct. The operating conclusion may still be wrong.

During those 30 days, the ASIN was out of stock for six full days, suppressed for one full day and unavailable for a combined 24 hours across two partial days. Customers could not buy the seller's offer for eight equivalent days. The product sold 1,140 units over 22 buyable-equivalent days, not 30.

The calendar metric is 26.7% below the buyable-day metric. Put differently, the buyable-day rate is 36.4% higher than the number being fed into the reorder system.

The mistake is not the formula. It is the denominator.

Most sales reports answer a historical question: how many units did the ASIN sell during the date range? Replenishment requires a different question: what was the rate of sale when the offer was in a condition to sell?

Important: buyable-day velocity is not a forecast by itself. Promotions, seasonality, price changes and advertising intensity must still be normalised.

What counts as a buyable day on Amazon?

A rigorous definition should be set before the metric is calculated. For most brands, a day should count as buyable only when the seller's offer has sellable inventory and is available for purchase for the required share of that day.

When hourly data is unavailable, use a consistent daily snapshot rule and label partial days rather than pretending the data is more precise than it is.

A worked replenishment example

The difference is 373 units before any safety stock is added. At a landed inventory cost of INR 240 per unit, the higher plan requires roughly INR 89,520 of additional working capital. At INR 180 contribution per unit, the same gap represents about INR 67,140 of contribution at risk if demand continues and inventory cannot be replenished in time.

Why the error is concentrated in your best ASINs

Slow-moving SKUs often remain continuously available, so calendar and buyable-day velocity are similar. Fast-moving SKUs are more likely to hit zero stock, face emergency price changes or experience inventory transfers and inbound delays. The products with the strongest demand therefore create the largest denominator distortion — and a damaging loop follows:

The Amazon data stack required

The metric can be built manually for a small catalogue, but a scaled implementation should combine four data layers via Amazon's SP-API, subject to account permissions and integration setup.

For seller-fulfilled inventory, quantity-change notifications and the live fulfilment-availability field can be used to distinguish submitted quantity from the quantity actually available for customers to purchase.

Five controls that prevent false precision

The operating principle

A sales report is not automatically a demand report. On Amazon, the difference between the two is availability. Brands that do not record when an ASIN was buyable will repeatedly under-read demand on the SKUs that stock out — and those are usually the SKUs carrying the account.

The answer is not a more complicated dashboard. It is a more accurate denominator.

Amazon can change report fields, API schemas and permissions; validate against your own Seller Central account and current SP-API documentation.