Skip to main content

SKU Performance

Which of your Amazon products earn their traffic, and which spend ad money without returning it.

90 second readAppears on: /amazon/sku

Some products pay for the rest of the catalogue. Some are being paid for.

This page tells you which is which, and whether ad money is the only thing keeping the second group alive.

The tables

Three questions get settled here: which SKU is eating the budget, which listing turns visitors into buyers, and which one is quietly losing the sale to another seller.

SKU Summary

One row = one SKU, for the marketplace chosen in Select Country.

The row is stitched from two sources. Sales, units, sessions, page views and Buy Box come from Seller Central; Ad SpendAd SpendRead it next to ROAS and sales; spend only helps if the return holds., ACOSACOSThe share of ad revenue eaten by the ads themselves.Under 20% is healthy and TACOSTACOSHow dependent your whole business is on ads.Under 8% is healthy come from Amazon Ads.

If the ad account for that marketplace is not connected, the ad columns stay blank while the sales columns fill in normally. Blank there means "not connected", not "no spend".

Columns, left to right: SKU · ASIN · Ordered Product SalesOrdered Product SalesShows gross product demand from Amazon orders. · Ad Spend · Units OrderedUnits OrderedThe volume side of demand, before value. · ASPASPYour average price point, and a lever for revenue. · Unit Session %Unit Session %How well your listings turn visits into sales. Counts units, so a 3-unit order counts as 3.12% or more is healthy · TACOS · ACOS · ROASROASRevenue earned for every ad dollar spent.4 or more is healthy · SessionsSessionsTraffic to your listings; only lifts sales if conversion holds. · % Sessions% SessionsShows this SKU's share of total traffic. · Page ViewsPage ViewsHigher means more product visibility. · % Page Views% Page ViewsShows this SKU's share of total visibility. · Buy Box %Buy Box %Higher is better; losing Buy Box usually hurts sales.

Each number is followed by a % △ column comparing it to the previous period. In org mode a Shop column appears after ASIN and the country picker disappears.

Scan 1 — find the SKU eating the budget

  1. Sort by Ad Spend, highest first
  2. Look across at ACOS, then at TACOS
  3. The row that matters high spend, ACOS above your product margin, TACOS not far below it — the ads are carrying that SKU rather than helping it

Scan 2 — traffic problem or listing problem

  1. Sort by Sessions, highest first
  2. Look across at Unit Session %
  3. The row that matters plenty of sessions, weak Unit Session % — people arrive and leave. That is a price, image, review or stock problem on the listing, not a traffic problem

Flip it for the opposite row: strong Unit Session % on few sessions is a listing that converts and nobody sees. That one is worth advertising.

Scan 3 — the quiet loss

  1. Sort by Buy Box %, lowest first
  2. Look across at Ordered Product Sales and Sessions
  3. The row that matters traffic holding, Buy Box % slipping, sales down — the clicks still arrive and another seller takes the order
ACOS + TACOS

ACOS only counts sales the ads were credited with. TACOS counts all of that SKU's sales. When the two sit close together, almost nothing sells organically — pausing the ads there stops the product, not the waste.

Sessions + % Sessions

Sessions is the count; % Sessions is that SKU's share of your whole Amazon catalogue's traffic. A SKU can gain sessions while losing share, which means the catalogue grew around it.

Filters

Country and date decide what the table contains. Search, the view switcher and Export only change what you do with it.

Select Country switches the marketplace. It only appears when more than one Amazon account is connected, and it shows one marketplace at a time — never a total across countries.

The date range at the top of the app drives the table and every % △, which compares against the period immediately before.

Search matches the SKU column and filters rows only. The summary row at the foot is hidden while a search is active, because it totals every SKU in range rather than the ones you searched for.

The view switcher turns the table into a bar, line or pie chart. It opens on Ordered Product Sales, takes up to three metrics and up to ten SKUs at a time, and Compare overlays the previous period.

Export downloads what is on screen.

Watch out for

“Sales are down on this SKU, so demand dropped.”

Check Buy Box % first. Losing the Buy Box drops sales while sessions stay flat, and no amount of ad spend fixes it.

“Ordered Product Sales is what I banked.”

It is ordered sales before refunds, and before Amazon's fees. For money that actually arrived, read Amazon P&L, which works off settlements.

“Buy Box % is 60%, so I lose it four days in ten.”

It is an average of Buy Box ownership across the days in your range, not a count of days. A short bad patch and a permanent partial loss can produce the same 60%.

“The ad columns are empty, so this SKU has no ads.”

Empty also means the Amazon Ads account for that marketplace is not connected. The Amazon Import chip at the top tells you which of Ads, Marketplace and Settlement have finished for each country.

Amazon backfills a limited stretch of history on connect, not your whole trading past. Pick a range that starts before your import began and the table will be thin for reasons that have nothing to do with your products.

Do this first

Sort by Ad Spend, take the top ten rows, and compare each ACOS against that product's margin. Anything above margin is losing money on every advertised sale, and it is losing it fastest at the top of that list.

See yoursYour SKU Summary for the last 30 days, ranked by sales.

Open SKU Performance