Summary
Did your Amazon business move this period, and how much of it is being carried by ads?
Amazon can grow and get less profitable at the same time, and on a sales chart the two look identical.
This page separates them, then tells you which country is responsible.
The cards
Whether Amazon grew is one number. Whether you got to keep the growth takes two more.
SalesSalesAmazon's fees and returns still come out of it. Compare with your payouts to see what you actually keep. · Ad SpendAd Spendother channels sit elsewhere. Rising spend is fine only if sales and the return keep pace. · Units OrderedUnits OrderedCounts units, not orders, so one big basket can lift it without a new buyer. Read it beside sales to see what each unit earned · Units RefundedUnits RefundedCounts units coming back, not the money lost. Break it down by product to see whether one listing is behind the rise · ASPASPThe average price shoppers actually paid, not your list price. A fall while units hold steady points to discounts or cheaper items selling · Unit Session %Unit Session %Counts units, not orders, so a 3-unit order counts as 3. If it slips while sessions hold steady, look at price, Buy Box or reviews12% or more is healthy · TACOSTACOSHow much of your Amazon business leans on ads, not just the ad-driven part. A climb while sales stay flat means you're paying more to stand stillUnder 8% is healthy · ACOSACOSabove that, those sales lose money.Under 20% is healthy · ROASROASCounts only sales Amazon credits to these ads, not your whole Amazon business. Compare it against your product margin before deciding a campaign is paying4 or more is healthy · SessionsSessionsone shopper checking back all week shows up several times. If it climbs while units stay flat, fix the listing, not the traffic.
Read them in three moves
- Did sales move? Sales. Nothing underneath means anything until you know the direction.
- Volume or price? Units Ordered and ASP. Sales is the two of them multiplied, and almost always only one of them moved.
- What did ads cost? TACOS first, ACOS second. TACOS is what ads took out of the whole business; ACOS is what they took out of the sales they actually caused.
Three pairs worth holding together
TACOS + ACOS. ACOS steady while TACOS climbs is the quiet one: your campaigns are running exactly as well as before, but organic sales are shrinking underneath them, so ads carry a bigger share of the business every week.
Sessions + Unit Session %. Sessions up with Unit Session % down means the extra traffic is not the right traffic. More listing views, no more buyers.
Units Ordered + Units Refunded. Refunds growing faster than orders points at the listing or the product, not at the advertising. No bid change fixes it.
Sales is measured before refunds. Units Refunded sits right beside it and is not taken out of it. Two cards, two different moments in the order's life.
The charts
A period total hides the day something broke. This is where you find that day.
All four are drawn from the same daily figures, so a day that looks odd in one is worth checking in the others.
TACOS & Sales
TACOS as a line on a percentage scale, Sales as bars in your currency.
Growing on your own
Sales rose and ads took a smaller cut of them. Organic rank and repeat buyers are doing part of the work.
You bought the growth
The extra sales came at a worse price for the business. A decision while ACOS stays under margin, a problem once it does not.
Pulled back
You spent less and sold less. Fine if it was deliberate, worth a look if it was not.
Paying more for less
Ads are taking a bigger share of a smaller business. Usually a competitor's price, a stock-out, or a lost Buy Box.
The hover text on this chart reads TACOS the other way round, treating a rising TACOS as a sign of profitable scaling. The TACOS card's own grading treats lower as better, and the formula agrees. Read lower as better.
ACOS & Ad Spend
ACOS as a line, Ad Spend as bars. This is the campaign-level view of the same week.
Found headroom
More money in and a cheaper price per advertised sale. Rare, and worth spending into.
Buying volume
Each advertised sale is costing more. Acceptable while ACOS sits below your product margin.
Trimmed the waste
You removed spend that was not converting and efficiency improved. Check units did not fall with it.
The real alarm
Spending less and paying more per sale. The clicks cost the same; the orders went elsewhere.
Sales & Ad Spend
Ad Spend as a line, Sales as bars, both on the same scale. When spend is small next to sales the line will sit near the bottom and look flat — that is the axis, not the spend.
Use it for one thing: finding the day where the two shapes disagree. A spend spike with no matching sales bar is the day worth asking about, and the day to check in Ads Analytics.
Conv Rate, Sales & Average Selling Price
Three series: a rate line, Sales as bars, and average selling price as a second line.
- Sales first. It is the outcome; the other two only explain it.
- Then Average Selling Price. If sales moved and price moved with it, you had a price change, not a demand change.
- Then the rate line. Price flat and sales moving means conversion did the work — and a drop there traces to the listing, the reviews, or somebody else's price.
The line labelled Conv Rate on this chart is Unit Session %, the same measure as the card above. It counts units against sessions, so a three-unit order counts as three. It is not an order-based conversion rate and will read higher than one.
The tables
One market is usually carrying the whole story, good or bad. This is where you find out which one.
Combined Summary
One row = one country marketplace.
Columns: Ordered Product SalesOrdered Product SalesShows gross product demand from Amazon orders · Ad Spend · Units Ordered · Units Refunded · ASP · Unit Session % · TACOS · ACOS · ROAS · Sessions, each followed by a % △ against your comparison period.
Ordered Product Sales here and the Sales card above it are the same number under two names. Nothing is being measured differently.
How to scan it
- Sort by Ordered Product Sales, highest first. It opens that way.
- Look across at TACOS, then Unit Session %.
- The row that matters: a large market whose TACOS runs above the others while its Unit Session % runs below them. You are paying more for ads in the place where the listing converts worst — fix the local listing and price before you add budget there.
Then run it a second way. Sort by Units Refunded, highest first, and read ASP beside it. A country returning more of a higher-priced item is the one eating your margin quietly.
Filters
Two of these change what the numbers say. The rest change how you look at them.
Select Country — the Amazon account picker. It appears only when more than one Amazon account is connected, and it is hidden entirely when all stores are selected.
Date range — drives the cards, the table and all four charts together.
Chart view — Combined Summary switches from table to bar, line or pie, opens on Ordered Product Sales, plots up to three metrics at once, and can overlay the previous period.
Import progress — tells you whether the numbers are finished. While it is running, the figures below are still filling.
Watch out for
Amazon data arrives on Amazon's schedule. The progress bar at the top tells you what has landed. A day read before it finishes will look weak and then quietly correct itself.
Sessions are Amazon product-page visits. They are not website sessions from Google Analytics and will never match. Different sites, different counters.
Unit Session % is not a per-order conversion rate. It counts units, so a three-unit order counts three times. Comparing it against a Shopify or GA4 conversion rate is comparing two different measures.
ACOS only sees advertised sales; TACOS sees the whole business. A flattering ACOS on campaigns that touch a tenth of your catalogue tells you very little on its own.
% △ needs a comparison period. With none set, every delta column on the table is empty.
Do this first
Read TACOS on the cards, then open Combined Summary and sort by Ordered Product Sales. Find which country is carrying that TACOS — the fix is almost always in one market, not across all of them.
See yoursYour Amazon sales, ad cost and conversion, by country and by day.
Open Amazon Summary →