Traffic but no sales: three numbers that tell you where the buyer stops
"The product is not selling" is a symptom, not a diagnosis. At least three different causes produce that same symptom, and they call for opposite fixes: one is an advertising problem, one is a listing problem, one is a price or offer problem. If you do not separate them first, there is a good chance you will spend the week improving the part that was already fine.
One report separates them. Every claim below points at a column you can open in your own account, so you do not have to take my word for it.
Where the report is
Seller Central, Reports, Business Reports, then Detail Page Sales and Traffic by Child Item. It requires a Professional selling plan, and it is per marketplace: your German report says nothing about the UK.
Five columns matter. Ignore the rest for now.
- Sessions. How many visitors opened the detail page. All activity by one visitor within 24 hours counts as one session.
- Page Views. How many times the page loaded. Always equal to or higher than Sessions, because the same visitor can come back.
- Buy Box Percentage. The share of page views where the Add to Cart button was pointing at your offer.
- Units Ordered. How many units were ordered.
- Unit Session Percentage. Your conversion rate.
Unit Session Percentage is not a separate measurement
It is Units Ordered divided by Sessions, as a percentage. It is not collected separately, it is the ratio of the other two columns, and that is exactly what makes it useful: when it drops, either the numerator fell or the denominator rose, and those are completely different stories.
If your conversion dropped because orders fell, the problem is on the product side. If it dropped because traffic grew, that can be your advertising working as designed: more visitors arriving, less precisely targeted. Same drop, opposite fixes. The ratio alone cannot tell you which one happened. Its two parts can.
So the first move is always to write the conversion rate down next to Sessions and Units Ordered, for the same period.
The four cases
Open the report for the last 30 days and read the three numbers for the product you are worried about.
1. Sessions are essentially zero
There is nothing to convert. Working on the listing here is wasted money, because nobody is seeing it. The question is findability: is there any search you actually rank for, is advertising running at all, and does the title (plus your backend keywords) contain the words a buyer would really type.
The common mistake is reading zero traffic as a weak listing and starting to swap images. A better image does not become more visible.
2. Buy Box Percentage is below 100
The page is taking traffic, but for part of the time the Add to Cart button was not yours. Those visitors could not buy from you even if you convinced them. Your conversion looks low partly because the denominator contains visits where you had nothing to sell them.
This is not always competition. Your own offer can lose the Buy Box, usually for one of four reasons: your price is above the external market price, you ran out of stock, your FBM delivery performance slipped, or a compliance restriction is sitting on the listing.
Zero percent is its own case, not a more extreme version of the previous one: the page took traffic for the whole period while not a single visitor could buy from you in one click.
3. Buy Box near 100, conversion low
This is the only case where the listing itself is the problem. The buyer opened it, could see the Add to Cart button, and left anyway. This is when it is worth touching images, title, price and reviews, because from here on it really is about persuasion.
Keep an order to it. The main image and the price are visible in search results too, so they filter buyers before anyone reaches the page. Everything that is only visible on the page is the second round.
4. Every ratio looks fine, revenue is still small
That is a volume question, not a ratio question. Conversion is good, not enough people see the product. The fix is the same as in case 1, but you start from a better place: you already know that the traffic you get converts, so buying more of it pays back.
Three traps in the report
Parent and child views are not the same thing. The by Parent Item view merges variations, so one strong size hides a weak one and the average looks comfortable. Diagnosis happens in the by Child Item view.
Conversion is per session, not per page view. Divide by Page Views and you will hand yourself a worse conversion rate than you have, because you are counting the same visitor more than once.
Buy Box Percentage is a share of time, not a state. 85 percent does not mean your listing is 85 percent healthy. It means that for 15 percent of page views the button was not yours. If that 15 percent lands on a weekend, you lost the whole weekend.
Ten minutes a week
Open the report for the last 7 days and the last 30 days, same product. Write down Sessions, Buy Box Percentage and Unit Session Percentage for both windows. Whichever number moved between the two tells you what changed, and the four cases above tell you where to reach.
The point is not the number. The point is to replace "it is not selling" with a question that can actually be answered. Three people saw it and left, or three people never saw it: those are different problems, and the report separates them in about a minute.