What the product performance report shows
Most analytics reporting stops at the order. This one goes a level deeper and lists each product on its own row, with the number of times it was viewed, added to a basket, bought, and the revenue it produced. It answers questions about the catalogue rather than about the shop.
The data comes from the item array your site attaches to ecommerce events. Every product view, basket addition and purchase carries a list of items, each with an identifier, a name, a price and often a category and brand. The report is a summary of those lists, so it only ever contains what your tracking bothered to send.
Why it matters
It separates a traffic problem from a product problem, which no site-wide metric can do. A product seen by many people and bought by almost none has an issue on its own page — the photography, the description, the price, the delivery estimate or a review that puts people off. A product bought by nearly everyone who sees it but seen by hardly anybody has a visibility problem, and the fix is merchandising, internal linking or advertising instead.
It also tells you where advertising money should go. Feeding a shopping campaign an entire catalogue treats a product that converts well and one that never sells as equals. Knowing which items earn their place lets you build product groups that match reality, and lets you stop paying for clicks on items the shop cannot sell.
Where it goes wrong
Missing or inconsistent identifiers cause the most damage. If a product’s identifier differs between the listing page, the basket and the purchase — or changes when a variant is chosen — the same item splits into several rows and none of them is right. The identifier you send should match the one in the shop platform and the product feed, so the same product means the same thing everywhere.
Incomplete instrumentation is the other pattern. Many shops track purchases properly but never send product views or basket additions, which leaves a report showing sales with nothing to compare them against. Variants are a third trap: reporting every size and colour separately buries the product, while collapsing them all hides the fact that one size never sells.
How to act on it
Start by proving the data. Open the report, pick a product you know well, and check that its view, basket and purchase numbers move in a sensible relationship to each other. If additions to basket exceed views, or a bestseller is missing entirely, fix the tracking before making a single decision from it.
Once it is trustworthy, use it for a short list of actions rather than a monthly read-through. Find the products with strong interest and weak conversion and rewrite those pages first, since the demand already exists. Find products that convert well but get little traffic and give them internal links, category placement or a campaign of their own. Pull the worst performers out of shopping feeds so budget stops leaking into them. If the item data itself is patchy, that belongs with the analytics and tracking work before any merchandising decision is made.
