Ecommerce Product Analytics: How to Understand Product Performance

Product analytics follows each product through the journey: how often it's shown in lists and clicked, viewed, added to cart, purchased, kept rather than returned, and how much margin it earns. Look at the whole path, because each drop points to different causes (visibility, appeal, information, checkout, expectations). Segment by traffic source, device and list, compare products with their own history and similar products rather than generic benchmarks, include returns and margin, and turn findings into merchandising, content and product page actions.

The flow above shows the stages a product passes through. A product can fail at any of them: rarely shown, often shown but rarely clicked, viewed but rarely added, added but rarely bought, bought but often returned, or sold at thin margin. Each failure has different causes and fixes. For the events that feed this analysis, see ecommerce event tracking.

Numbers point to where; evidence explains why. Combine metrics with product page reviews, session recordings, return reasons, reviews and search terms before changing anything. A low add-to-cart rate on a bestseller's new colour may mean poor imagery; on a new product it may mean mismatched traffic from a campaign.

Product performance differs by traffic source (paid social vs returning email), device, market and list (search results vs category vs recommendations). A product may convert well from search and poorly from social ads. Segment by these dimensions before deciding a product is weak. See customer segmentation.

Categories are where most product discovery happens. Analyse entries, click-through to products, filter and sort usage, conversion, revenue, returns and margin per category, and how well default sorting surfaces sellers. Categories with many entries but low click-through often have card, sorting or relevance issues. See category page optimization.

Gross revenue flatters products that are frequently returned or heavily discounted. Include refunds (from refund events or order data), return reasons and product costs from your platform or ERP to see net revenue and margin by product. This often changes which products deserve promotion. See analytics architecture.

Add-to-cart and conversion rates vary widely by category, price point, traffic mix and store. Published averages rarely apply to your situation. Compare each product with its own history (before and after changes) and with similar products in your catalog. Use minimum traffic thresholds before drawing conclusions from small numbers.

Review top products and categories weekly (changes in exposure, conversion, stock), run deeper monthly reviews that include returns and margin, and seasonal assortment reviews. Present products in tables with the full path rather than single metrics, and flag significant changes. See ecommerce dashboard design.

An illustrative scenario: a homeware store's top-selling cushion also has the highest return rate, while a similar cushion with fewer sales has low returns and higher margin but appears on page three of its category. Return reasons for the first mention colour differences. The team improves colour accuracy and descriptions for the first cushion and moves the second higher in default sort and collections. It tracks exposure, conversion, returns and margin for both over the following month.

A useful weekly product report shows the whole path in one table: list views, click-through, product views, add-to-cart rate, purchases, revenue, return rate and margin, with change versus the previous period and flags for significant movement. Sort by revenue or by opportunity (high views, low conversion). See dashboard design and merchandising.

Product analytics is about the whole path: exposure, clicks, adds, purchases, returns and margin, segmented and compared sensibly, then turned into actions. For how products are organized, see ecommerce merchandising.

For related guides, see returns management.