← Friday Seller Tools

Seller guide

Why do my eBay listings get impressions but few views or sales?

Split a seller-supplied eBay traffic funnel into impression-to-view and view-to-transaction gaps, compare each listing with explicit targets, and review likely causes without treating a funnel label as a diagnosis.

Use one comparable eBay traffic window

Collect listing reference, impressions, views, transactions, and the report's click-through and sales-conversion rates for the same marketplace and dates. Preserve zero rows and report rates rather than calculating only from rounded display values. Exclude or label listings that were inactive, out of stock, newly launched, ended, or materially repriced during the window so unlike exposure is not ranked as if it were comparable.

Separate the two funnel stages

Impression-to-view rate asks whether an exposed listing earned a visit. View-to-transaction rate asks whether a visit became a transaction. Compare each supplied rate with a seller-chosen review target and calculate the relative gap. A row can miss the first stage, the second stage, both, or neither. That label states where the observed gap appears; it does not state why the gap exists.

Weight the queue by the opportunity you actually observed

A large percentage gap on ten impressions is usually less urgent than a moderate gap on thousands. Weight first-stage gaps by impressions and second-stage gaps by views, then rank within the supplied batch. Keep raw counts and count-derived rates beside the reported rates as a cross-check. Transactions are not units, revenue, contribution, or profit, so do not turn the ranking into a revenue forecast.

Diagnose the live listing before drafting a change

For a first-stage gap, review query relevance, title facts, item specifics, main-image clarity, price context, shipping promise, and traffic source. For a second-stage gap, review offer competitiveness, full gallery, description, condition, returns, seller signals, variation availability, and buyer expectation. These are investigation prompts, not causes inferred from the traffic export. Verify each against the live listing and current report.

Change one bounded thing and compare again

Choose the smallest supportable edit for a high-volume row, record the exact change and date, and leave unrelated variables stable where practical. Compare a later like-for-like window only after enough exposure accumulates. Do not revise, promote, end, or reprice every flagged listing at once: that destroys the comparison and can turn a diagnostic queue into uncontrolled account work.

Tools that help with this

Every paid output is a draft for your review before you publish, appeal, or activate anything. Friday does not access your seller account.

More seller guides

Browse all guides →