Explore demand through queries and individual words
Review frequent searches, individual terms and changes from the previous period. A word appearing in several searches is not the same as a confirmed purchase. Queries sent while someone types may also count; reports explain that distinction.
Find zero-result searches and results without clicks
Both deserve attention, but they mean different things. Empty results can come from filters or an incorrectly interpreted name. Results without clicks can suggest poor relevance, but they do not prove a lost sale. Check the catalog before making inventory decisions.
Understand which products get attention
Explore rendered cards, observed views, clicks, click position and recorded cart additions. Loading a card does not prove a visitor saw it: new tracking distinguishes cards observed with at least 50% of their area visible.
Follow shopping journeys and the funnel
Review query reformulations, recovery after empty results and search → click → cart → purchase sequences. The funnel requires events in order within the same session and reporting period. It shows observed association, not proof of incremental revenue.
Work with your data
Filter by period and device, check activity by UTC hour, sort tables and export CSV. Reports can cover up to 365 days of retained data. New tracking does not reconstruct missing history, and different currencies are reported separately.
Build a practical improvement routine
Start with recurring failed terms. Check filters, stock and synonyms, then review products that appear but receive no clicks. After a change, compare equivalent periods and record campaigns or traffic changes that may affect the results.