Site search audit

A site search audit is a one-time expert review of how the internal search works from the visitor's perspective: is the search box noticeable, convenient and easy to find, are there suggestions and autocomplete, does the search forgive typos, word forms and synonyms, how relevant are the results and is the sorting clear, is "nothing found" handled correctly (are there hints on what to do next rather than an empty screen), can the results be refined with filters, are price and availability shown clearly in the results, and is everything convenient on a phone. Honestly about the main point: this is diagnosis, not design, development or search-engine configuration; and it is not a guarantee of a rise in conversion or sales. Search is only one of the paths: people also find products through the catalog, through filters and from ads. So the recommendations are reasoned hypotheses to test, not "we will raise conversion by X%". The relevance assessment is partly expert judgment (typical queries we check by hand), and it is more accurate with access to your search-query logs (what people actually search for and where they get empty results) — without such access some problems remain unseen (for example, we will not see for which real queries you most often get "nothing found"). It is a snapshot in time: the catalog and the search change. You get a report with prioritized problems and recommendations; configuring the search engine, synonyms, design and development are done separately or by your team. An important boundary: this is an audit of search specifically, not an audit of catalog filters (how results are narrowed — separate) and not an audit of product cards (what a person sees once they reach the product — separate) — those are adjacent but separate services. If the catalog is very small, there is no site search, or it is almost never used, there is almost nothing to assess and you most likely do not need the audit. Picture this: instead of "people search the site and leave" you learn that the search box hides behind an icon, a one-letter typo gives an empty screen, a synonym finds nothing, and the results show no price — and you see what to fix first. The base price starts from 12,000 ₽; it depends on the size of the catalog and the complexity of the search.

• People use the search but do not find what they need and leave.
• A typo or a synonym often gives empty results.
• Search results seem irrelevant or are sorted unclearly.
• You are not sure the search is convenient on a phone.

• Noticeability and convenience of the search box
• Suggestions and autocomplete
• Tolerance to typos, word forms and synonyms
• Relevance and clarity of result sorting
• Handling of "nothing found" (hints instead of an empty screen)
• Refining results with filters; price and availability in results
• Convenience on a phone; query-log analysis when access is available
• A prioritized report with hypotheses to test

• Clear understanding of why people fail to find things via search
• Priorities: what to improve in the search first
• Understanding how typos, synonyms and empty results are handled
• Understanding where the problem is in the search and where in the catalog or assortment

• We clarify the site, typical queries, goals and collect access to the site and search logs
• We test the search by hand on typical and problem queries and review the results
• We prepare a report with priorities and hypotheses and review it with you

• Will sales grow after the audit?

  The audit itself does not raise them. Search is only one of the paths to a product (also the catalog, filters, ads). We find the weak spots and give hypotheses, but what will work is verified by tests; growth is possible only after implementation and is not guaranteed.

• Will you configure the search, synonyms and engine?

  No, that is separate work. The audit gives diagnosis and hypotheses: what to improve and why. Configuring the search engine, synonym dictionaries, design and development are done separately or handled by your team.

• Do you need access to search-query logs?

  It is desirable. With it you can see what people actually search for and where they get empty results — the assessment is more accurate. Without access it is an expert review on typical queries; some problems will remain unseen without logs.