Site search analytics

Description

Analyse the queries people type into a site's search to learn their vocabulary, find what they cannot find, and fix labels, content and results, starting with the most frequent queries.

Search queries are users' own words for what they want and could not reach by browsing. Query frequency usually follows a power law, so a small head of phrases accounts for much of all searching and is optimised first. Core metrics are zero-result rate, click-through, refinement and search exits; reading top queries reveals acronyms, jargon, product names and mismatches with navigation labels. Actions include adding synonyms and preferred terms, renaming labels, writing missing content and curating Best Bets for top queries. Comparing internal queries with external search-engine queries exposes SEO and IA gaps. More rigorous variants categorise queries into need-based groups with weighted conversion metrics, or follow Jansen's three stages: collection, cleaning (removing corrupt records, separating page requests from new queries, choosing a session cut-off) and analysis at term, query and session levels.

Attributes

Loss type
Active
Effort
Days

Steps

  1. Ensure search logging captures query, timestamp, session or user identifier, result count and clicks, and arrange regular access.
  2. Export a representative period, clean corrupted records, and separate results-page requests from new queries; choose a session cut-off (studies use 5 to 120 minutes).
  3. Rank unique queries by frequency with cumulative percentages, grouping misspellings and variants.
  4. Compute zero-result, click-through, refinement and exit rates for the head queries.
  5. Read the top queries for vocabulary, acronyms and label mismatches; optionally categorise queries by need and attach weighted conversion metrics per category.
  6. Act: add synonyms, rename labels, create or fix content, and set Best Bets for top queries; compare internal with external queries for SEO gaps.
  7. Report key metrics monthly on one page and repeat the analysis regularly; use interviews or usability tests to explain anomalies.

Sources

  • Universal Methods of Design (Martin & Hanington, 2019)
  • UX For Beginners (Marsh, 2016)
  • Information Architecture (Rosenfeld et al., 2015)
  • Search Patterns (Morville & Callender, 2010)
  • Practical Web Analytics for User Experience (Beasley, 2013)
  • Understanding Audiences via Analytics (Jansen et al., 2024)

Also known as

  • SSA
  • Search log analysis
  • Site search log analysis
  • On-site search query analysis
  • Internal search logs
  • Search analytics
  • Query log analysis
  • Best Bets
  • Search keyword analysis
  • Search term categorisation
  • Discount keyword analysis
  • Transaction log analysis of search
  • SLA

Categories

  • Content
  • Information Architecture & Content