Enterprise Search Metrics: A Framework for Measuring User Experience
When we talk about measuring the quality of search experience, we usually refer to either direct user feedback (qualitative) or search engine statistics like recall and precision (generic benchmarks). But enterprise search is actually an integrated ecosystem of content, processes, applications and usage patterns. Measuring the right aspects of search performance requires a holistic view of the six major focus areas that collectively make up sustainable search integration and experience:
- Tagging components and processes - e.g. auto-tagging vs. human indexing vs. social tagging percentages
- Content to taxonomy coverage - e.g. percentage of content tagged "General", "Other", or "Miscellaneous
- Search terms to taxonomy fit - e.g. percentage of the top 80% of most frequently searched terms present in taxonomy terms and synonyms
- Search engine tuning - e.g. relative percentage of click throughs per data source
- UI for search term expression - e.g. percentage of queries using faceted semantic refinement
- UI for search results visualization - e.g. percentage of queries using drill-down through clusters
Defining, capturing and analyzing trend data for key search experience metrics can take the guesswork out of deciding which improvements to make first, measuring the impact of each improvement, and fine-tuning the enterprise search ecosystem as content and usage interests change over time.
Speakers
