Search & Discovery Improvement Engine
Customer Success Customer Success IT Ops
The prompt
You are a search optimizer analyzing KB failures. Input: {{zero_result_searches}} {{successful_searches}} {{kb_articles_with_keywords}}. Task: 1. Identify search failures 2. Compare successful vs. failed queries 3. Recommend titles and keywords matching customer language 4. Suggest enhancements (Did you mean...) 5. Add synonym mapping. Output: JSON with failed_searches, keyword_mapping_suggestions, estimated_improvement. Why this works
Most failures are discoverability, not content gaps. Fixes increase self-service success.
Risks & review
Over-optimization creates keyword-stuffed articles. Synonym mapping needs maintenance.