Measure what your agent resolves
See your agent’s resolution rate, the quality of its citations, the knowledge gaps it hits, and readiness scores that tell you what to fix next.
Your support metrics at a glance
Total searches
AI answers given
Success rate
Tickets deflected
Example dashboard metrics
Insights that drive improvement
Data-driven documentation gets better results
Search analytics
See exactly what customers search for. Understand their language and anticipate their needs.
Article performance
Which articles get viewed most? Which lead to further questions? Optimize based on real usage.
Knowledge gaps
Every question the agent can’t answer confidently rolls up into a knowledge gap, so you can fill it before more customers hit it.
Resolution rate
See what share of conversations your agent resolves autonomously versus escalates to a human — the number that tracks real deflection.
Citation quality
See which sources ground your agent’s answers and how often replies are backed by curated content — no uncited confidence slips through.
Readiness scores
Three readiness gauges — answer, sales, and agent — score how ready your agent is and hand you the next best action to raise them.
Turn insights into action
Find content gaps
See what customers search for but don't find. Create content that matters.
Improve weak articles
Identify articles with low satisfaction or high bounce rates.
Prove ROI
Show stakeholders exactly how many tickets self-service deflects.
Optimize AI
Track AI performance and improve responses over time.
Frequently asked questions
HelpShelf tracks search queries, article views, AI answer performance, content gaps (searches with no results), user navigation patterns, and estimated ticket deflection. All data is presented in an easy-to-understand dashboard.
The analytics dashboard shows searches that returned no results or low-confidence AI answers. These are content gaps—topics customers ask about but your documentation doesn't cover well. Filling these gaps improves self-service rates.
Yes! HelpShelf shows AI answer confidence scores, customer feedback on AI responses (when enabled), and the percentage of questions AI handles vs routes to humans. You can see exactly how automation is performing.
Ticket deflection is estimated based on successful self-service interactions—searches that led to article views without subsequent support contacts, and AI answers marked as helpful. It's an estimate, but a useful one for measuring ROI.
Basic analytics are included on all plans, including Free. Advanced analytics—including detailed search queries, AI performance tracking, and export capabilities—are available on paid plans.