How to Use SearchStax Analytics to Improve Search Relevance

Updated

Where this fits in the Get Started sequence: This is article 3 of 5. Use it after Why Analytics Matters for Relevance and Understand Analytics at the App, Profile, and Language Levels.

Use SearchStax Site Search analytics to make relevance decisions with real user data. This guide gives you a repeatable process to improve click-through rate, reduce no-result searches, and track whether your changes are working.

Before You Start

Before you begin, make sure your search UI is live and collecting user data.

  • Your Search App should be active with a working search experience.
  • You should have enough analytics data to spot meaningful patterns.

Tip: If profile-level analytics looks incomplete, revisit Understand Analytics at the App, Profile, and Language Levels in this section before you run this workflow.

Step 1: Set the Scope on the Dashboard

Open the Dashboard first. It's the best place to start when you want to understand search performance.

  1. Select the Search App you want to review.
  2. If your App has multiple languages, select the target language.
  3. Choose either All Profiles or a specific profile.
  4. Set the date range so you're comparing a meaningful period.

Tip: Analytics tables and graphs refresh hourly. Wait for the next refresh window before you judge the impact of a recent change.

Note: If a selected profile shows zeros, switch to All Profiles to confirm overall activity. Then ask your technical team to verify that tracking includes model (Search Profile name).

Metrics to Watch in This Workflow

Use this smaller set of metrics while you work so you stay focused on the signals that matter most for relevance:

  • Total Searches: Confirms demand and sample size for your comparison window.
  • Click-Through Rate: Shows how often searches lead to at least one click.
  • % No Results: Shows how often users get no results.
  • Searches with No Clicks: Shows where users saw results but didn't engage.
  • Average Click Position: Shows whether useful results appear near the top.
  • Average Search Latency: Adds performance context when engagement shifts.

Step 2: Review Core Relevance Signals

Use the Dashboard summary metrics and charts to see where relevance is helping users and where it isn't.

  • Click-Through Rate
  • No Results Searches
  • Average Click Position

If you need deeper definitions while you review trends, use the Analytics Glossary and Why Analytics Matters for Relevance in this section.

Step 3: Investigate the Cause with Analytics Pages

Once you identify a trend on the Dashboard, use the analytics pages to diagnose the issue.

  • Searches: Review top searches, no-result searches, and searches with no clicks to find high-impact query gaps.
  • Content Items: Compare impressions, clicks, and click-through rate by item to see what content is underperforming.
  • Search Feedback: Review ratings and comments to understand user-reported quality issues.
  • Custom Reports: Build focused reports and export results for deeper analysis.

Quick Map: Which Analytics Page to Use

Check Tracking Before You Trust Zeros

  • Profile-level reporting: Track model so profile dashboards populate correctly.
  • Language-level reporting: Track language when you need language-specific analysis.
  • No-results accuracy: Track totalHits as 0 for no-result searches.
  • Latency reporting: Track latency to calculate average search latency.
  • Item-level reporting: Track both impressions and clicks for reliable Content Items analysis.

Step 4: Apply the Right Relevance Lever

Match each pattern to the relevance feature most likely to improve it.

  • Users search with different words for the same concept: Add or refine Synonyms.
  • You need search results to change for specific queries: Use Rules.
  • You need tighter ordering control for important content: Tune Ranking.
  • Users miss expected results because of spelling issues: Improve Spell Check.
  • Users need better follow-up discovery: Refine Related Searches and Auto-Suggest.
  • You want to answer direct questions faster: Evaluate Smart Answers.

Note: If users repeatedly search for important topics and get no results, create or publish content for those topics, then tune your search settings.

Step 5: Measure Impact and Repeat

Relevance tuning is iterative. After each change:

  1. Wait for the analytics refresh.
  2. Recheck the same metrics, charts, and date window.
  3. Keep the changes that improve outcomes and continue refining.

Over time, this process helps you build a search experience that reflects how your audience actually searches.

Next Steps

Next, read Levers You Can Use to Influence Relevance (article 4 of 5). Then continue to Understand Cookies and Sessions (article 5 of 5) before you move into Use Searches to Find High-Impact Queries and Content Gaps.

Articles in this section