Adaptive Search
Adaptive Search personalizes search results for each visitor. Rank and filter results based on predicted intent and affinities.

What adaptive search does
Standard search treats all visitors the same. Adaptive search customizes results:
Rank by intent - Predict what visitors want to find
Boost by affinity - Highlight products matching interests
Filter by context - Show relevant subsets
Personalize facets - Sort filters by relevance
How it works
Visitor signals - AdaptiveCX collects behavior, affinities, and predictions
Search query - Visitor enters a search term
Real-time ranking - Results scored and ranked for this visitor
Personal delivery - Results shown in optimal order
Adaptive search list
Track performance of each search configuration:
Status - Active or archived
Uplift - Improvement vs baseline
Sessions - Total searches
Conversion impact - Effect on purchases
Tabs
Ongoing Adaptive Searches - Active configurations
Archived Adaptive Searches - Inactive configurations
Create adaptive search
Configure basic settings
Set up the search configuration:
Name - Give your search a descriptive name
Schedule - Choose when to start (immediately or scheduled)
Live QA - Enable quality assurance testing
In the What To Show section:
Number of suggestions - How many results to display (default: 7)
Prioritize by - Most recent, relevance, or custom rule
Categories & Keywords - Select affinity categories to boost
Colors - Enable color-based recommendations
Brands - Enable brand-based recommendations
In the When To Show section:
Until visitor types - Show suggestions until visitor types N characters
Always show - Always display personalized suggestions
Set targeting rules
In the Who To Target section:
Traffic Source - Target visitors from specific sources
Device - Target by device type (mobile, desktop, tablet)
Audience - Target specific audiences
Click Edit to configure targeting conditions.
Use Terms to Promote/Exclude to specify keywords that should always be boosted or hidden.
Configure test/control split
Set the percentage split between test (personalized) and control (baseline) groups:
Move slider to allocate traffic
100% test means all visitors see personalized results
Lower percentages create a control group for comparison
Add custom code (optional)
Click Next to access the Custom Code tab:
Add JavaScript for custom ranking logic
Implement client-side personalization hooks
Integrate with third-party search providers
Save and publish
Click Save as Draft to save without activating, or publish to go live.
Last updated
Was this helpful?


