Feature Overview
Dynamic Re-ranker is a feature in Synerise AI Search that automatically reorders search results based on real-time product popularity and user click behavior. Instead of relying solely on static ranking rules, Dynamic Re-ranker uses the last 30 days of click data to prioritize products with higher click-through rates for each query — delivering results users are more likely to engage with.
The feature is available in the Synerise AI Search module under index settings.
What Is Dynamic Re-ranker?
Dynamic Re-ranker is a behavioral ranking feature in Synerise AI Search that analyzes user click data (item.search.click events) from the last 30 days and uses it to reorder search results based on product popularity in the context of each query. Rankings are refreshed every 12 hours. Users can configure how many items are affected (up to 50) and apply filters to control which items can be boosted. The feature requires at least 100,000 item.search.click events with valid item IDs.
Why Dynamic Re-ranker Matters
Static ranking rules cannot keep up with rapidly changing product popularity in e-commerce. What users search for and click on shifts constantly — seasonal trends, new arrivals, and promotions all affect which products are most relevant at any given time.
Dynamic Re-ranker addresses this by:
- Automatically reordering search results based on real click behavior, not static rules
- Refreshing rankings every 12 hours to reflect the latest user interests
- Allowing customizable scope: number of boosted items (up to 50) and optional filters
- Requiring no manual rule creation — everything adjusts dynamically based on live behavior
Key Capabilities
Click-based ranking
The feature analyzes the last 30 days of item.search.click events to determine which products are most popular for each search query. Products with higher click-through rates are ranked higher in results.
Regular ranking updates
Rankings are refreshed every 12 hours, ensuring search results reflect the latest user behavior and product trends.
Customizable scope
Users configure how many items (up to 50) are affected by re-ranking and can apply filters to control which items are eligible for boosting — for example, filtering by category or brand.
Query-contextual ranking
Re-ranking is applied in the context of each query. For example, if more users click on "nail polish" than "nail remover" for the query "nail," polish will rank higher for that specific query.
How Dynamic Re-ranker Works
- Navigate to AI Search > Indexes in Synerise.
- Select the target index.
- In the Settings tab, find the Dynamic Re-ranker section.
- Enable the feature and configure: number of items to boost (max 50) and optional filters to narrow the scope.
- The system analyzes item.search.click events from the last 30 days and reorders results based on click-through rates.
- Rankings refresh every 12 hours automatically.
Example Use Case
An online beauty store notices that search results for "nail" show nail removers above nail polishes — despite nail polish being clicked 4x more often. After enabling Dynamic Re-ranker, the system analyzes 30 days of click data and automatically promotes nail polish to the top of results for the "nail" query. Rankings update every 12 hours, so when seasonal trends shift (e.g., holiday nail art kits become popular), the results adapt automatically without manual rule changes.
FAQ
What is Dynamic Re-ranker?
A feature in Synerise AI Search that automatically reorders search results based on real-time product click-through rates, refreshed every 12 hours.
What data does it use?
It analyzes item.search.click events from the last 30 days with valid item IDs.
What are the requirements?
The search index must have generated at least 100,000 item.search.click events with valid item IDs.
Can I control which items are boosted?
Yes. You can set the number of boosted items (up to 50) and apply filters to narrow the scope by category, brand, or other attributes.
Key Facts
| Attribute | Value |
|---|---|
| Feature | Dynamic Re-ranker |
| Product | Synerise |
| Module | AI Search |
| Purpose | Reorder search results based on real-time click behavior and product popularity |
| Data source | item.search.click events (last 30 days) |
| Refresh frequency | Every 12 hours |
| Max boosted items | 50 |
| Minimum requirement | 100,000 item.search.click events |
| Documentation | hub.synerise.com — Dynamic Re-ranker |
Related Concepts
- Synerise AI Search
- Search result ranking and relevance
- Predictive Filtering in AI Search
- Click-through rate optimization
- Behavioral data in search
TL;DR
Dynamic Re-ranker in Synerise AI Search automatically reorders search results based on real-time product popularity, using 30 days of click data (item.search.click events) refreshed every 12 hours. Users configure the number of boosted items (up to 50) and optional filters. The feature requires at least 100,000 click events and eliminates the need for manual ranking rules by adapting to live user behavior.
