Release Summary 25.11 | May 29, 2025
The following key features and improvements, along with bug fixes, have been released in Algonomy CXP products in the release version 25.11.
Enterprise Dashboard
Introducing Product Comparison Configuration
Product Comparison Configuration is now officially available and can be accessed from the ‘Optimization’ section in the navigation. This new feature makes it easier for users to manage and customize product comparison experiences.
The configuration allows users to define comparison rules at the category level, specifying which product attributes should appear in the comparison table. This ensures that shoppers see the most relevant information based on the type of products they are viewing. The setup works seamlessly with the Product Comparison Dynamic Experiences template, enabling a personalized and informative comparison experience tailored to user intent.
Jira: ENG-30274
Improved User Affinity Scoring with Dynamic and Combined Attribute Configuration
The User Affinity Configuration has been upgraded to offer greater control and accuracy in modeling shopper preferences. The attributes table now supports merging custom attributes using drag-and-drop, allowing affinity scores to be calculated based on products that match both attribute values. This helps capture more specific and relevant shopper behavior. Brand and Category rows remain fixed, while new attributes can be added through an input row with auto-suggest and support for multiple entries using semicolon-separated values.
By combining two custom attributes into a single row, users can better reflect real-world preferences—for example, identifying affinities based on combinations like Color and Gender or Region and Varietal. The system also supports scoring with AND logic, where affinity is counted only when both attribute values are present in a product. These enhancements make affinity scoring more flexible, precise, and aligned with actual user behavior.
Jira: ENG-30006, ENG-29988, ENG-29843
Autosuggestions for User IDs and Attributes in User Affinity Configuration
To streamline workflows and improve usability, the User Affinity Configuration page now supports autosuggestions for both user IDs and attributes. This enhancement helps optimization managers quickly access previously used inputs without needing to retype them.
When entering a user ID, the system now suggests values based on past usage, shared across relevant areas like the User Profile and Affinity Configuration pages. This reduces manual effort and enables faster access to affinity scores and configuration setup.
Jira: ENG-29510
Custom Attribute Support for Primary Category Assignment
Users can now assign a product's primary category based on a custom attribute, allowing better alignment with internal categorization rules. This new option is available on the Primary Category Configuration page and offers greater flexibility for organizations that maintain their own category structures.
When enabled, the interface hides default category selectors and instead provides an input field with auto-suggest to help users choose an existing attribute. This enhancement ensures that the assigned primary category reflects the organization’s preferred classification method.
Jira: ENG-30146
Recommend
B2B Visibility Filtering Now Supported Across Recommend
Recommend now supports B2B visibility filtering across all APIs, including recsForPlacements, p13n.js, and recommendations in email. This enhancement allows merchandisers to apply availability and pricing rules based on user-specific visibility settings, improving control and consistency in B2B personalization.
When enabled through site configuration, the system filters products using parameters such as visibility=true, visibilityGroups, userContexts, and lang, with values supplied via customer instrumentation. Filtering is applied before strategy selection, and fallback strategies are used when all products are excluded. This ensures that only eligible products are recommended while maintaining a seamless experience across all strategy types and delivery channels.
Jira: ENG-29794, ENG-28660, ENG-29978
Ensemble AI
Ensemble AI: Merchandising Report with Visualizations
A new merchandising report has been introduced in the UI to help merchandisers analyze shopper interactions with products, brands, and categories from ensembles. The report provides key metrics such as attributable sales, CTR, clicks, items from ensembles, overall items, page views, and sales. Users can filter the results by product, brand, category, region, and currency.
The report includes bar charts to highlight top-performing entities and line graphs to show trends over time. A detailed table view is also available, enabling users to explore all metrics across selected dimensions and date ranges, with support for data download.
Transparent Image Configuration for Ensemble AI
Added support to configure transparent image generation for Ensemble AI from the model options page. Merchandisers can now enable or disable the job and select product categories for which transparent images should be created. The Ensemble AI API returns the image URLs with transparent backgrounds, using a backend table to track generated images, which is now accessible from the front end.
Jira: ENG-29574
Other Feature Enhancements
The following feature enhancements and upgrades have been made in the release version 25.11.
|
Jira # |
Module/Title |
Summary |
General Availability |
|---|---|---|---|
|
Enterprise Dashboard: Support for A/A Testing in Social Proof MVT |
The Social Proof MVT configuration now allows users to create A/A tests by exposing the "Enable Social Proof API Service" option for both control and treatment. Users can manually configure and edit this setting, enabling consistent behavior across both groups when needed. |
29-May-25 |
|
|
Enterprise Dashboard: Optimized Social Proof API Calls for Single Experience |
Social Proof now avoids making multiple API calls when only one message is displayed on a page. The update consolidates API calls into a single request per page load, ensuring more accurate reporting and improved performance. |
29-May-25 |
|
|
Data Engineering: Added Ensemble AI API Report Visualization IDs to Production |
The visualization IDs for the Ensemble AI API report have now been added to the production consul. This ensures the IDs are available for use in the production environment. |
29-May-25 |
|
|
Science, Ensemble AI: Support for Generating Transparent Product Images in Ensemble AI |
A new enhancement enables generation of product images with transparent backgrounds to support visually engaging collages and creative merchandising. A configurable job can now be scheduled to process product images by site and category, generating transparent versions stored in CacheFly for use in shopper-facing content and Active Content. This enhancement ensures images are only generated when needed and avoids duplication by checking for existing transparent image attributes. Merchandisers can configure the job from the model options page, monitor image generation status, and rely on fallback logic when transparent images are not available. |
29-May-25 |
|
|
Science, Ensemble AI: Shopper Feedback Integration for Dynamic Ensemble Ranking |
Ensemble AI now incorporates shopper feedback to refine how outfits are ranked and displayed. When shoppers interact using "more like this" or "less like this," the system applies a Feedback Score to adjust the ensemble order, enhancing personalization based on real-time preferences. The Feedback Score is used alongside Ensemble and User Affinity scores to re-rank results, and API responses reflect this updated order when relevant. Additionally, feedback-only updates can be processed without altering the visible experience, ensuring seamless integration with existing flows. |
29-May-25 |
|
|
Find: Updated Default to cxp-gl-j11 for Solr 9 Stack |
With all clients now migrated to the Solr 9 stack, cxp-gl-j11 has been set as the default. The older cxp-gl version will only be used if explicitly configured in the portal site configuration.
|
29-May-25 |
Bug and Support Fixes
The following issues have been fixed in the release version 25.11.
|
Jira # |
Module/Title |
Summary |
General Availability |
|---|---|---|---|
|
Enterprise Dashboard: Display and Configuration Issues in Product Comparison DynEx Template |
We have fixed several issues in the Product Comparison DynEx template. The placement now renders inline when added within a specific HTML element, and the attribute order in the comparison table correctly follows the recsForPlacements response. The preview product input field has been added, and the fallback base URL in the JavaScript has been updated to recs.richrelevance.com. |
29-May-25 |
|
|
Recommend: Encoding Issue with Accented Characters in DSW Keys |
We have resolved an issue where accented and capital characters in DSW strategy keys were not properly encoded, resulting in missing recommendations. The model files now retain accented characters correctly, ensuring that recommendations are returned as expected for all valid search terms. |
29-May-25 |
|
|
|
Enterprise Dashboard: Environment Display in Recs Test Drive |
The 'View recs' button now correctly reflects the active environment in QA and Staging, and the recsForPlacements call points to the appropriate endpoint. The 'View recs in Integration' button is disabled in non-production environments. |
29-May-25 |
|
Attribute Selection Issue in Configurable Strategies
|
We have resolved a UI issue that prevented attribute seeds from being saved when used with user history in Configurable Strategies. Attributes like product_color and style now save and display correctly when configured with user history.
|
29-May-25 |
|
|
Find: Duplicate Category Issue in FIND Facets |
We have resolved an issue where duplicate categories appeared in FIND responses due to category name conflicts within the same hierarchy. The system now properly handles updates when multiple category IDs share the same name, ensuring accurate category and facet data. |
29-May-25 |
|
|
Inconsistent Recommendation Behavior in Cart Page Placement |
We have resolved an issue where a cart page placement was returning fewer products than configured due to unexpected behavior in a boosting rule. The placement now correctly returns recommendations based on manual and strategy rules, adhering to the defined minimum product count. |
29-May-25 |
|
|
Issue with Multiple Placements in App Channel |
Resolved an issue where certain API calls using multiple placements in the App channel were intermittently failing. The failure occurred under specific conditions such as reuse of the same strategy across placements or conflicts between incompatible strategies. The issue has been addressed and the calls now execute as expected across all placement configurations. |
30-May-25 |