Release Summary 25.13 | Jun 26, 2025
The following key features and improvements, along with bug fixes, have been released in Algonomy CXP products in the release version 25.13.
Recommend
Advanced Merchandising: Export and Import Rules in JSON Format
Managing Advanced Merchandising rules across sites is now easier with the new Export/Import Rules feature. The previous “Upload Rules” button has been replaced with a combo button labeled ‘Export/Import Rules’, providing more flexible options for exporting and importing rule files.
Users can choose to export production-enabled rules (default), all rules, or only selected rules. The exported file uses the same JSON format as the upload process, making it simple to transfer rules between sites. The export also supports recent enhancements including system filters, deduplication, translated rule messages, backfilling recommendations, and variety settings. The import flow remains unchanged, allowing users to upload rule files via the existing file picker.
Advanced Merchandising: Edit Rules with Preview Turned Off for Large Catalogs
A new mode has been introduced to help users manage Advanced Merchandising rules more efficiently on sites with large catalogs. When the "Enable large catalog mode" setting is turned on in Site Configuration, users can create and edit rules without triggering the preview functionality by default. This helps avoid delays and performance issues that may occur during rule editing on high-volume catalogs.
In this mode, preview actions like fetching seed products or recommendations are performed only on demand, using a simplified interface. Users can still fully edit and save rules without being impacted by preview-related processing.
Recommend and Engage
User Affinity: Support for Browsed Category Events
User Affinity configurations now include a new event type, Browsed Category, to better capture shopper intent during their browsing journey. This enhancement ensures that categories a user explores before interacting with a product are factored into affinity scoring, rather than assigning equal relevance to all categories linked to a product.
By incorporating signals from browsed category pages, the system provides more accurate affinity scores that reflect real user behavior. New configuration fields have been added to specify scores and lookback periods for both recent and older browsed categories. These updates are available in both new and existing affinity configurations and are reflected in the affinity preview interface.
Auto-Suggest for User IDs in Test Drives
To speed up the testing process, auto-suggest has been added for user IDs in Recommend and Content Test Drives. When entering a user ID, a list of previously used IDs now appears, allowing for quicker selection and reuse.
This list is shared across related areas like User Profile, User Affinity Configuration, and Test Drives, making it easier to preview personalized results for known users.
Jira: ENG-30369
Engage
Enhanced Experience in Content Test Drive
The Content Test Drive has been improved to provide clearer visibility into how content is delivered and displayed. The Content tab now shows the image associated with each content item based on the configuration defined in the site settings. The displayed label reflects the content's campaign value, and if available, each content name is linked directly to its corresponding catalog entry for easy access.
Additionally, a new Campaigns tab has been added. This tab displays the campaign associated with each placement and includes a direct link to the campaign setup page in the portal. These enhancements make it easier to validate, explore, and troubleshoot content experiences in real time.
Jira: ENG-30282
Click Tracking Added to Dynamic Experiences templates for Category Recs, Tabbed Recs, Bundle Placement, and Comparison Placement
Click tracking is now supported in the Dynamic Experiences templates for various experiences, providing better visibility into shopper engagement. This enhancement applies to Category Recommendations, Tabbed Recommendations, Product Bundle Placement, and Comparison Placement.
Clicks are tracked on the outer product elements, and for Product Bundles, also on the “Add bundle to basket” button. This helps measure the effectiveness of each experience more accurately.
Jira: ENG-30350
Channel Targeting in Dynamic Experiences Based on User Agent
Dynamic Experiences now automatically detects the shopper’s channel—Desktop, Tablet, or Phone—based on the user agent, instead of relying solely on the API client key. This makes it easier to target configurations like Social Proof for specific devices without requiring additional setup. The update applies to all sites and works out of the box.
Jira: ENG-30142
Discover
Discover Workbench Now Uses the Updated Find Discover API
The Discover Workbench has been updated to use the latest Find Discover API, replacing the older PCS Simulate API. This change ensures that the product ordering and global weight metrics are now sourced directly from the new API, providing a more consistent and accurate testing experience.
To support the updated API, the workbench now includes additional input fields for required parameters such as Placement, Channel, and Language. These inputs are saved locally once selected, streamlining future usage. Users can also reveal or adjust these settings using a "View more" option. A Region input has been added when applicable for sites that support regional configurations.
Jira: ENG-29295
Front-end Integration
Simplified User Context Configuration for p13n.js
The p13n.js library has been updated to simplify how user context values are configured. Previously, setting user contexts required providing both a key and a value. With this update, the R3_COMMON.setUserContexts() method now accepts a simple value instead of a key-value pair. This change streamlines configuration for customers who use single-value contexts.
For example, instead of calling R3_COMMON.setUserContexts('userContext_key1','userContext_value1'), you can now simply use R3_COMMON.setUserContexts('userContext_value1'). Multiple calls to the method are still supported, and the values will be combined and passed as a comma-separated list in the p13n_generated.js request.
Jira: ENG-30477
Other Feature Enhancements
The following feature enhancements and upgrades have been made in the release version 25.13.
|
Jira # |
Module/Title |
Summary |
General Availability |
|---|---|---|---|
|
Recommend: Limit on Product Count in Discover Requests |
A new site configuration allows setting a maximum number of products returned in Discover requests. If a request exceeds this limit, only the configured maximum is returned. This helps manage performance for high-volume requests, especially when compression is not used. Metrics are also logged to monitor usage and identify sites where the cap should be enabled. |
26-Jun-25 |
|
|
Enterprise Dashboard: Low Code JS Enhancements |
The low-code JavaScript setup now defaults to using the latest version of p13n.js, with an option to select older versions if needed. Script delivery has been moved to Cloudflare, and the latest p13n.js content is now used dynamically through configuration for better flexibility and maintenance. |
26-Jun-25 |
|
|
Engage: Carousel support added in DynEx chatbot templates |
We’ve added carousel functionality to both the custom and default chatbot templates in the DynEx experience. This enhancement allows product recommendations to be displayed in a carousel layout, improving engagement and usability. |
26-Jun-25 |
|
|
Enterprise Dashboard: Report visibility now based on enabled features per site |
The reports interface now displays only the reports relevant to the site's enabled features. For example, if Engage is not enabled for a site, all Engage-related reports will be hidden. This streamlines the experience by reducing clutter and improving focus. Additionally, all available reports are now sorted alphabetically for easier navigation. |
26-Jun-25 |
|
|
Ensemble AI: View and CTR metrics added to Ensemble AI reports |
View and click-through rate (CTR) metrics have been added to Ensemble AI reports, including Style Performance, Outfit Performance, and Merchandising. These metrics help measure engagement by tracking how often styles and recommendations are viewed and clicked. Time series visualizations are also available for deeper insight. |
26-Jun-25 |
|
|
Ensemble AI: Updated placement for Ensemble AI recommendations |
Ensemble AI recommendations can now be displayed just below the product image, aligning with client preferences. This update ensures the curated ensembles appear prominently at the top of all recommendations, improving visibility and user engagement. |
26-Jun-25 |
|
|
Find: Numeric search queries excluded from query writer job |
The query writer job now filters out purely numeric search queries and queries shorter than four characters when retrieving data from the database. This enhancement improves the relevance of queries used in vector search processing by eliminating low-value inputs. |
26-Jun-25 |
|
|
Find: Site-specific schema support enabled in batch builds |
Batch index builds now support site-specific schema configurations by incorporating the use of ManagedIndexSchemaFactory and adding schema payload support in the Find deployer. This enhancement enables greater flexibility by allowing customized schema changes for features like autocomplete, links, boosts, and merchandising rules across different sites. |
26-Jun-25 |
|
|
Find: API added for language resource cache invalidation in Find deployer |
API added for language resource cache invalidation in Find deployer A new API is now available to invalidate language resource cache entries by key. This helps resolve issues where cache is created before the language resource container is ready, allowing deployment retries to load the correct data without waiting for the default cache expiration. |
26-Jun-25 |
|
|
Find: AWS VA support added to DC comparison report |
The DC comparison report now includes data from AWS Solr hosts by reading full hostnames from the configuration. This update ensures that the dc_doc_stats_report job reflects a complete view across both on-prem and AWS VA environments. |
26-Jun-25 |
|
|
Find: Cluster pool parameter added to search service metrics |
The find.searchservice.api.latency metric now includes a new parameter for the cluster pool. This addition enhances observability by allowing latency data to be segmented and monitored by cluster, aiding in performance analysis and troubleshooting. |
26-Jun-25 |
|
|
Enterprise Dashboard: Chatbot now filters out deleted products from recommendations |
The chatbot experience has been improved to automatically exclude products that have been deleted from the catalog. Products lacking essential information like images or prices are now filtered out from the display list, ensuring users only see valid, purchasable items. |
26-Jun-25 |
|
|
Improved handling of non-scoped ingests during ProductMatch scoped actions |
The ProductMatch scopedAction logic has been enhanced to prevent unintended deletions of items ingested outside the scoped context during an active scoped action. Now, non-scoped ingests occurring within the scoped action’s execution window for configured sites are retained, ensuring data integrity. This refinement helps maintain the correct item state throughout the scoped action lifecycle while allowing unrelated processes to function as usual. |
26-Jun-25 |
|
|
Find, Science: Enhanced delta handling in catalog vector updates |
The catalog embedding process has been improved to detect and respond to changes in product attributes such as category, brand, and other metadata. Now, whenever any relevant attribute is updated, the affected products are automatically included in the next catalog vector generation cycle, ensuring embeddings stay accurate and up to date. |
26-Jun-25 |
|
|
Discover, Find: Search/Browse menu now hidden when not applicable |
The Search/Browse menu item is now automatically disabled if neither Discover nor Find is enabled for a site. This streamlines the navigation by removing unused options and keeping the interface relevant to each site’s configuration. |
26-Jun-25 |
Bug and Support Fixes
The following issues have been fixed in the release version 25.13.
|
Jira # |
Module/Title |
Summary |
General Availability |
|---|---|---|---|
|
Recommend: Cart items filter excluded for cart products model |
We have fixed an issue where the "don't recommend cart items" filter was incorrectly applied to strategies using the cart products model. This caused expected recommendations to be excluded. The filter is now bypassed when the cart products model is used, ensuring the correct products appear in the recommendations. |
26-Jun-25 |
|
|
DSW, Recommend: Incorrect display of schedule time during edit in Data Science Workbench |
We have fixed an issue where scheduled times appeared in reverse order when editing a strategy. Times now retain the correct sequence as originally configured. |
26-Jun-25 |
|
|
Recommend: Query failures in Airflow for Data Science Workbench |
We have fixed issues causing certain Data Science Workbench queries to fail in Airflow due to incorrect syntax. The query format has been corrected to ensure compatibility. Additionally, updates have been made to improve error handling and accurately reflect strategy update statuses, including better messaging for decommissioned sites. |
26-Jun-25 |
|
|
Find: Brand name matching issue in autocomplete for accented characters |
We have fixed an issue where brand names containing accented or special characters were not returned in autocomplete unless typed with exact spelling. The improvement ensures more consistent brand suggestions, especially for inputs missing accent marks, aligning behavior across languages and reducing gaps between autocomplete and search results. |
26-Jun-25 |
|
|
Find: Fixed search handling for brand names with acute and circumflex accents |
We have resolved an issue where brand names containing acute or circumflex accents were not being matched unless users entered the exact spelling. Search and autocomplete now support both accented and unaccented variations, improving discoverability without relying on manual synonym configuration. |
26-Jun-25 |