Release Summary 26.08 | Apr 16, 2026
The following key features and improvements, along with bug fixes, have been released in Algonomy DXP products in the release version 26.08.
Configurable Strategies
New Configurable Strategies Page Now Available from Navigation
We are excited to announce that the new Configurable Strategies page is now officially available to all customers directly from the portal dashboard left navigation. Previously accessible only via direct URL as early access, the redesigned page is now fully integrated into the main navigation.
The new interface improves usability and organization. The “Choose A Model” section is structured into three sub-sections: users first select a model category, then choose from the corresponding models, and finally configure additional options such as region-based seed selection. Models are logically grouped by seed type, including sitewide, category, brand, product context, user history, attribute value, and other specialized strategies.
Set Strategy Message Directly in Configurable Strategies
Users can now set and manage the Strategy Message directly within the new Configurable Strategies page. A dedicated section below the Strategy Name displays the current message and translation count, with an option to edit via a sidebar.
The system checks existing messages across page types when loading a strategy and provides warnings if inconsistencies exist, along with an overwrite option. The Save and Publish action is independent of the main strategy save, and users are prompted to save or discard unsaved changes.
Jira: ENG-31524
Support for Using Segment as a Seed
Configurable Strategies now support segment-based models through a new “Use Segment Models” option. This allows users to apply segment-specific recommendation models with affinity sorting and category recommendations.
Users can select from whitelisted segments, reorder them via drag and drop to define priority, or apply all available segments. This option is available for selected models including Viewed Together, Viewed Then Bought, Bought Together by User, Top Sellers, Movers and Shakers, Top Views, and Top Products. The View Results tab reflects segment usage in the Algorithm Assembly and requires a user ID for preview.
Jira: ENG-30815
New Models: Product List and Categories Bought Together
Two new models have been added to the Configurable Strategies page under the “Other” model type group:
-
Product List Model: Enables users to define a fixed list of products within a strategy. Supports unlimited product IDs entered as a comma-separated list. Includes affinity-based sorting, brand and category diversification, category recommendation enablement, and the option to include the seed product.
-
Categories Bought Together (CategoryCP2) Model: Enables category recommendations based on category co-occurrence data. Supports Category Affinity, Category Context, and Fixed Category Seed as personalization seed options, along with User Affinity, Smart Shuffle, and Margin sorting. Users can specify the category tier and include the seed product in results.
Jira: ENG-32015, ENG-32016, ENG-32211
Preview with Multiple Product Seeds
The Configurable Strategies preview now supports selecting multiple product seeds when using the Product Context seed type. This enhancement allows users to evaluate how recommendations perform across more than one seed product within the same preview session.
Find
Enhanced Query Understanding Filter with Multi-Token Category Matching
The Query Understanding filter has been enhanced to support multi-token matching, allowing search queries to match multiple relevant categories instead of only the closest term. Previously, the filter returned results from a single matched category. For example, a search for “yellow kurta for men” would return results only from the “Kurta” category and exclude related categories such as “Kurta sets.”
With this update, the system identifies all matching tokens within the defined QU scope and returns results from all relevant categories. This ensures broader and more accurate category coverage in search results.
To support this feature, a new query tag property collection must be created using the new propertyType: querytag. Customers who want to enable this enhancement will need to create a new snapshot, which can be copied from an existing snapshot using the same property collection.
Jira: ENG-31878
Personalization Settings Integrated into Search Engine Configuration
Search personalization settings have been fully integrated into the Search Engine configuration, ensuring that personalization behavior is managed as part of the core search setup rather than through separate controls. This centralizes configuration, improves consistency, and simplifies ongoing management.
The update includes end-to-end support across all services. Personalization settings are now stored and persisted within search configurations in playservice, supported by updated client integration with backward compatibility. The dashboard reads and manages these settings directly within the Search Engine configuration screen, allowing users to enable or adjust personalization options alongside other search parameters. Updates have also been implemented in OC master and rrserver to publish, cache, and apply personalization settings as part of the standard search configuration payload, ensuring consistent execution across search requests.
Jira: ENG-31641
Engage
Multi-Value Attribute Support for Content Affinity Sorting
Content affinity sorting now supports multiple attribute values per content item, enabling more accurate and comprehensive personalization. Previously, when a content item contained multiple values for the same attribute, only one value was considered during affinity calculation. This limited how effectively content could align with user preferences.
With this enhancement, all attribute values associated with a content item are considered when calculating affinity scores. For example, a content asset featuring multiple designers can include multiple enhanced tags for the designer attribute, allowing the system to account for a shopper’s affinities across all associated designers.
This allows content tagged with multiple designers, brands, or other attributes to be ranked based on the user’s combined affinities across all relevant values. The update includes full support in the data model and UI, along with validation across Content Reports, Content Catalog pages, and layout qualifiers to ensure consistent behavior throughout the platform.
Jira: ENG-31603
Product List Option for Content+Recs in Engage Campaigns
Engage campaigns now support a Product List seed option for content+recs when using “Seed defined within selected content.” This allows digital marketers to dynamically define the exact products to be recommended alongside promotional content, ensuring tighter alignment between offer messaging and featured products.
When Product List is selected as the seed type, users are guided to create a Configurable Strategy using the Product List model with Product Context and Affinity Sorting for personalization. The Rule Summary clearly reflects the selected seed type. Campaigns configured with this option return the content together with the products defined in the reserved content attribute (algo_recs_manual_products), enabling seamless manual product control within personalized campaigns.
Jira: ENG-32012, ENG-31970, ENG-32280
Content Overview Reporting: Full Placement Name with Page Type
Content Overview Reporting and Campaign Overview Reporting have been updated to display the full placement name, including both page type and page area, in the format page_type.page_area. Previously, only the page area was shown.
A new Page Type column has also been added to the reports, consistent with other reports such as Site Analytics. This update applies to both rollup-level reporting and TS visualizations, ensuring clearer placement identification across reporting views.
Jira: ENG-32178
Enterprise Dashboard
Multiple Email Placement Support in Content and Recs Test Drive
Content Test Drive and Recs Test Drive now support multiple email placements, enabling users to preview de-duplication behavior across placements within a single email campaign. Previously limited to one placement, this enhancement aligns preview functionality with live email configurations.
When Email is selected as the channel, users can configure multiple placements, each with its own layout and slot settings. In Recs Test Drive, a “Show strategy message” option is available and enabled by default. All requests now include the dedupWith parameter, constructed from the selected placement and layout pairs, ensuring consistent de-duplication across slots and accurate preview behavior.
Additionally, product cards in both Recs Test Drive and Content Test Drive have been updated to match the design and layout used in the new Configurable Strategies preview.
Jira: ENG-32085, ENG-32014, ENG-31947
Placement Profile Page: Enhanced Strategy Configuration View
The Placement Profile page now includes a redesigned Strategies view, replacing the previous table with a consolidated, read-only configuration overview. Strategy Rules are grouped by context and displayed in a clear waterfall layout, showing execution order, precedence, and fallback behavior.
Each context group highlights its targeting conditions, associated rules by importance, preferred strategies, and strategy messages. Default and Conditional contexts are clearly distinguished, and Global Fallback strategies are shown when applicable. The page prioritizes loading strategy data first for improved performance and is designed to support inline editing in a future release.
Jira: ENG-31703
Recipes: Product Name Search with Auto-Complete
The Recipes page now supports searching by product name with auto-complete, replacing the previous requirement to enter a product ID. The experience aligns with the product seed selection workflow used in the Configurable Strategies View Results page, providing a more intuitive and efficient search experience.
A picker has also been added for the User ID field, allowing users to search and select from available values. Clear validation messages are displayed if required fields are missing, and a specific error message appears if no recipes are found for the specified product name and locale.
Jira: ENG-31945
Expanded Activity Log Coverage
The Activity Log has been enhanced to provide more complete auditing for Dynamic Experiences and User Affinity Configurations. Updates to experiences and variations are now logged with relevant change details. Any publish action, whether triggered directly or as a result of a save, is also recorded.
Creation, edits, and deletions of User Affinity Configurations are now logged. Improvements have also been made to better capture detailed change information in the audit log.
Jira: ENG-31653
Recommend
Customer-Facing API for Managing Strategy Messages
A new customer-facing API is now available to retrieve, set, and update Strategy Messages for Configurable Strategies. This enables translation and localization teams to manage multi-language messaging programmatically, streamlining workflows and reducing reliance on manual dashboard updates.
The API uses OAuth authentication via the API gateway and supports bulk updates across multiple strategies and page types. It validates inputs, returns clear status responses, and allows publishing messages to Production per strategy and page type for controlled rollout.
Jira: ENG-31299
Other Feature Enhancements
The following feature enhancements and upgrades have been made in the release version 26.08.
|
Jira # |
Module/Title |
Summary |
General Availability |
|---|---|---|---|
|
Enterprise Dashboard: Site Selection Dropdown: Site ID for Algonomy Employees |
The site selection dropdown now displays the Site ID alongside the site name for Algonomy employees, in the format Site Name (SiteID). Employees can search using partial site names, partial Site IDs, or any substring within the combined text. This update is visible only to Algonomy employees; non-employees continue to see site names without the Site ID. |
16-Apr-26 |
|
|
Enterprise Dashboard: Domain Alignment for Product and Click URLs in Chatbot Experiences |
Product navigation in Chatbot Dynamic Experiences has been updated to ensure that both productURL and the clickURL ct query parameter align with the current window domain. This resolves an issue where users accessing the chatbot on one domain could be redirected to a different domain, resulting in incorrect navigation and loss of chat session history. With this enhancement, recommended product links now dynamically match the active domain, preserving correct navigation and maintaining chat continuity. |
16-Apr-26 |
|
|
Chatbot: Voice Mode Support for Algo AI Sales Assistant |
The Algo AI Sales Assistant chatbot now supports a voice mode, allowing users to ask questions and hear responses without typing or reading text. Users can toggle voice mode on or off, with the system automatically detecting when the user has finished speaking and processing the input accordingly. |
16-Apr-26 |
|
|
Find: DC Solr Count Report: Summary Added |
A summary section has been added to the DC Solr count mismatch report to provide a clear overview of batch and non-batch discrepancies across data centers. The report now highlights critical and normal mismatches separately, along with total counts, and confirms email notification status. This improves visibility into collection-level differences and makes it easier to quickly assess overall impact across sites and DCs. |
16-Apr-26 |
|
|
Find: Related Searches: Case-Insensitive Query Matching |
Related Searches query matching has been updated to be case-insensitive, ensuring consistent behavior regardless of how the query is capitalized. Previously, suggestions were indexed in lower case and matched using a string field, resulting in case-sensitive behavior. With this enhancement, queries such as harry potter, Harry Potter, and HARRY POTTER now return equivalent relatedSearches and relatedSearchProducts results when data exists. The update preserves existing indexed data and applies consistently across data centers, with no regression for lower-case queries or other clients. |
16-Apr-26 |
|
|
Recommend: Guardrails for Precision Mode in Affinity Configuration |
Precision Mode in affinity configurations has been enhanced with built-in safeguards to prevent performance issues when applied to attributes with extremely high numbers of values. A maximum supported threshold for attribute values is now enforced. When Precision Mode is enabled, each attribute is evaluated against this threshold. If an attribute exceeds the limit, Precision Mode is automatically skipped for that attribute, and standard scoring is applied instead. |
16-Apr-26 |
|
|
Recommend: Configurable Strategies: Multi-Product Category Context |
Configurable Strategies using a Best Seller model with Category Context now correctly use categories from all seed products when seedStrategy: "ACCUMULATE" is applied. Previously, only one product’s category was used. If a product has a primary category, that category is used as the seed; otherwise, all associated categories are used. This ensures more relevant recommendations when multiple products are provided as seeds. |
16-Apr-26 |
|
|
Enterprise Dashboard: Runtime Affinity Computation Efficiency Improvements |
Runtime affinity computation has been optimized to improve performance and reduce unnecessary overhead. Updates address inefficiencies such as excessive object allocation and repeated operations, including redundant string processing within request handling. |
16-Apr-26 |
|
|
Data Engineering: Social Proof Coverage Analysis: Rollup Stability Improvements |
Social Proof Coverage Analysis: Rollup Stability Improvements The Social Proof Coverage Analysis rollup has been enhanced to ensure stability when the FrontEnd DC is out of rotation. The rollup now uses backend DC tables, preventing failures caused by FE DC outages. Performance optimizations have been applied, backward fill support has been added, and related data issues have been addressed to ensure timely and reliable rollup execution. |
16-Apr-26 |
|
|
Platform: Streaming Ingest: Validation for Empty Payload Items |
Validation logic has been added to detect and handle empty ingest payload items, defined as entries containing only an externalId with no additional data. Previously, such items were silently ignored by the streaming engine. With this enhancement, empty items are now marked as invalid and routed to invalidItems for proper tracking. They are visible through both the invalidItems API and the streaming-status API, improving traceability and debugging. |
16-Apr-26 |
|
|
Social Proof Reporting: All Categories and Category-Level View |
The Social Proof report has been enhanced to support both all categories combined and individual category-level performance views. This enables users to evaluate overall performance alongside category-specific insights. The report now includes two selectable options: Detailed Report (All Categories) and Category Level Report. By default, the detailed report across all categories is displayed. Users can switch to category-level reporting or select a specific category to view performance exclusively for that category. |
16-Apr-26 |
Bug and Support Fixes
The following issues have been fixed in the release version 26.08.
|
Jira # |
Module/Title |
Summary |
General Availability |
|---|---|---|---|
|
New DynEx UI: Fix for Change Template in Creation Flow |
An issue in the New Dynamic Experiences UI caused a 404 error when users selected a template, proceeded in the creation flow, then used the Change Template option and attempted to save. This issue has been resolved. Users can now change templates during the creation flow and save successfully without encountering errors. |
16-Apr-26 |
|
|
New DynEx UI: Fix for Adding Targeting Context Conditions |
An issue in the New Dynamic Experiences UI prevented users from adding conditions when creating a new Targeting Context. This issue has been resolved. Users can now successfully add conditions while creating or editing a Targeting Context. |
16-Apr-26 |
|
|
Channel Evaluation Based on User Agent for All Requests |
Channel detection has been extended to evaluate the User Agent across all requests, including /experiences calls. Previously, this logic was applied only to /targeting API calls. With the adoption of showOnlyQualified=true, where targeting calls are no longer made, channel evaluation now occurs server-side for all relevant requests. The request-type condition has been removed, ensuring consistent channel evaluation across APIs and enabling accurate experience targeting based on device context. |
16-Apr-26 |
|
|
Enterprise Dashboard: New Configurable Strategies Page: Additional Fixes |
Several usability fixes have been implemented on the new Configurable Strategies page. The Personalization Seed is now correctly highlighted for existing strategies using the DSW Strategy model. Default personalization seeds are also automatically selected for applicable models to prevent incomplete configurations. Product Context is set by default for Product-to-Product and Product Attribute-Based models, while Category Context is set for the Categories Bought Together model. |
16-Apr-26 |
|
|
Enterprise Dashboard: Fix for Exact Phrase Match Reset After Save & Publish |
An issue was identified in Search Test Drive where the Exact Phrase Match value was resetting to 100 after clicking Save & Publish, even when configured to a different value such as 0. This issue has been resolved. The Exact Phrase Match setting now correctly retains the configured value after Save & Publish, ensuring that search optimization changes persist as expected. |
16-Apr-26 |
|
|
Find: Fix for Query Understanding enableFilters Default Behavior |
An issue was identified where the Query Understanding configuration parameter enableFilters was not behaving as true by default, despite documentation indicating that it should. As a result, requests that omitted enableFilters did not apply filtering correctly, while explicitly setting "enableFilters": true worked as expected. This issue has been resolved. The enableFilters configuration now correctly defaults to true when omitted, ensuring consistent filtering behavior across requests and environments. |
16-Apr-26 |
|
|
Social Proof: Coverage Analysis Report Loading Issue Resolved |
An issue was identified where the Coverage Analysis Report was not loading for Stadium.se in the Production environment. This issue has been resolved, and the report now loads correctly as expected. |
16-Apr-26 |
|
|
Recoomend: Prevent Airflow Job Trigger on Strategy Deletion |
An issue was identified where deleting a strategy in the UI could still trigger its associated Airflow job, resulting in unnecessary processing for strategies that no longer exist. This has been resolved. The system now validates whether a strategy is active before triggering the related Airflow workflow, ensuring no jobs are executed for deleted strategies. |
16-Apr-26 |
|
|
Enterprise Dashboard: Dynamic Experiences Rendering Issue Due to client.js Error |
An issue was identified where Dynamic Experiences campaigns were not rendering for multiple clients, including Epoca Cosméticos, Renner, and Camicado. The Experiences API was returning empty responses, and the Targeting API was not being triggered. The root cause was a client.js error (TypeError: Cannot read properties of undefined (reading 'indexOf')) that prevented the Targeting API call from executing, blocking campaign rendering. This issue has been resolved, and Dynamic Experiences now render correctly across the affected sites. |
16-Apr-26 |
|
|
Data Engineering: MVT Merchandising Test Dates Not Updating |
An issue was identified where updates made to Merchandising test rules in MVT, such as changes to start and end dates, were not reflected after saving. The previously configured dates continued to display even after the rule was updated. This issue has been resolved, and date changes in MVT test rules now save and display correctly. |
16-Apr-26 |