Skip to main content

Release Summary - Aug 24, 2023

The following key features and improvements, along with bug fixes, have been released in Algonomy products in the release Version 23.19 during Aug 11 - Aug 24, 2023.

Enterprise Dashboard

Enhance Social Proof Experience Campaigns with Importance Indicator

In this release, we have introduced a new enhancement to Social Proof, enabling users to assign Importance levels to their experiences. This empowers Digital Experience Optimizers to ensure that the most vital social proof messages are displayed when conflicts arise between multiple experiences.

Now, when configuring a Social Proof experience, you can specify its Importance level. The system will automatically prioritize experiences based on their Importance values, allowing for precise control over message display. This enhancement provides users with a powerful tool to optimize customer engagement strategies and ensure maximum impact.

Aug 24 2023_1.png

Jira: ENG-26565

Display Interval Values for Each Message Type in Social Proof UI

In this release, we have enhanced Social Proof by allowing Digital Optimization Managers to conveniently view and adjust interval settings within the user interface for each message type, eliminating the need for JavaScript code modification.

Within the Social Proof campaign configuration, a dedicated section for interval settings is available for each message type. Here, you can set numerical values and units (Minutes, Hours, Days, or Weeks) for intervals, with default settings displayed as a reference.

This streamlines interval management, enabling you to optimize campaigns effortlessly. Please note that interval values are optional for specific message types, simplifying your campaign setup.

Aug 24 2023_2 - Copy.png

Jira: ENG-26577

Enhanced Social Proof by Displaying Dates Alongside Campaign Names

We have enhanced our Social Proof feature to provide clearer campaign management. You can now view campaign dates alongside the campaign name, both in view and edit modes. This update streamlines the editing process, making it easier for you to manage campaign date ranges with confidence.

Jira: ENG-26510

Composite Outfit UI - Enable to Select Category/Sub Category from Category Hierarchy

We have made it easier for merchandisers to select categories and subcategories within our Composite Outfit feature. With the addition of a category hierarchy view, you can now effortlessly choose categories for defining styles and include/exclude subcategories without needing to remember category IDs or names. This update streamlines the category selection process, enhancing your overall experience when creating outfits.

Aug 24 2023_3.png

Jira: ENG-25811

Resolve CXP Portal Issue with UPS Category View

We have resolved an issue in CXP Portal UPS, where some users encountered excessive category views being displayed inaccurately.

Upon thorough investigation, we identified that all category views were mistakenly added to all sessionID-date sections, leading to an inaccurate representation of user activities. To rectify this, we have refined the categorization of category views, ensuring they are now accurately displayed only within the relevant sessionID-date sections.

Jira: ENG-26436

Add a diversify by attribute filter to Only Recommend rules

We've introduced a new feature that provides digital merchandisers with even greater control over recommendation diversity. Now, within the "Only Recommend" rules, merchandisers can apply diversity filters based on specific product attributes. This means you have the flexibility to precisely tailor diversity settings to meet your unique requirements.

For example, imagine you want to showcase product recommendations while ensuring that products with the same "product type" attribute are not overly repeated. With this feature, you can easily achieve this goal. You can specify the attribute name, set the maximum number of products per attribute value (e.g., limit each color to 3 products), and fine-tune your recommendations to provide more personalized and engaging shopping experiences.

This enhancement isn't limited to a single strategy type; it's adaptable across various strategy types, making it versatile and suitable for a wide range of merchandising needs.

Jira: ENG-26179

Enterprise Dashboard & Recommend

Enhance Configurable Strategies Best Offers Model with Personalization Seed Options

In this release, we have improved the Best Offers model within Configurable Strategies. Now, you can easily customize your top offer recommendations based on specific categories or brands.

Within the 'Strategy Composition,' you'll find Personalization Seed options that match those for Top Sellers, including User's History-based preferences like Category and Brand Affinity, as well as Context-based options like Category and Brand Context.

In the 'View Results' tab, input options align with your selected seed for straightforward configuration, supported by auto-complete functionality.

These enhancements offer you more flexibility and precision when configuring your Best Offers strategy.

Jira: ENG-26100 & ENG-25731 


Ability to Set TTL for Email Recommendation Cache (Early Access)

We have introduced the ability for merchandisers to define the time-to-live (TTL) for email recommendation caches at the site level. This enhancement allows merchandisers to have precise control over how long recommendation caches are maintained, offering a more tailored and dynamic shopping experience for your customers.

With this feature, you can configure the cache refresh rate to align with your specific needs. For example, you can set a cache TTL of 12 hours for one site, ensuring that recommendations are refreshed twice daily, while another site might benefit from a 6-hour TTL, delivering entirely different recommendations to shoppers who open their emails within that timeframe.

This level of customization empowers you to optimize your email marketing strategies by providing fresh, relevant product recommendations to your customers. Whether it's promoting new arrivals, showcasing trending products, or highlighting personalized suggestions, this feature allows you to fine-tune your email recommendation caching for maximum impact.

Aug 24 2023.png

Jira: PLAT-3473

Feed Processing & Streaming Catalog

Catalog Enrichment data import for feed clients

In this release, we have introduced a streamlined integration for Catalog Enrichment data, enhancing the capabilities of our feed clients. This improvement enables seamless synchronization of approved Catalog Enrichment attributes with product IDs in the Primary catalog, ensuring your product recommendations are both accurate and effective.

Here's how it works: During the feed herder process, the system automatically collects accepted Catalog Enrichment attributes and synchronizes them with corresponding product IDs in the Primary catalog. This synchronization occurs each time the feed herder process runs, guaranteeing that your product catalog is consistently updated with the latest enrichment data.

To maintain data relevance, the process refreshes Catalog Enrichment attributes every 24 hours. This advancement not only simplifies your workflow but also empowers you to utilize these attributes for attribute top views, attribute top sellers, user affinity, rules, and guided selling filters.

Jira: PLAT-3527

Other Feature Enhancements

The following feature enhancements and upgrades have been made in the release Version 23.19 during Aug 11 - Aug 24, 2023.

Jira #



General Availability



Enterprise Dashboard:

Customizable Lookback Period for NLP Sort Options

We have enhanced our NLP sorting feature to give you more control over your search results. You can now configure the lookback period for user history, allowing you to fine-tune sorting based on historical data. In the configurable strategy settings, you'll find the "NLP Look Back Period" option, which lets you set the number of days for the lookback period. The default is 720 days, but you can adjust it from 1 to 720 days to suit your needs.





Enhanced Related Searches in Response Sections

In this release, we have introduced the ability for clients to specify related searches in the response sections, enhancing the functionality of the Context Solr Plugin. Now, clients can activate related searches logic by including relatedSearches in their responseSections. This allows for a more comprehensive exploration of related search terms and their associated products.





Surface ClusterPools for the sites

We have introduced the ability to reverse query sites with cluster pools. Now, when using the ‘getClusterInfo’ API with a specific site ID, you can obtain a list of cluster associations for which that site is eligible.





Integrate Datadog Metrics in Cluster service

We have integrated Datadog metrics into the Cluster service for enhanced monitoring. New metrics encompass startup events, cache loads, cache load failures, total connection attempts, connection failures, cache hits, successful data updates in Consul, and Consul data reads. These metrics empower us to better track and respond to service issues, ensuring smoother operations.





Pass suggestionSections

We have introduced improved SuggestionSections functionality. Now, you can leverage relatedSearches and relatedSearchProducts in addition to assortmentSection. These sections are seamlessly integrated into both the Solr plugin and the search service.





Composite Outfit - Use color predictions to identify a complementary color score

We have integrated color predictions to enhance outfit generation. When behavioral data is limited, color becomes a vital factor in crafting appealing outfits. We identify complementary colors for each product, assigning scores accordingly. This enhances the variety and visual appeal of outfit suggestions.




Data Engineering:

Update Session Report Visualizations to Use Worksheets Instead of Tables

We have improved the Session Report's liveboards for the top 1000 users by optimizing the way visualizations are created. Previously, visualizations were directly generated from tables, which occasionally led to filter-related challenges in the user interface. Now, we have enhanced the system to retrieve data directly from the underlying worksheets instead of tables. This change ensures smoother navigation and more efficient data visualization within the Session Report.




Data Engineering:

Load the dimension tables into HDFS for Custom Reporting Automation

To enhance efficiency, we have introduced a process to load dimension tables into HDFS, eliminating the need for repeated table joins with Postgres or Thoughtspot. Initially, these tables were loaded into Thoughtspot once a day, but now we have extended this process to include HDFS as well. With the ongoing TS Migration, we're transitioning the dimension tables to be loaded into both S3 and HDFS in the future.




Platform - Feed Processing:

Ability to control the update of catalog enrichment attributes with Primary catalog

We've introduced a feature to enhance catalog enrichment, allowing optimization managers to have better control over product attributes. Now, if an attribute does not exist for a product ID, it's automatically added along with the value. If the attribute exists but the value doesn't, the system replaces the value.

This ensures catalogs are always complete and up to date, saving time and effort.





Pass suggestionSections

Added suggestionSections:relatedSearches, relatedSearchProducts similar to assortmentSection.


Integrated Solr plugin suggestionSections in search service.

Integrated service suggestionSections in rrserver.





Update context Solr plugin to perform related searches

We have updated the context-solr-plugin jar and solrconfig to the language containers.





Ability to accept and process multiple placement ID in rrserver/Mailview API

Recommend can now accept multiple placement IDs via the Mail view API from Mail service and dedupe the product recommendations across the placements which are passed in Mail view API request.


Customized the Mailview API so that RRserver can accept multiple placement IDs along with the layout ID in a single API request so that de-duplication of the products being recommended can happen during the runtime processing.

This also required modification to how rrmail processes the requests from runtime during email open, rather than making one rrserver call against each placement ID in an email layout, Mail service shall make a single call with all placementID along with the layout ID to get the product recommendations with no duplicates.





Test the bearer token TTL with UPS and Omni channel APIs

Validated the working on the bearer token TTL that has been configured in Kong for UPS and Omni channel APIs. The TTL has been configured for 8 hours. Thus, after 8 hours, client system will have to request for a new bearer token.



Bug and Support Fixes

The following issues have been fixed in the release Version 23.19 during Aug 11 - Aug 24, 2023.





General Availability



Enterprise Dashboard:

Error while saving Affinity Configuration

We've fixed an issue where users encountered an error while trying to save Affinity Configuration changes for two clients. Previously, the changes didn't reflect as expected, even though the system indicated a successful update. Now, users can reliably update and save their Affinity Configuration values without encountering errors.





Diversified by brand and category filters are not working as expected when use multiple items from history is checked

We have addressed a bug related to diversified brand and category filters when using multiple items from a user's history. Previously, products from the same brand were appearing together, contrary to the filter settings. This issue has now been fixed, ensuring that diversified results by brand or category function as expected when the 'multiple items from history' option is checked.




Disable response cookies siteconfig is not working in p13n_generated.js call

We have fixed the issue with the 'Disable response cookies' site configuration not working in p13n_generated.js calls. Previously, it only applied to click calls, but now it functions correctly for p13n_generated.js calls as well. This ensures consistent control over response cookies across various call types.


  • Was this article helpful?