Skip to content

Changelog: January 2025

API

Added

  • Added additional metadata to the response when a Search Index is being rebuilt. This can be used to gain insight into the recent changes that have been applied to the search index.

  • Improved multi-word predictions by further using statistical data to determine the best stemmed version of a word to use in a given context.

  • Retail Media: Introduced BidRelativeProbability promotion computation strategy. It calculates each campaign’s time window by comparing its bid size with other active campaigns, so each campaign’s time window is proportional to its bid.

  • Search Highlight is now released. This feature allows you to return an overview of why a specific search request returned the results that it did.

Changed

  • Relewise.Client NuGet Package now contains more information and a link on how to get started.

Fixed

  • Fixed a bug in our optimization layer that would not trigger updates when active/disabled state changed for the ClearAndReplace ProductUpdate Kind.
  • Fixed a bug in search results where variant matches were prioritized over product-only matches, even when the variant matches were less relevant.
  • Fixed a bug in the multi-word prediction handler that would cause an error during evaluation. We have added safeguards in the relevant areas to ensure that similar issues won’t reoccur.
  • Fixed an error that occurred when an EqualsCondition was used without a value.
  • Fixed an issue where categories that were automatically deleted after being disabled for a long period of time would not be removed from products and content, which left empty category paths on those entities.
  • Setting ExplodeVariants = 0 in the ProductSearchRequest settings no longer returns a VariantResult.

My Relewise

Added

  • Added client-side validation to search index languages to help users identify what might be wrong with the values in the configuration.
  • Added the option to only include approved or unapproved synonyms when exporting.
  • Added validation messages to the input fields, which display why a field might be invalid. This will help users identify what is required of the value they input into a given field.
  • Front Page: Added a link to our changelog.
  • Merchandising: Added support for the "Product has variants" filter.

Changed

  • New and improved input validations for synonyms now help users create rules that follow the logic of the API. Rules cannot contain special characters except / or - provided they are surrounded by numbers. E.g., 1/2 would be valid while a/b wouldn't be.

Fixed

  • Fixed a bug where having selected a filter option on a merchandising rule could cause the page to display no rules when navigating back to the Merchandising overview page.
  • Fixed a bug where the input validation refused to confirm passwords for MyRelewise users.
  • Fixed a bug where updating a search index would not update the raw JSON immediately, but require a refresh of the page.
  • Retail Media - Campaigns Performance: Added a missing "back-to-overview" button, to align with the details-tab.

Integration

Added

  • Java SDK: Added support for the search highlighting feature.
  • JS SDK: Added the option to include disabled variants in the ProductHasVariantsFilter.
  • JS SDK: We now include snippets in search highlight results.

Changed

  • Demo Shop: Reduced the number of popular category recommendation on the front page. This will make it more easy to demo product recommendations.
  • Demo Shop: We now keep the context switcher open when making changes to the context, to give a more smooth demo experience.

Fixed

  • Demo Shop: Fixed an issue where the "On Sale" recommendation header was being shown even when the recommendation was empty.
  • Demo Shop: The Retail Media products on the category page rendered to the right of the results would always be rendered outside the viewport and cause overflow.
  • Shopify: Fixed an issue where webhooks would sometimes send line items without any product ID on one or more line items.

Documentation

Added

  • Added a How-to guide for creating a merchandising rule that boosts or buries products based on their profit margin. Available here.
  • Added examples for Object and ObjectList data to the JSON product integration example page.
  • Added page for Recommendation Usage, offering quick and commercially relevant insights into the use of Relewise recommendations for common usecases. Available here.
  • FAQ: Added an entry detailing why prices might not show as updated on the frontend of a solution, or in an email recommendation. Available here.

Changed

  • Moved available information on the Fill function to the Recommendations Types page.

Don't know us? Don't worry - you can find more information about us, by visiting our main page www.relewise.com