2026 Product Recommendations release notes
5 min
Subscribe to updates on this page using the notification and RSS feed options.
October
- Added a Language filter to the Orders report, letting you view product recommendation performance for a specific site locale. See Product Recommendations reportsProduct Recommendations reports for more information.
- Added the Algorithm Performance report, letting you compare clicks, purchases, and revenue for each algorithm across widgets and pages. You can filter by time range, currency, and widget, and export the data to CSV. See Product Recommendations reportsProduct Recommendations reports for more information.
September
- Released the pr_delete_custom_page tool in Agent Platform to help you delete an existing custom page from your website. See Page and position management for more information.
- Added prefix-match locale matching to Triggered Messages campaigns, letting you select a de*-style option in the Locale field to match every regional variant of that locale (such as de, de-ch, and de-de) with a single trigger instead of one trigger per variant. Locale values without a trailing * keep their existing exact-match behavior. See Set up an Optimizely Triggered Messages campaignSet up an Optimizely Triggered Messages campaign for more information.
July
- Added feed notification management to the Configure tab in the Admin section, letting you set alerts for selected accounts whenever a feed import succeeds, fails, or both, so teams learn about import problems without manually checking the feed. Each account is assigned a Success, Failure, or Both notification type, and you can search or filter recipients by name, email, or notification type. See Administer the Product Recommendations portalAdminister the Product Recommendations portal for more information.
- Added an AND variant to the Bestsellers by customer segments and Viewed together by segment algorithms. The AND variant targets shoppers who belong to all selected segments, letting you build more precise segment-based recommendations. See Algorithms for widgetsAlgorithms for widgets for more information.
- Released additional System tools for Product RecommendationsSystem tools for Product Recommendations in Agent PlatformAgent Platform to retrieve algorithm performance reports for a date range to compare which recommendation algorithms drive the most revenue.
May
- Added pagination to the Search reportAdded pagination to the Search report, improving load times and readability for sites with large volumes of unmatched search terms.
- (Beta) Released Mark chat for Product Recommendations. Mark is an AI assistant that provides contextual guidance directly within the product. Take Mark's help to manage high-level workspaces, retrieve algorithms, define website layouts, manage recommendation blocks, and analyze the impacts of your product recommendations. See Mark chat for Product RecommendationsMark chat for Product Recommendations.
April
- Added Widget Breakdown to self-service A/B test resultsself-service A/B test results, giving you total revenue by variation and key metrics like product refCode, brand, impressions, clicks, items bought, and revenue for each widget within a variation.
- Added three new customer segment-based algorithms, Added to basket after viewing by segmentAdded to basket after viewing by segment, Viewed together by segmentViewed together by segment, and Purchased after viewing by segmentPurchased after viewing by segment, to help you target recommendations more precisely.
Agent Platform releases
Released the following Optimizely Product Recommendations toolsOptimizely Product Recommendations tools in Agent overviewAgent overview to help you monitor and analyze product, search, and category performance directly in Mark.
- pr_get_category_report – Retrieves a category report containing performance metrics based on a provided date range and site configuration.
- pr_get_search_report – Generates a search report detailing search performance and metrics for a specific site and timeframe.
- pr_get_trend_report – Retrieves a trend report analyzing performance by product, brand, or category over a designated period.
March
Agent Platform releases
Released the following Optimizely Product Recommendations toolsOptimizely Product Recommendations tools in Agent Platform to help you configure sites, manage recommendation widgets, and analyze performance data.
- pr_get_sites – Retrieves the available sites for your organization.
- pr_switch_site – Switches the active site context.
- pr_get_locations – Retrieves a list of locations or regions from a specified server and site.
- pr_get_currencies – Retrieves all available currencies for a site.
- pr_get_channels – Retrieves all available widget channels for a site.
- pr_get_algorithms – Retrieves a list of available algorithms for a specific project or environment.
- pr_get_algorithm_hints – Retrieves available quick filters (hints) for a specific algorithm.
- pr_get_attributes – Retrieves specific attributes for product recommendations based on a search term.
- pr_get_categories – Retrieves a list of categories based on a search term for a specific site and server.
- pr_get_product_sets – Retrieves all available product sets for a site.
- pr_get_pages – Retrieves all pages for a site.
- pr_create_page – Creates a new page within a site.
- pr_get_positions – Retrieves the available widget positions for a specific page.
- pr_create_position – Creates a new position on a specific page for placing widgets.
- pr_get_widgets – Retrieves all available widgets for a site.
- pr_get_widget_config – Retrieves the configuration details for a specific widget.
- pr_create_widget – Creates a new recommendation widget. Optimizely currently supports quick filters only. Advanced filters support will be added in a future release.
- pr_get_order_report – Retrieves order reports for a specified date range.
- pr_get_products_report – Retrieves a products report based on specified criteria such as date range, location, and sorting preferences.
- pr_get_conversion_report – Retrieves a conversion report for a specific site or location within a given date range, letting you filter by traffic, conversion type, and currency, with pagination and sorting options.
- pr_get_engagement_report – Retrieves an engagement report for a specific site within a given date range.
- pr_get_pagetype_report – Retrieves a performance report broken down by page type for Product Recommendations.
February
- Added the ability to create widgetsability to create widgets. Previously, you could only edit or duplicate existing widgets. Now, when creating widgets, you can place them on pages, select their positions, choose the number of recommendations, select your channels, and configure whether recommendations always appear on your pages.
- Self-service A/B testingSelf-service A/B testing is now generally available. Using this feature, you can independently create, configure, and launch experiments without relying on developer support.