System tools for Product Recommendations

  • Updated

System tools are built-in features that help Optimizely Opal take action. Each tool performs a specific task, such as creating a campaign, uploading files, or generating images. Think of tools like attachments on a Swiss Army knife. Each one has a distinct purpose that helps you get work done.

In addition to the system tools available in Opal, Optimizely Product Recommendations includes a set of system tools designed to help you improve your recommendations.

You can ask Opal what tools it has at any time. For example, enter "Please list the tools you have with a brief description of what they do and the parameters" into Opal Chat.

Click a tool's name to expand it and learn when to use it, its required and optional parameters, and example prompts to calling the tool. If you do not provide a required parameter, Opal prompts you for it.

System tools for Product Recommendations are in beta. Contact your Customer Success Manager for information.

Site and environment management

These tools help you set up and manage your high-level workspace, including selecting the active site, regional settings, and foundational configurations.

pr_get_channels – Retrieves all available widget channels for a site.
  • When to use
    • To see the list of channels that widgets can be assigned to.
    • When creating a widget and you want to specify a channel other than the default.
    • To understand how your site's channels are configured for recommendations.
  • Parameters
    • serverUrl – The server URL for the site.
    • site – The name of the site.
  • Example prompts
    • List all available widget channels for my site.
    • What channels can I use for a new widget?
    • Show me the channels.
pr_get_currencies – Retrieves all available currencies for a site.
  • When to use
    • To find out the default currency for the site's reports.
    • To see a list of all supported currencies for reporting.
    • When you need to specify a currency for an order report.
  • Parameters
    • serverUrl – The server URL for the site.
    • site – The name of the site.
  • Example prompts
    • What currencies are available for my site?
    • Show me the list of supported currencies.
    • What is the default currency?
pr_get_locations – Retrieves a list of locations or regions from a specified server and site.
  • When to use
    • Retrieve available geographic locations or regions configured for a specific site.
    • Fetch a list of store locations or delivery areas from the server.
    • Get location data to set up location-based personalization or targeting.
  • Parameters
    • serverUrl – The URL of the server to query for locations.
    • site – The specific site identifier or name to fetch locations for.
  • Example prompts
    • Get the locations for the site "main-store" from the server "https://api.optimizely.com".
    • Can you fetch the available regions for site "eu-region" using server URL "https://eu.api.optimizely.com"?
    • I need to see all the locations configured on "https://test.optimizely.com" for the "beta-site".
pr_get_sites – Retrieves the available sites for your organization.
  • When to use
    • To see a list of all sites you have access to.
    • To select a site to work with before performing other actions like creating pages or widgets.
    • At the beginning of a session to set the context to a specific site.
  • Parameters
    • None
  • Example prompts
    • List all my sites.
    • Show me the available sites.
    • Which sites can I access?
pr_switch_site – Switches the active site context.
  • When to use
    • To change the active site you are working on.
    • When you need to perform actions on a different site than the one currently selected.
    • To ensure subsequent commands are executed against the correct site.
  • Parameters
    • None
  • Example prompts
    • Switch to the 'My Test Site' site.
    • I want to work on the 'Staging Environment' site now.
    • Change the site to 'Production'.

Algorithms and catalog data

These tools help you set up and manage your high-level workspace, including selecting the active site, regional settings, and foundational configurations.

pr_get_algorithm_hints – Retrieves available quick filters (hints) for a specific algorithm.
  • When to use
    • To see what filtering options are available for a particular recommendation algorithm.
    • To understand how you can refine the results of an algorithm. For example, by category or brand.
    • Before creating or updating a widget to know which hints can be applied to its algorithms.
  • Parameters
    • id – The ID of the algorithm.
    • serverUrl – The server URL for the site.
    • site – The name of the site.
  • Example prompts
    • What hints are available for algorithm ID 7?
    • Show me the quick filters for the 'Bestsellers' algorithm.
    • Can I filter the 'Recently Viewed' algorithm? What are the options?
pr_get_algorithms – Retrieves a list of available algorithms for a specific project or environment.
  • When to use
    • View the available recommendation algorithms to configure a new Product Recommendations campaign.
    • Check which algorithms are currently supported or enabled for a specific project.
    • Retrieve algorithm IDs to use in subsequent API calls or tool actions for setting up recommendations.
  • Parameters
    • serverUrl – The URL of the server.
    • site – The name of the site. 
    • (Optional) pageType – Page type to filter algorithms. If you provide, only algorithms compatible with this page type are returned.
  • Example prompts
    • Get the product recommendation algorithms for the 'home' page type on the 'eu-store' site using server 'https://api.recs.optimizely.com'.
    • What algorithms are configured for the 'product' page on site 'us-main'? The server URL is 'https://recs.example.com'.
    • Can you fetch the algorithms for the 'category' pageType for the 'global' site?
pr_get_attributes – Retrieves specific attributes for product recommendations based on a search term.
  • When to use
    • Fetch specific product attributes (like category, brand, or price) from the Product Recommendations server.
    • Search for available attribute values matching a specific term within a given site.
    • Retrieve configuration or catalog data needed to configure product recommendation campaigns.
  • Parameters
    • serverUrl – The URL of the Product Recommendations server.
    • site – The specific site identifier or name to query the attributes from.
    • (Optional) attr – The specific attribute to retrieve or search against (for instance, brand or category).
    • (Optional) term – The search term or keyword used to filter the attributes.
  • Example prompts
    • Get the "brand" attributes matching the term "Nike" from the server at https://pr.example.com for the "US-Main" site.
    • Can you use pr_get_attributes to find the "category" attribute for "shoes" on site "EU-Store" (serverUrl: https://pr-eu.example.com)?
    • I need to retrieve the "color" attributes using the term "blue" from the product recommendations server (https://api.pr.optimizely.com) for site "Global".
pr_get_categories – Retrieves a list of categories based on a search term for a specific site and server.
  • When to use
    • Fetch available product or content categories matching a specific keyword.
    • Retrieve category data from a specific server URL and site configuration.
    • Look up category identifiers or names to use in subsequent product recommendation API calls.
  • Parameters
    • serverUrl – The base URL of the server to query for categories.
    • site – The specific site identifier or name to retrieve categories from.
    • (Optional) term – The search term or keyword used to filter the categories.
  • Example prompts
    • Get the categories matching the term "shoes" from the site "retail-us" on server "https://api.example.com".
    • Can you find the "electronics" categories for the "global" site using the server URL "https://recommendations.api.com"?
    • I need to retrieve all categories related to "accessories" from the "eu-store" site at "https://api.optimizely.com".
pr_get_product_sets – Retrieves all available product sets for a site.
  • When to use
    • To see the list of product sets that can be used as fallbacks for widgets.
    • When creating a widget and you want to specify a fallback product set.
    • To manage and review the product sets configured for your site.
  • Parameters
    • serverUrl – The server URL for the site.
    • site – The name of the site.
  • Example prompts
    • List all product sets for my site.
    • What product sets can I use as a fallback?
    • Show me the available product sets.

Page and position management

These tools are used to define the layout and scaffolding of your website to prepare it for recommendation widgets.

pr_create_page – Creates a new page within a site.
  • When to use
    • When you need to add a new page to your site for placing recommendation widgets.
    • To create a custom page that is not one of the standard types (like Homepage, Product page, and so on).
    • Before creating a widget if the desired page for it does not exist yet.
  • Parameters
    • name – The name for the new page.
    • serverUrl – The server URL for the site.
    • site – The name of the site where the page is created.
  • Example prompts
    • Create a new page named 'Special Promotions'.
    • I need a new page for our 'Summer Collection' campaign.
    • Add a page called 'Brand Collaborations' to my site.
pr_create_position – Creates a new position on a specific page for placing widgets.
  • When to use
    • To define a new area on a page where a recommendation widget can be displayed.
    • When the existing positions on a page are already in use or not suitable for a new widget.
    • To organize and name different widget locations on a page for clarity.
  • Parameters
    • page – The ID of the page where the position will be created.
    • isCustomPage – A boolean indicating if the page is a custom page.
    • name – The name for the new position.
    • serverUrl – The server URL for the site.
    • site – The name of the site.
  • Example prompts
    • Create a new position named 'Header Banner' on page ID 123.
    • On the 'Homepage', which is not a custom page, add a position called 'Footer Recommendations'.
    • I need a new position for a widget on my custom page with ID 456. Call it 'Sidebar Ad'.
pr_get_pages – Retrieves all pages for a site.
  • When to use
    • To get a list of all pages where you can place recommendation widgets.
    • To find the ID of a specific page before creating a position or a widget.
    • To see all standard and custom pages available on your site.
  • Parameters
    • serverUrl – The server URL for the site.
    • site – The name of the site.
  • Example prompts
    • List all pages on my site.
    • What pages are available for widgets?
    • Show me the pages.
pr_get_positions – Retrieves the available widget positions for a specific page.
  • When to use
    • To see where you can place a widget on a particular page.
    • To find the ID of a position before creating a widget.
    • To check the available widget slots on a page like the 'Homepage' or a 'Product Detail Page'.
  • Parameters
    • page – The ID of the page.
    • isCustomPage – A boolean indicating if the page is a custom page.
    • serverUrl – The server URL for the site.
    • site – The name of the site.
  • Example prompts
    • What positions are available on the 'Homepage'?
    • Show me the positions for page ID 123.
    • List the widget positions on my custom page 'Special Promotions'.

Widget configuration and management

These tools are focused directly on creating, reviewing, and managing the actual recommendation blocks (widgets) shown to the end user.

pr_create_widget – Creates a new recommendation widget. Optimizely currently supports quick filters only. Advanced filters support will be added in a future release.
  • When to use
    • To add a new set of product recommendations to a page.
    • To configure a widget with specific algorithms, a certain number of recommendations, and a fallback product set.
    • To place a recommendation block in a predefined position on a page.
  • Parameters
    • name – The name of the widget.
    • page – The ID of the page where the widget displays.
    • position – The ID of the position on the page for the widget.
    • isCustomPage – A boolean indicating if the page is a custom page.
    • serverUrl – The server URL for the site.
    • site – The name of the site.
    • (Optional) numOfRecs  – The number of recommendations to display.
    • (Optional) alias – An alternate name for the widget.
    • (Optional) channel – The channel ID for the widget.
    • (Optional) fallbackProdSet – The ID of a fallback product set.
    • (Optional) recsExpected – A boolean indicating if recommendations are expected.
    • (Optional) widgetAlgorithms – A JSON string of algorithms to apply.
  • Example prompts
    • Create a new widget named 'Homepage Bestsellers' on the 'Homepage' in the 'Main Content' position.
    • I want to create a widget on the 'Product Detail Page' in the 'Cross-sell' position. It should show five recommendations.
    • Add a new widget called 'Recently Viewed' to the 'Sidebar' on page ID 789. Use the 'Recently viewed' algorithm.
pr_get_widget_config – Retrieves the configuration details for a specific widget.
  • When to use
    • To see the detailed configuration of an existing widget, including its name, status, and associated page and position.
    • To check which algorithms are currently being used by a widget.
    • To review a widget's configuration before making changes.
  • Parameters
    • id – The ID of the widget.
    • serverUrl – The server URL for the site.
    • site – The name of the site.
  • Example prompts
    • Show me the configuration for widget ID 3059898.
    • What are the details of the 'Homepage Bestsellers' widget?
    • Get the config for widget 12345.
pr_get_widgets – Retrieves all available widgets for a site.
  • When to use
    • To get a list of all widgets currently configured on your site.
    • To find the ID of a specific widget you want to examine or modify.
    • To get an overview of all recommendation widgets running on the site.
  • Parameters
    • serverUrl – The server URL for the site.
    • site – The name of the site.
  • Example prompts
    • List all widgets on my site.
    • Show me all the recommendation widgets.
    • What widgets are currently active?

Reports and performance analytics

These tools help you analyze the success, engagement, and revenue impact of your product recommendations.

pr_get_category_report – Retrieves a category report containing performance metrics based on a provided date range and site configuration.
  • When to use
    • View performance data for specific product categories over a designated time period.
    • Analyze category metrics like revenue or conversions within a selected date range.
    • Gather category-level insights (such as comparing level 1 vs level 2 categories) for a given website.
  • Parameters
    • startDate – The start date for the report in DD-MM-YYYY format.
    • stopDate – The stop date for the report in DD-MM-YYYY format.
    • serverUrl – The URL of the server.
    • site – The name of the site.
    • (Optional) – currency – The currency for the report.
    • (Optional) – level – Category level: 1, 2, 3, 4, and so on. Default is 1.
    • (Optional) – locationId – The location ID.
    • (Optional) – page – The page number for pagination. Default is 1..
    • (Optional) – size – The number of items per page. Default is 10.
    • (Optional) – sortBy – Field to sort results by (such as, "+date", "-name").
  • Example prompts
    • Get the category report for the site 'mysite' from 01-11-2025 to 30-11-2025.
    • Show me the level 2 category performance for last month on server 'example.com'.
    • Can you pull a category report for 'storefront' sorted by descending revenue?
    • Generate a category report for February 2026 at level 1. Which top-level categories are generating the most revenue through recommendations? I need this for our merchandising team's quarterly planning.
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.
  • When to use
    • Generate a report to analyze conversion rates and traffic for a specific website over a defined date range.
    • Retrieve paginated conversion data for a particular physical location or store to evaluate performance.
    • Pull a sorted list of conversions filtered by a specific currency to analyze revenue metrics.
    • Export traffic and conversion metrics for a specific server environment to monitor campaign success.
  • Parameters
    • startDate – The start date for the report's date range in DD-MM-YYYY format. Example: 20-11-2025.
    • stopDate – The end date for the report's date range in DD-MM-YYYY format. For example: 20-12-2026.
    • site – The specific site or website identifier to pull the report for.
    • traffic – Filter by traffic value: high or low.
    • conversion – Filter by specific conversion value: high or low.
    • serverUrl – The server URL to query for the report data.
    • (Optional) locationId – The identifier for a specific physical or logical location.
    • (Optional) currency – The currency code to filter or display revenue data. You can get the list of supported currencies using the pr_get_currencies tool. 
    • (Optional) page – The page number for pagination. Default is 1.
    • (Optional) size – The number of items per page. Default is 10.
    • (Optional) sortBy – The field and direction to sort the report results by.
  • Example prompts
    • Get the conversion report for site optimizely-main from 2025-01-01 to 2025-01-31, sorted by conversions.
    • Pull the conversion data for location loc-123 in USD currency, showing page 1 with a size of 50 records.
    • I need a conversion report showing organic traffic conversions for the last week on server https://api.example.com.
    • Can you fetch the conversion report for the UK site, starting from March 1st to March 8th, and sort it by traffic volume?
    • Show me products with high traffic but poor conversion for the last 60 days. I want to create a targeted promotion campaign for products that are getting eyeballs but not converting.
pr_get_engagement_report – Retrieves an engagement report for a specific site within a given date range.
  • When to use
    • Retrieve user engagement metrics for a specific site over a defined time period.
    • Analyze engagement data filtered by specific locations or currencies.
    • Fetch paginated engagement reports for dashboard displays, data exports, or performance reviews.
  • Parameters
    • site – The name of the site to retrieve the engagement report for.
    • startDate – The beginning date for the report data in DD-MM-YYYY format. For example: 20-11-2025.
    • stopDate – The end date for the report data in DD-MM-YYYY format. For example: 20-12-2025.
    • serverUrl – The server URL. 
    • (Optional) currency – The currency code for financial engagement metrics. You can get the list of supported currencies using the pr_get_currencies tool.
    • (Optional) locationId – The specific location identifier to filter the report. You can get the list of supported locations using the pr_get_locations tool.
    • (Optional) page – The page number for paginated results.
    • (Optional) size – The number of records to return per page.
    • (Optional) sortBy – The field to sort the report results by.
  • Example prompts
    • Get the engagement report for site 'US-Main' from 2025-01-01 to 2025-01-31.
    • Fetch the first page of the engagement report for location 'NYC', showing 50 results per page sorted by date.
    • Can you pull the engagement metrics for the EU site in EUR currency for last week?
    • Show me the engagement report for the past 30 days. I need to understand our CTR and click-to-purchase conversion rates to evaluate if our mobile recommendation experience needs improvement.
pr_get_order_report – Retrieves order reports for a specified date range.
  • When to use
    • To analyze the performance of product recommendations over a period.
    • To see how much revenue is being generated by Optimizely Product Recommendations.
    • To compare total orders, revenue, and products sold with those influenced by recommendations.
  • Parameters
    • serverUrl – The server URL for the site.
    • site – The name of the site.
    • startDate – The start date for the report (DD-MM-YYYY).
    • stopDate – The end date for the report (DD-MM-YYYY).
    • currency – The currency for the report.
    • (Optional) page – The page number for pagination.
    • (Optional) size – The number of items per page.
    • (Optional) sortBy – The field to sort the results by.
  • Example prompts
    • Show me the order report for last month.
    • Generate an order report from January 1st to today, sorted by revenue.
    • I need the order report for the past seven days in USD.
    • Show me the order report for Q1 2026 (January 1 to March 31). I need to present to the board how our recommendation engine impacted overall revenue and AOV compared to non-recommendation purchases.
pr_get_pagetype_report – Retrieves a performance report broken down by page type for Product Recommendations.
  • When to use
    • Analyze the performance of Product Recommendations across different page types (for example, home, product, and cart).
    • Retrieve revenue and engagement metrics for specific page types within a given date range.
    • Compare recommendation performance across different sites or locations.
  • Parameters
    • site – The name of the site.
    • startDate – The beginning date for the report's time period in DD-MM-YYYY format. For example: 20-11-2025
    • stopDate – The ending date for the report's time period in DD-MM-YYYY format. For example: 20-12-2025
    • serverUrl – The server URL or environment endpoint.
    • (Optional) currency – The currency code for revenue metrics. You can get the list of supported currencies using the pr_get_currencies tool. 
    • (Optional) locationId – The identifier for a specific location or region.
    • (Optional) page – The page number for paginated results.
    • (Optional)size – The number of results to return per page.
    • (Optional)sortBy – The metric or field to sort the report by (for example, revenue or clicks).
  • Example prompts
    • Get the page type report for site "my-store" from January 1st to January 31st, sorted by revenue.
    • Show me the product recommendations performance by page type for location 123 in USD.
    • Retrieve the first page of the page type report for the EU site, with a size of 50 results per page.
    • Generate a page type report for last month. I want to see how our product detail page recommendations are performing versus cart and checkout pages. Are we leaving money on the table?
pr_get_products_report – Retrieves a products report based on specified criteria such as date range, location, and sorting preferences.
  • When to use
    • Generate a report of product performance over a specific date range.
    • Retrieve a paginated list of products for a specific site or location.
    • Analyze product data sorted by specific metrics (e.g., revenue, views) in a given currency.
  • Parameters
    • site – The name of the site.
    • startDate – The beginning date for the report's time period in DD-MM-YYYY format. For example: 20-11-2025
    • stopDate – The ending date for the report's time period in DD-MM-YYYY format. For example: 20-12-2025
    • serverUrl – The server URL or environment endpoint.
    • (Optional) currency – The currency code for revenue metrics. You can get the list of supported currencies using the pr_get_currencies tool. 
    • (Optional) locationId – The identifier for a specific location or region.
    • (Optional) page – The page number for paginated results.
    • (Optional)size – The number of results to return per page.
    • (Optional)sortBy – The metric or field to sort the report by (for example, revenue or clicks).
  • Example prompts
    • Get a products report for site mysite from 2023-01-01 to 2023-01-31 sorted by revenue.
    • Retrieve page 2 of the products report for location loc-123 with a page size of 50.
    • Can you pull the products report in USD currency for the last week?
    • Show me the products report for the past 7 days sorted by revenue. Which recommended products are our top earners this week? I need to brief the inventory team on restocking priorities.
pr_get_search_report – Generates a search report detailing search performance and metrics for a specific site and timeframe.
  • When to use
    • Analyze the most popular search terms used by visitors on a site.
    • Review search conversion metrics within a specific date range.
    • Identify trends in user search behavior on the configured server.
  • Parameters
    • startDate – The start date for the report in DD-MM-YYYY format.
    • stopDate – The stop date for the report in DD-MM-YYYY format.
    • serverUrl – The URL of the server.
    • site – The name of the site.
    • (Optional) – currency – The currency for the report.
    • (Optional) – locationId – The location ID.
    • (Optional) – page – The page number for pagination. Default is 1.
    • (Optional) – size – The number of items per page. Default is 10.
    • (Optional) – sortBy – Field to sort results by.
  • Example prompts
    • Generate a search report for 'mysite' between 10-10-2025 and 20-10-2025.
    • Show me the top search terms sorted by volume for the main site.
    • I need a search performance report for 'eu-store' including location data.
    • Pull the search report for March 2026. What are the top search terms our customers are using? I want to align our Google Ads keywords and product page optimization with actual user behavior.
pr_get_trend_report – Retrieves a trend report analyzing performance by product, brand, or category over a designated period.
  • When to use
    • Compare the performance of different products, brands, or categories over time.
    • Identify the best or worst performing attributes based on revenue, page views, or conversions.
    • View unique conversion trends applying specific filters to hone in on granular data.
  • Parameters
    • startDate – The start date for the report in DD-MM-YYYY format.
    • stopDate – The stop date for the report in DD-MM-YYYY format.
    • serverUrl – The URL of the server.
    • site – The name of the site.
    • attribute – The attribute to analyze trends by. For example, "product", "brand", or "category".
    • sortOrder – "desc" for best performing, "asc" for least performing.
    • currentViewBy – The metric to view by. Such as, "conversion", "pageView", or "revenue".
    • (Optional) – currency – The currency for the report.
    • (Optional) – locationId – The location ID.
    • (Optional) – page – The page number for pagination. Default is 1.
    • (Optional) – size – The number of items per page. Default is 10.
    • (Optional) – sortBy – Field to sort results by.
    • (Optional) – selectedFilters – Array of filter values to apply to the report.
    • (Optional) – unique – Whether to return unique views or purchases only. Default is false.
  • Example prompts
    • Get a trend report for the best performing products by revenue for 'mysite' in November 2025.
    • Show me the worst performing brands by page views on 'global-store'.
    • Can you pull a category trend report showing unique conversions from 01-12-2025 to 31-12-2025?
    • What are the best performing products by revenue for the past 14 days? I need to feature trending items in our Instagram stories and email newsletter this week.
    • What brands had the best conversion rates through recommendations during the holiday season (December 1-31, 2025)? I'm planning our 2026 holiday featured brand strategy.

Example prompts

The following prompts help you get meaningful insights from your Product Recommendations data without needing to know which specific reports to run. Simply describe what you need in plain business language in Opal Chat and Opal automatically identifies the right reports, apply the correct filters, and surface the metrics that matter most to your use case. 

Q1 performance review
  • Prompt – How did our recommendations impact sales in Q1 2026? I need to show the board the revenue lift and whether customers who clicked on recommendations bought more than others.
  • What it does – Automatically generates order report showing revenue from recommendations and totals, order penetration, AOV with and without recommendations, and daily trends.
  • Business context – Executive reporting, board presentations, ROI analysis.
Homepage optimization analysis
  • Prompt – Which recommendation sections on our website are getting the most clicks and driving sales over the past month? I want to know what's working and what's not.
  • What it does – Pulls widget report with click-through rate (CTR), conversion, and revenue by placement or section.
  • Business context – UX optimization, homepage redesign, A/B testing validation.
Category merchandising strategy
  • Prompt – Which product categories are making us the most money through recommendations in February? I need this for our quarterly merchandising meeting.
  • What it does – Generates category report showing top revenue-generating categories with drill-down by widget.
  • Business context – Merchandising planning, category management, and quarterly reviews.
Low-converting product rescue campaign
  • Prompt – Show me products that are getting a lot of views but aren't selling well in the last 2 months. I want to create a promo campaign to move this inventory.
  • What it does – Identifies high-traffic, low-conversion products with performance metrics.
  • Business context – Promotional campaigns, inventory clearance, and pricing optimization.
Trending products for social media content
  • Prompt – What products are selling like crazy in the past 2 weeks through our recommendations? I need to feature hot items in our email blast and Instagram stories.
  • What it does – Shows best-performing products by revenue and conversion for content marketing.
  • Business context – Content planning, social media marketing, and email campaigns.
Search intent analysis for Search Engine Optimization (SEO) or pay-per-click (PPC)
  • Prompt – What are people searching for on our site this month? I want to align our Google Ads and SEO strategy with what customers actually want.
  • What it does – Lists top search terms by frequency and unique users.
  • Business context – SEO strategy, PPC keyword planning, and customer intent analysis.
Mobile and desktop engagement comparison
  • Prompt – Are customers clicking on our recommendations, and how often do those clicks turn into purchases? I'm concerned our mobile experience might be underperforming.
  • What it does – Displays CTR and click-to-purchase rates over time.
  • Business context – Mobile optimization, UX evaluation, and engagement analysis.
Product detail page (PDP) upsell performance
  • Prompt – Where on our site are recommendations driving the most revenue—product pages, cart, checkout, or homepage? I think we might be missing opportunities.
  • What it does – Compares performance across different page types.
  • Business context – Cross-sell and upsell optimization, funnel analysis, and revenue opportunities.
Seasonal brand performance (holiday campaign)
  • Prompt – Which brands converted best through recommendations during the December holiday season? I'm planning our 2026 holiday featured brand strategy.
  • What it does – Identifies conversion leaders among brands for seasonal planning.
  • Business context – Seasonal planning, brand partnerships, and holiday campaigns.
Weekly product performance check-in
  • Prompt – What were our top revenue-generating recommended products this week? The inventory team needs to know what to restock urgently.
  • What it does – Sorts products by revenue showing units sold and recommendation performance.
  • Business context – Inventory management, restocking priorities, and weekly reviews.
Complete email campaign planning package
  • Prompt – I'm planning next week's email campaign. What products are trending right now, which categories are hot, and what are people searching for? I need the full picture to create a high-converting email.
  • What it does – Pulls multiple reports (Trend, Category, and Search) to provide a comprehensive view of customer interest and product performance for email content planning.
  • Reports involved – Products and Trend Report, Category Report, and Search Report.
  • Business context – Email marketing, content curation, and campaign planning.
  • Complexity – Multi-report analysis.
Revenue drop investigation
  • Prompt – Our recommendation revenue dropped 15% this month compared to last month. Where did we lose ground? Is it engagement, conversion, or specific categories underperforming?
  • What it does – Compares multiple reports across time periods to diagnose revenue decline. Checks engagement metrics, conversion rates by product or category, and identifies problem areas.
  • Reports involved – Order Report (monthly comparison), Engagement Report, Category Report, and Products Report.
  • Business context – Performance troubleshooting, executive reporting, and optimization.
  • Complexity – Multi-report comparison.
Black Friday post-mortem analysis
  • Prompt – How did recommendations perform during Black Friday (November 24–27)? I need to know what products flew off the shelves, which widgets crushed it, and whether our homepage changes actually worked.
  • What it does – Comprehensive event analysis pulling widget performance, trending products, page type effectiveness, and overall revenue impact.
  • Reports involved – Widget Report, Trend Report, Page Type Report, and Order Report.
  • Business context – Post-campaign analysis, seasonal review, and strategy validation.
  • Complexity – Multi-report event analysis.
Customer journey optimization
  • Prompt – Walk me through the customer journey, from homepage to checkout. Where are recommendations working best, and where are we losing customers along the funnel?
  • What it does – Analyzes performance across all page types (home, product pages, cart, and checkout) to map the conversion funnel and identify drop-off points.
  • Reports involved – Page Type Report, Engagement Report, and Widget Report.
  • Business context – Funnel optimization, UX improvement, and conversion rate optimization.
  • Complexity – Multi-report funnel analysis.
Quarterly business review (QBR) package
  • Prompt – I need a complete quarterly review deck for Q1 2026. Show me overall ROI, top performing categories, trending products, engagement trends, and where we should focus next quarter.
  • What it does – Generates comprehensive executive dashboard pulling all key metrics, including revenue impact, category winners, product trends, engagement health, and strategic recommendations.
  • Reports involved – Order Report, Category Report, Products Report, Trend Report, and Engagement Report.
  • Business context – Executive presentations, strategic planning, and board meetings.
  • Complexity – Complete performance package
New product launch readiness
  • Prompt – We're launching a new product line in the 'Electronics' category next week. Show me how that category is currently performing in recommendations and what similar products are trending so I can set realistic targets.
  • What it does – Analyzes category performance and identifies comparable trending products to benchmark expectations for new launch.
  • Reports involved – Category Report, Trend Report (filtered by category), and Products Report.
  • Business context – Product launches, target setting, and competitive benchmarking.
  • Complexity – Multi-report strategic analysis.
Budget allocation decision
  • Prompt – I have $50K to spend on optimizing our recommendation system. Should I invest in improving homepage widgets, product page upsells, or cart abandonment recommendations? Show me where the biggest opportunity is.
  • What it does – Compares revenue potential and current performance across page types and widgets to identify highest-ROI optimization opportunities.
  • Reports involved – Page Type Report, Widget Report, and Order Report.
  • Business context – Budget planning, resource allocation, and ROI optimization.
  • Complexity – Multi-report investment analysis.
Underperforming widget deep dive
  • Prompt – Our 'Customers Also Bought' widget on product pages has terrible click-through rates. What products are being shown there, and are they actually relevant? Maybe we should change the algorithm or kill the widget entirely.
  • What it does – Analyzes specific widget performance, products being recommended, and compares against other widgets to inform optimization or removal decision.
  • Reports involved – Widget Report, Products Report (filtered by widget), and Page Type Report.
  • Business context – Widget optimization, product relevance, and UX improvement.
  • Complexity – Multi-report diagnostic analysis.
Competitor category comparison
  • Prompt – Our competitor is crushing it in the 'Home Décor' category. How are we performing in that category through recommendations? Are we showing the right products, and what's our conversion rate compared to other categories?
  • What it does – Analyzes specific category performance with competitive context, product mix, and relative conversion rates.
  • Reports involved – Category Report (Home Décor), Trend Report (category filter), and Products Report (category filter).
  • Business context – Competitive analysis, category strategy, and market positioning.
  • Complexity – Multi-report competitive analysis.
Search-to-recommendation gap analysis
  • Prompt – What are customers searching for, and are we actually recommending those products? I think there's a disconnect between search intent and what our recommendation engine is showing.
  • What it does – Compares top search terms against trending and recommended products to identify gaps in recommendation strategy.
  • Reports involved – Search Report, Products Report, and Trend Report.
  • Business context – Search-recommendation alignment, product discovery, customer intent matching
  • Complexity – Multi-report gap analysis.
Seasonal trend forecasting
  • Prompt – Summer is coming. What products and categories performed best through recommendations last summer (June–August 2025)? I need to plan inventory and marketing campaigns.
  • What it does – Historical performance analysis of products, categories, and brands during comparable seasonal period for forecasting and planning.
  • Reports involved – Trend Report (seasonal period), Category Report (seasonal), and Products Report (seasonal).
  • Business context – Seasonal planning, inventory forecasting, and campaign preparation.
  • Complexity – Multi-report historical analysis.
Mobile app and website performance
  • Prompt – Is our mobile app recommendation performance keeping up with our website? I'm worried we're not optimizing for mobile shoppers. Show me engagement and conversion across platforms.
  • What it does – Compares engagement metrics, CTR, and conversion rates to identify platform-specific optimization needs.
  • Reports involved – Engagement Report and Order Report (platform segmentation if available).
  • Business context – Platform optimization, mobile strategy, and app development priorities.
  • Complexity – Cross-platform comparison.
Brand partnership ROI validation
  • Prompt – We're paying $30K/month to feature 'Nike' products prominently in recommendations. Is it worth it? Show me Nike's performance compared to other brands and whether it's actually driving incremental revenue.
  • What it does – Analyzes specific brand performance, revenue contribution, and comparative metrics to validate partnership ROI.
  • Reports involved – Trend Report (brand: Nike), Products Report (Nike products), and Order Report.
  • Business context – Partnership evaluation, vendor negotiations, and budget justification.
  • Complexity – Multi-report ROI analysis.
Clearance sale product selection
  • Prompt – I need to run a clearance sale next month. Which products have high visibility in recommendations but aren't converting, and which categories have the most dead inventory being shown?
  • What it does – Identifies high-impression, low-conversion products and underperforming categories to target for clearance promotions.
  • Reports involved – Conversion Report (high traffic, poor conversion), Category Report, and Products Report.
  • Business context – Clearance planning, inventory liquidation, and promotional strategy.
  • Complexity – Multi-report product selection.
Homepage redesign validation
  • Prompt – We redesigned the homepage 2 weeks ago. Did recommendation performance improve? Compare engagement and revenue before and after the redesign—I need to justify this to the CEO.
  • What it does – Time-based comparison of engagement metrics, widget performance, and revenue before and after homepage changes.
  • Reports involved – Engagement Report (comparative), Widget Report (comparative), and Order Report (comparative).
  • Business context – Design validation, stakeholder reporting, and A/B test analysis.
  • Complexity – Multi-report before and after comparison.
Customer lifetime value (CLV) by recommendation engagement
  • Prompt – Are customers who engage with recommendations more valuable long-term? Show me purchase frequency and revenue for customers who click on recommendations versus those who don't.
  • What it does – Analyzes customer segmentation and value metrics based on recommendation engagement behavior.
  • Reports involved – Order Report and Engagement Report (customer segmentation).
  • Business context – Customer value analysis, retention strategy, and personalization ROI.
  • Complexity – Multi-report customer segmentation.
Cross-category recommendation strategy
  • Prompt – Are we successfully cross-selling between categories? For example, when someone buys 'Electronics,' are we recommending 'Accessories' and do those actually convert?
  • What it does – Analyzes cross-category recommendation patterns and conversion effectiveness.
  • Reports involved – Category Report, Products Report, and Widget Report (cross-category).
  • Business context – Cross-sell strategy, product bundling, and category relationships.
  • Complexity – Multi-report relationship analysis.
Influencer campaign product selection
  • Prompt – Which products under $100 are trending right now with high engagement but aren't mainstream yet? I'm briefing influencers for an 'undiscovered gems' campaign.
  • What it does – Identifies trending products in specific price range with growth potential but not yet saturated.
  • Reports involved – Trend Report (filtered), Products Report (price filter), and Engagement Report.
  • Business context – Influencer marketing, product curation, and campaign planning.
  • Complexity – Multi-report filtered analysis.
Year-over-year (YoY) growth analysis
  • Prompt – How did March 2026 recommendation performance compare to March 2025? Show me YoY growth in revenue, engagement, and top performing categories.
  • What it does – Comprehensive year-over-year comparison across all key metrics to measure growth and identify trends.
  • Reports involved – Order Report (YoY), Engagement Report (YoY), and Category Report (YoY).
  • Business context – Annual planning, growth tracking, and investor reporting.
  • Complexity – Multi-report temporal comparison.
Abandoned cart recovery opportunity
  • Prompt – What products are frequently added to cart but not purchased? Are we showing effective recommendations on the cart page to save those sales?
  • What it does – Identifies cart abandonment patterns and evaluates cart page recommendation effectiveness.
  • Reports involved – Products Report, Page Type Report (cart page), and Engagement Report.
  • Business context – Cart abandonment, conversion optimization, and revenue recovery.
  • Complexity – Multi-report abandonment analysis.

Output

All reports generate a canvas with the following:

  • Summary cards with key metrics.
  • Interactive expandable tables (where applicable).
  • Tooltips for detailed values.
  • Responsive design for sharing.
  • Ready for stakeholder presentations.

The canvas contains self-contained HTML files that can be

  • Opened in any browser.
  • Shared through email.
  • Embedded in presentations.
  • Exported to PDF for reports.
Keep in mind that the canvas reports Opal generates can vary across different conversation threads.

Additional example prompts

The following are quick-fire examples showing the flexibility of natural language queries:

Executive and strategy

  • Are our recommendations actually influencing what people buy?
  • Show me our recommendation ROI for the past quarter.
  • What's the business impact of our recommendation engine?
  • Which business metrics improved since we launched recommendations?
  • Is the recommendation engine paying for itself?

Optimization and testing

  • Which of our "You May Also Like" sections should we kill or redesign?
  • Are recommendations on mobile converting as well as desktop?
  • What page types need better recommendation placement?
  • Should we show more or fewer products in each recommendation widget?
  • What's the optimal number of recommendations to show on product pages?

Campaign planning

  • What products should we promote based on what's trending right now?
  • Which categories should we push harder in our email campaigns?
  • What brands are customers responding to in recommendations?
  • What should be in our Memorial Day sale based on recommendation data?
  • Which products would make the best flash sale candidates?

Inventory and operations

  • What products are people viewing but not buying?
  • Which high-traffic products need price adjustments?
  • What should we restock based on recommendation performance?
  • Are we showing out-of-stock products in recommendations?
  • What slow-moving inventory should we feature in recommendations?

Customer insights

  • Is our search driving sales or just traffic?
  • What are customers looking for that we might not be promoting?
  • How engaged are customers with our recommendations?
  • Do returning customers engage with recommendations more than new visitors?
  • What are the behavioral differences between recommendation clickers and non-clickers?

Pricing and promotions

  • Are discounted products performing better in recommendations than full-price items?
  • Should we promote high-margin or high-volume products in recommendations?
  • What price points convert best in our recommendation widgets?
  • Are premium products getting enough visibility in recommendations?

Content and SEO

  • What search terms should we target based on on-site search behavior?
  • Are recommended products aligned with our content marketing strategy?
  • What products should we create buying guides for?
  • Which categories need more SEO investment based on recommendation performance?

Personalization and targeting

  • Are personalized recommendations outperforming generic ones?
  • What customer segments engage most with recommendations?
  • Should we show different recommendations to first-time vs. returning visitors?
  • Are our VIP customers seeing and buying recommended products?

Multi-report scenarios quick reference

The following advanced prompts require insights from multiple reports to provide complete answers:

Scenario Business question Reports required Complexity
Email campaign planning What products, categories, and search trends should drive my email? Trend + Category + Search Medium
Revenue drop diagnosis Why did revenue drop? Where did we lose ground? Order + Engagement + Category + Products High
Event post-mortem How did Black Friday perform across all touchpoints? Widget + Trend + Page Type + Order High
Funnel optimization Where in the customer journey are recommendations failing? Page Type + Engagement + Widget Medium
QBR package Give me a complete quarterly business review Order + Category + Products + Trend + Engagement Very High
Launch benchmarking What should I expect from my new product launch? Category + Trend + Products Medium
Budget allocation Where should I invest my optimization budget? Page Type + Widget + Order High
Widget deep dive Should I optimize or kill this underperforming widget? Widget + Products + Page Type Medium
Search recommendation alignment Are we recommending what customers are searching for? Search + Products + Trend Medium
Seasonal forecasting What performed well last summer to plan for this year? Trend + Category + Products (historical) High
Platform comparison Is mobile app performance matching website? Engagement + Order (segmented) Medium
Partnership ROI Is our brand partnership delivering value? Trend + Products + Order High
Clearance selection Which products should go into clearance sale? Conversion + Category + Products Medium
Redesign validation Did the homepage redesign improve performance? Engagement + Widget + Order (comparative) High
Cross-sell analysis Are cross-category recommendations effective? Category + Products + Widget Medium
YoY growth How does this year compare to last year? Order + Engagement + Category (temporal) High
Cart recovery How can we recover abandoned cart revenue? Products + Page Type + Engagement Medium

Quick reference use cases

Use this table to find the reports most relevant to your role.

Function Primary use cases Recommended reports
Executive leadership ROI reporting, board presentations, strategic decisions Order, Products, Engagement
Ecommerce manager Homepage optimization, widget performance, conversion Widget, Page Type, Engagement
Merchandising Category performance, product mix, seasonal planning Category, Products, Trend
Digital marketing SEO/PPC strategy, content planning, campaign ROI Search, Trend, Products
Email marketing Product selection, trending items, personalization data Trend, Products, Category
Social media Trending products, content ideas, product features Trend, Products
Inventory and operations Restocking priorities, slow movers, clearance candidates Products, Conversion
UX and product Team Engagement analysis, page performance, placement testing Engagement, Widget, Page Type
Analytics team Deep dive analysis, customer behavior, search patterns All reports, Search

Tips for best results

Getting the most out of Opal starts with how you ask. When using Opal with Product Recommendations, the more context and clarity you provide, the more accurate and actionable your results are. These tips help you craft better prompts so you spend less time refining outputs and more time acting on insights.

  • Be specific with date ranges
    • Good – "past 30 days" or "January 1 to March 31, 2026".
    • Bad – "recently" or "a while ago"
  • Define your success metric
    • Good – "by revenue" or "by conversion rate".
    • Bad – "best performing" (without context).
  • State your business goal
    • Good – "I want to create a promo campaign" or "for our quarterly merchandising meeting".
      • This helps Opal provide contextually relevant insights
  • Use business terms, not technical jargon
    • Good – "recommendation sections"
    • Bad – "widgets"
  • Ask follow-up questions
    • After getting a report, ask "Why is product X performing poorly?" or "Show me more details on the top category".

If you use Opti ID, administrators can turn off generative AI in the Opti ID Admin Center. See Turn generative AI off across Optimizely applications.