Configure a Recommendations Action with Slotting

A recommendations slider with slotting action allows you to use up to four recommendation strategies in a single action to yield more diversified results.

When you're configuring this type of action, keep in mind that each recommendation strategy that you select has its own combination of recommendation algorithm and filters. For example, you could use two recommendation strategies configured with the Viewed and Also Viewed recommendation algorithm, and one is configured to show items within the same category but the other is configured to show complementary items.

Follow these steps to configure a recommendations slider with slotting action.

  1. Create a new Web experience, or navigate to the Web Experiences list page and identify an existing draft experience to which you want to add the recommendations action.
  2. Configure or modify the WHY and WHO settings.
  3. Click WHAT and then click ADD ACTION.

    Callout of the WHAT settings and of the ADD ACTION button

  4. Click Product Recommendations on the Action Type panel.

    Callout of the Product Recommendations option on the Action Type panel

  5. Click an action template for a recommendations slider with slotting.

    The recommendations action template options shown here may not reflect the options that appear in your account.

    The action template options on the Product Recommendations panel in the WHAT configuration. One of the options is a recommendations slider with slotting action template.

  6. Determine where on the page of your site you want the recommendations action to appear.
    1. Input the element selector into Element Selector to define where on the page the recommendations action should appear. See Element Selectors for more information.

      Callout of the Element Selector field on the 'Recommendation (with slotting)' action template

    2. Select an option from Insert Method to determine where the action should be inserted relative to the selected node you input in the previous step.

      Callout of the Insert Method selector on the 'Recommendation (with slotting)' action template

    3. Toggle RecheckForElement to NO if you do not want Monetate to recheck for the selected node for up to 3 seconds if it doesn't find it immediately. Toggle SelectMultipleElements to YES if you want the platform to display the recommendations at every occurrence of the selected node on the page if it finds multiple occurrences.

      Callout of the RecheckForElement option and SelectMultipleElements option on the 'Recommendation (with slotting)' action template

  7. Configure the required recommendations-related inputs for the action.
    1. Select up to four recommendation strategies.
      • When selecting multiple recommendation strategies, ensure that they all use the same product catalog. Using strategies that are configured with different catalogs can produce unexpected results.
      • Scroll through the list of recommendation strategies in a selector, or use the search bar to find one by name.
      • After you select a strategy from one selector, it still appears in the list for each of the other selectors.
      • Select None to remove a previously selected strategy before saving the configuration.

      Rec Strategy selectors A through D, with a view of the search field for a selector and with the 'None' option at the top of the list of the options in the selector, on the 'Recommendation (with slotting)' action template

    2. Type into Recommendation strategy sequence how many recommended products should come from each strategy that you selected in the previous step. See Setting the Recommendation Strategy Sequence for more information.

      Callout of the 'Recommendation strategy sequence' field on the 'Recommendation (with slotting)' action template

    3. Enter the minimum number of products the recommendation strategies must display before the action triggers a fallback recommendation strategy.

      Setting the minimum to zero prevents the fallback recommendation strategy from rendering and thus causes the action to fire in all scenarios, even if no products are recommended. This situation can also result in customers being counted in the experience without seeing recommendations.

      If in step 7a you selected any recommendation strategy with Prepend context item in recommendation enabled, then the context item counts toward meeting the threshold you set in Minimum products returned.

      Callout of the 'Minimum products returned' field on the 'Recommendation (with slotting)' action template

    4. Enter the maximum number of products to include in the recommendations results.

      If in step 7a you selected any recommendation strategy with Prepend context item in recommendation enabled, then the context item counts toward meeting the threshold you set in Maximum products returned.

      Callout of the 'Maximum products returned' field on the 'Recommendation (with slotting)' action template

    5. Optionally, enter into Pinned products the value of the id attribute for each product that you want to always appear at the beginning of the recommendations results.

      Pinned products are not subject to any recommendation filters configured in the selected recommendation strategies or to the recommendation filters configured on the Global Settings tab of the Product Recommendations page.

      If you enter product IDs into Pinned products and if in step 7a you selected one or more recommendation strategies with Prepend context item in recommendation enabled, then the products identified at the action level appear before the context product in the recommendations results.

      Callout of the 'Pinned products' field on the 'Recommendation (with slotting)' action template

    6. Optionally, select a product catalog attribute on which duplicate recommended products are removed from the results.

      Ensure that the attribute that you select is present in the product catalog configured within each recommendation strategy that you selected in step 7a. This requirement doesn't apply to any fallback recommendation strategy that you select because the Remove duplicate products based on setting doesn't apply to fallback recommendation strategies.

      Callout of the'Remove duplicate products based on' selector on the 'Recommendation (with slotting)' action template

  8. Select backup recommendation strategies from Fallback Strategy 1 and Fallback Strategy 2 that you want used to supplement the strategies that you selected in step 7a if they can't display enough recommended products to meet the minimum that you set in step 7c.

    The Fallback Strategy 1 and Fallback Strategy 2 selectors on the 'Recommendation (with slotting)' action template

  9. Select from Orientation the movement of each panel of product images in the slider.

    Callout of the Orientation selector on the 'Recommendation (with slotting)' action template

  10. If necessary, modify the code in Slider Container Template that builds the slider structure.
    • Optionally, add slide navigation buttons by adding the attributes data-prev-button and data-next-button. If you include optional arrow images, you must include elements with the attribute data-prev-button and data-next-button for them to appear.
    • Optionally, add location-related dynamic text by selecting an option from SELECT DYNAMIC TEXT and then pasting the code snippet into the code editor.

    Callout of the Slider Container Template editor and the SELECT DYNAMIC TEXT selector on the 'Recommendation (with slotting)' action template

  11. If necessary, modify the code in Slide Template that builds the product slides. See Modifying the Slide Template in Build an Intelligent Recommendation Slider with Slotting Action for more information.

    The Slide Template editor on the 'Recommendation (with slotting)' action template

  12. Optionally, modify the slider CSS. Either click CHOOSE SAVED CSS or UPLOAD CSS to insert CSS that is appended to the DOM in a <style> or <link> element. Use the placeholder #MONETATE-ID as the selector for the action's wrapper ID.

    Callout of the CHOOSE SAVED CSS and UPLOAD CSS buttons atop the CSS editor on the 'Recommendation (with slotting)' action template

  13. If necessary, modify the code in Slides Visible to change the slide visibility options. See Modifying the Slide Visibility Template in Build an Intelligent Recommendation Slider with Slotting Action for more information.

    The Slides Visible editor on the 'Recommendation (with slotting)' action template

  14. Optionally, add image files to be used for the slider navigation buttons by clicking CHOOSE EXISTING, GENERATE IMAGE, or UPLOAD for Previous Button Creative and Next Button Creative, respectively.

    Callout of the CHOOSE EXISTING, GENERATE IMAGE, and UPLOAD buttons for the Previous Button Creative field and the Next Button Creative field on the 'Recommendation (with slotting)' action template

  15. Optionally, configure the action with JavaScript.

    You must contact your dedicated Customer Success Manager (CSM) and request that the action template that includes the JavaScript editor be added to your account.

    Example of the JavaScript editor on the 'Recommendation (with slotting)' action template

  16. Optionally, add one or more action conditions. Click ADD CONDITION and then navigate the categories and options to set any conditions that must be met for the action to fire. See Action Conditions for more information.

    The action condition category options accessible from the ADD CONDITION selector on the 'Recommendation (with slotting)' action template

  17. Click CREATE.

Finish building the experience by configuring the WHEN settings. Ensure that you preview the experience before activating it. See Activate an Experience for Preview and Testing for more information.

Setting the Recommendation Strategy Sequence

The number that you enter into Recommendation strategy sequence determines how many products are identified using the strategy you selected for Rec Strategy A before the same number are next identified using the strategy you selected for Rec Strategy B and so forth. The default value for this field is 1, which means the recommendations slider displays one product from each strategy and then begins the cycle again until the display is complete. A value of 2 means that two recommended products come from strategy A, and then two from strategy B, and so forth.

For example, if you configure a recommendations action with two recommendation strategies and set the strategy sequence to 1, its configuration and output would be as follows:

Illustration of a recommendation strategy sequence of 1 using two recommendation strategies

Perhaps you configure the recommendations action with two strategies and set the strategy sequence to 2. The output would be as shown here.

Illustration of a recommendation strategy sequence of 2 using two recommendation strategies

Use Cases

There are many combinations and ways to leverage a recommendations with slotting action. Here are some approaches you can use.

Encourage Cross-Selling with Similar Products

Use a recommendation strategy configured with the Viewed and Also Viewed recommendation algorithm and with a filter to show products from the same category as the one the customer is viewing. You can encourage cross-selling by also using a recommendation strategy configured with the Viewed and Later Purchased recommendation algorithm and with a filter to exclude products within the category the user is viewing.

Leverage Machine Learning and Marketer Expertise Simultaneously

With a recommendations with slotting action, you can combine algorithmically generated recommendations with manually onboarded recommendations datasets.

For example, a retailer sells winter coats. For recommendations that appear on coats pages, a recommendation strategy configured with the Viewed and Also Viewed recommendation algorithm likely recommends a variety of other coats. After the marketing team determines that scarves are a great product to cross-sell, it uploads a Recommendations dataset that contains only scarves and configures a recommendation strategy with that dataset. The retailer then creates a recommendations with slotting action that uses both of these recommendation strategies, effectively augmenting the machine's results with additional products.

Mix Algorithms

By configuring a recommendations with slotting action, you no longer need to decide, for example, between using a recommendation strategy with the Viewed and Also Viewed recommendation algorithm or one with the Purchased and Also Purchased recommendation algorithm because you can slot each of them into a single recommendations action to get the best of both worlds.

Related Articles