Skip to main content
Setting up AI-based recommendation

In this article, we show you how to setup AI-based recommendation widget.

Kady avatar
Written by Kady
Updated over 2 weeks ago

📖 In this guide, we will walk you through the process of setting up AI-based Boost AI Search & Discovery Recommendation widgets to enhance your storefront's appeal and functionality.


⚠️ This feature only works with Boost AI Search & Discovery V3 and above. To find out which version of our app you are using, please follow this article.


Understanding AI-based recommendation

The AI-based recommendation is automatically generated by our AI models, which are based on products commonly purchased together, have similar product data, and become more accurate over time as more order and product information becomes available.

  • [Recommendation type] recommendation cannot work without read_orders and other permissions:

    • 3 recommendation types will need your permission: FBT, Related items, and Trending products


Set up AI-based recommendation

To set up an AI-based recommendation, we need the view_order and read_order permissions to run data for the model.

As a basic flow to set up a recommendation widget, create an AI-based recommendation widget in the app’s dashboard:

1. Go to Recommendation > Recommendation widgets.

2. Click Add new widget.

3. On the Recommendation Type section, select Product or Cart page > select Frequently bought together or Related items > click Next.

⚠️ Currently, AI-based recommendation is only available for Frequently Bought Together and Related items recommendation type.

4. On Recommended product settings, select AI-based recommendation, in the previous step, if you have chosen:

Frequently Bought Together

  • The Primary algorithm is set by default as AI-based Frequently Bought Together and will recommend products that are more often browsed and purchased by other customers in a single transaction, based on items added to the customer’s cart, and the shopping history of all customers on your store.

  • You can choose the Secondary algorithm just in case the Primary algorithm data is insufficient:

    • Bestsellers

    • Most viewed

    • Newest arrivals

    • AI-based Alternative products

    • Not set

Related Items

  • You can select 1 of the 3 options for Primary Algorithm:

    • AI-based Complementary products

      • Show products a user would purchase with a particular product or set.

      • For example: if a user is looking for toothpaste, this algorithm shows toothbrushes and dental floss.

    • AI-based Alternative products

      • Show products a user might consider an alternative to a particular product or set.

      • For example: if a user is looking for toothpaste, this algorithm shows other kinds.

    • AI-based Complementary products and Alternative products

      • Show products that could be complementary products and/or alternative products.

      • Ex: if a user is looking at an iPhone, this algorithm shows phone cases, screen protectors, and other types of iPhone.

  • For the Secondary algorithm, we have these options available just in case the Primary Algorithm data is insufficient:

    • Bestsellers

    • Most viewed

    • Newest arrivals

    • Not set

Personalized Recommendation

image.png

⚠️ Personalized Recommendation only applies to the Product page

  • The primary algorithm is by default AI-based personalization, providing tailored suggestions to individual shoppers based on their website activity, including purchase history, clicks, conversions, and demographics.

  • For the Secondary algorithm, we have these options available just in case the Primary Algorithm data is insufficient:

    • Bestsellers

    • Most viewed

    • Newest arrivals

    • AI-based alternative products

    • AI-based complementary products

    • Frequently bought together

    • Not set

💪 For more tips and tricks on how to ultilize Personalized Recommendation, check out our tutorial video below.

5. Click Next.


Tutorial Video

Watch our tutorial video to see how to set up a Product Recommendation Widget:

If you have any questions or need further assistance, please do not hesitate to reach out to our dedicated support team at [email protected].

Did this answer your question?