Recommended product sets - decorations

Creating product sets can be very helpful in increasing the value of the whole basket. That’s why it is important to create such sets and offer them to your clients. You can compose them as you wish, according to different variables and attributes that connect the whole set. These can include color, size, brand or any other attribute that connects them. These similarities give you a better chance of appealing to your customers.

Example of use - Retail Industry

Challenge

We prepared some recommendation sets for a client in the retail industry connected with the decoration category.

Solution

We proposed an AI-driven mechanism offering the possibility of composing their own product sets based on the same value of some attributes, which in this case were color, category and series.

Based on that, clients visiting curtains or blinds could get some decorations in the same color and series as the product they were viewing. In our case, the main product was curtains and the recommended products were candles and pillows.

Users can also add the whole set to the cart, using one button.

Screenshot presenting sets

Requirements

  • Similar model trained
  • Color and series of products in the product feed – as custom attributes
  • Tracking code on the website
  • OG tags implemented – especially product:retailer_part_no which is the same as <g:id> in the product feed and og:category which is also the same as <g:product_type> in the product feed
  • Your website should have a function of adding product to cart and it should be possible to use it in JavaScript, basing on ids of product from the product feed

How to do it

Preparation of AI campaign

  1. Go to Campaign – Recommendations and create new campaign.

  2. Choose similar products recommendations because we want to get products that have common values of attributes with our main product. Define the number of products which will be returned as a set, ex. 2 or 3.

  3. Define the category from which you want to serve your products in the set. In this case it will be the “Decoration” category as we only want to serve other decorations to those products

  4. To match the color of the recommended set to the main product, use information about the item color from the product feed. And set it up with filters in such way as we present below - by choosing “current” value. Based on that, recommendations will only display products that match the color of the curtain. Also, set it in the same way regarding series attribute – choose the proper attribute and set its value as “current”

    Screenshot presenting sets

  5. Also, mark those attributes as elastic, because it will help when any products with a matching color or series are found.

  6. After you change all those settings you may go to a preview of the campaign and verify your selections.

Preparation of dynamic content

When you set your AI model, it’s time to display product sets on the website.

  1. Create a dynamic content and write an HTML + CSS + JS template which will correspond with your website and will be responsible for displaying the frame with product sets. It should contain the frame of each product – image, title, link, price etc and a button with function to add those products to the cart.

  2. Implement the insert with the AI recommendation campaign that you prepared in the first part

  3. Now we need to set displaying the campaign only for the chosen categories of products. In our case it was curtains, blinds and carpets. You can set it in the few ways, depending on how your website is built. For example, you can change the settings to display the campaign only on pages with a URL containing “curtains”, “blinds” or “carpets”. If it is not possible to do so in your case, you may refer to the OG tags and implement an additional logic in JavaScript to show the frame only if we are on one of the mentioned categories – in the separate dynamic content.

Read more

This use case belongs to the following categories:

Challenges:

Features:

😕

We are sorry to hear that

Thank you for helping improve out documentation. If you need help or have any questions, please consider contacting support.

😉

Awesome!

Thank you for helping improve out documentation. If you need help or have any questions, please consider contacting support.