Set Up Product Recommendations

Welcome to this demo of Adobe Commerce Product Recommendations powered by Adobe Sensei. Today we’ll cover both the shopper experience on the storefront and the admin experience in the back office. Product Recommendations is a powerful marketing tool to increase conversion, boost revenue, and stimulate shopper engagement. In fact, over 30% of ecommerce revenues come from Product Recommendations. To kick off our demo, we’ll start here, in Luma, our B2C company selling consumer fitness apparel. While this is B2C, Product Recommendations also supports customer-specific catalogs and pricing for B2B companies or companies with hybrid business models. Let’s start with the shopper experience. If we scroll down on our Luma homepage, we can see that Product Recommendations are service to customers in the form of Recommendation Units. Here you can see two Recommendation Units, our Recommendations based on your style, and Trending Accessories. Product Recommendations has 13 unique Recommendation types right out of the box, including behavior-based, such as Recently Viewed or Bought This, Bought That, personalized, such as Recommended For You, Item-Based, such as Visually Similar To, and Popularity-Based, such as Most Viewed or Most Purchased. For any Recommendation type you choose, Adobe Commerce automatically tags storefront pages to capture and analyze shopper behavior. There’s no need for you to add code, create data streams, or manually create eventing. It’s all done for you. If we click into a specific product and scroll down, you can see two more Recommendation Units here, Related Products, which is based on similar metadata, such as name, description, category assignment, and attributes. And we found Other Products You Might Like, which is based on our View This, View That algorithm. We’re not showing it here, but we also have a Visual Similarity Recommendation type that’s powered by Adobe Sensei and uses the same Visual Similarity technology as Adobe Stock for images. Product Recommendations can be easily deployed across different types of Commerce pages, from the Home page to Product Detail pages, Cart page, and more. We’ll see how easy it is to deploy across your site in just a bit. And of course, these Recommendation Units can be resized, branded, and updated to show or hide different data or attribute fields. Now let’s see what Product Recommendations Admin looks like. As a SaaS service included in your Adobe Commerce license, Product Recommendations has a dedicated UI directly within the Adobe Commerce Admin. Here we see the existing recommendations set up on my site, as well as reporting on key KPIs, such as click-through rate, revenue, and more, which help determine the engagement and performance of each of those Recommendation Units. Let’s set up a new Product Recommendation Unit here in the Admin to see how easy it is to create and deploy a new Recommendation Unit for our top-selling watches. So I’ll give it an internal reference name, Watches, and then I’ll select where I want to deploy this Recommendation Unit on my site. In this case, I’ll select Home page, but you can see I’m able to place this Recommendation nearly anywhere on my site. Next I can select which Recommendation Type I want. In this case, I’ll select Most Purchased, because this is meant to show our top sellers. You can already see Recommended Products update in the preview right here in the Admin to show us what our shoppers will see on the storefront. I’ll then give it a shopper-facing name, which is the label over the Recommendation Unit that your shoppers see. I’ll call it Top Selling Watches. I’ll leave 5 products in this carousel, but I can select up to 20 products to add to this Recommendation Unit. And I want this Recommendation Unit to be placed at the bottom of the main content on the Home page. Up next, we’ll need to determine which products can or cannot show up in my Product Recommendation Unit. For that, we can use Inclusions and Exclusions filtering to merchandise by things like category, price ranges, specific products, stock status, and product type. You can select multiple Inclusions and Exclusions for fine control over what appears here. In this case, I want to enable this filter and use a Category filter to show only products from the Watches category. As soon as we do that, we see the Recommended Products preview update in real-time, so we can see how our shopper’s experience differs depending on the rules we set. I can then activate the unit, and as the unit lives on the site, it will start gathering performance data, like click-through rate and revenue. It’s as easy as that. Now we can see the new Product Recommendations Unit show up on the Home page. A few other key points to note about product recommendations. It is a SaaS-based service in the form of an extension in the Adobe Commerce Marketplace, created for free with your Adobe Commerce license. It’s available for PWA, AEM, and other headless implementations. And it’s super simple to install and configure, with storefront tagging and Catalog Sync automatically deployed. It requires no provisioning or configuration of infrastructure, data pipelines, or machine learning. You just install event tracking, plug in your API key in the admin, and you’re ready to go. We have hundreds of other merchants using product recommendations today and seeing major results, such as a 20% increase in AOV, 15% increase in click-through rates on PDPs. That rounds out our demo of product recommendations in Adobe Commerce. Thanks for watching!