Personalizing ecommerce merchandising with AI in Adobe Commerce

Successful ecommerce merchandising means relevant, contextualized experiences — every time.

As shoppers, we know that when we walk into a physical store, our experience will look and feel just like any other customer who walks into that store. We may engage more with some displays than others, touch and feel products we like, or explore parts of the store relevant to us. But foundationally, product selection, placement, promotion, and other pieces of the merchandising puzzle hold constant for every customer in that store.

In the digital world, our expectations are quite different. We know that when we log into our favorite streaming service, the shows and movies we see are a unique selection specifically curated to our preferences. When we log into ecommerce marketplaces like Amazon, we are immediately presented with recommendations that meet our needs. We've come to expect that Netflix-like level of personalization whenever we shop. In fact, 84% of Gen Z says their favorite brand treats them like an individual. This holds across B2B buying and B2C shopping — customers are looking for product discovery experiences that feel hyper-relevant based on their context and affinities.

While personalizing ecommerce merchandising to this level may feel like an impossibility, with the right AI tools it becomes possible to roll out and manage at scale — with powerful positive impacts on the metrics that matter, like average order value (AOV) and conversion rate.

What to personalize — breaking down the product discovery experience

The short answer of what to personalize — all of it. Diving into shopper traffic data, we see that just over 40% of customers go directly to the search bar when they arrive on a website. However, that leaves the other ~60% of customers starting their journey by browsing through category pages. We also know that many customers find products through product recommendations with ~30% of ecommerce revenues coming from that path.

That means that rather than viewing search, category browsing, and recommendations in silos, the best brands out there consider these as pillars of a cohesive experience. Given the importance of each, Adobe Commerce provides easy-to-use AI-powered tools that leverage behavioral data to personalize product discovery across search results, category browsing, and product recommendations. Regardless of how your customer chooses to find and buy products, we’ve got you covered. Let’s explore each pillar and how you can use AI tools to unlock personalized merchandising.

Tailoring the browse experience with Intelligent Category Merchandising

Merchandisers often seek to optimize category pages by manually boosting, burying, pinning, and hiding products. While these efforts are important and help drive conversion by presenting customers with top selling products, there are a few downsides to a purely manual approach:

1. Manual merchandising is just as it sounds — manual. Boosting, burying, pinning, and hiding products takes time, effort, and careful analysis of data on each product. In fact, nearly 60% of merchandising teams spend at least 20 hours per week on manual merchandising activities.
2. The approach doesn’t scale. Merchants with vast catalogs have a lot of trouble with manual merchandising. If you have hundreds or even thousands of categories, boosting and burying products seems like an endless uphill battle.
3. It isn’t personalization. Even after manually merchandising, most category browsing pages are static, meaning they are the same for every customer and don’t change as product sales, product views, add-to-carts, and other site behaviors or engagement trends change.

With the introduction of Intelligent Category Merchandising, Adobe Commerce empowers you with an AI tool that automatically re-ranks products on each category page to boost relevance and conversion for each shopper.

Let’s look at how it works.

Built within Adobe Commerce Live Search, you can select from a set of five AI re-ranking algorithms, including Recommended for You, Most Viewed, Most Purchased, Most Added to Cart, and Trending.

You can then apply those re-ranking rules directly to any level of your category tree, from a single subcategory all the way up to your entire store (all category pages at the root level) with just a few clicks.

Once you’ve created these AI-powered rules, Adobe Sensei uses shopper data and affinities to ensure that the most relevant results are shown to each shopper immediately as they browse. Rather than bouncing or churning, shoppers can immediately find products they love. Let’s check out two examples of how this works.

Examples:

• A clothing retailer has seasonal new item releases. That retailer may choose to set the "Trending" algorithm on its "New Releases" page. Now, shoppers will immediately see the new releases with the highest momentum first, turning those winning new items into conversions on your site.
• A B2B tools and hardware company places the "Recommended for You" algorithm on a parent category of "Tools." When a buyer who only purchases DeWalt tools explores the tools category, she would see DeWalt items that are most relevant to her purchase behavior first, boosting relevance and allowing her to get what she needs to check out.

Merchandising with Adobe Commerce places you in the driver's seat, meaning you can set rules for automatic AI-powered category merchandising, then refine results with manual pinning, boosting, burying, and hiding. In-session category merchandising happens at scale, optimizing discovery, conversion, and revenue.

Personalizing site search with Intelligent Live Search Results Optimization

Just as with category browsing pages, search results are also often static, with searches returning results that are most textually relevant (i.e., closest to search terms used) without considering shopper or product context. However, curated search results can have a powerful impact on conversion rates and sales.

Intelligent Results Optimization built into Live Search mirrors the AI ranking capabilities of Intelligent Category Merchandising, but instead of anchoring on categories, rules can be set for specific queries. Live Search then blends textual relevance and AI-driven optimization to re-rank products for a highly relevant product discovery experience for searchers.

For example, if a telecommunications company sets the "Recommended for You" algorithm on a search query containing "phone" and that customer has previously purchased only Samsung devices, when the customer searches for "smartphone" or "phone," Samsung smartphones will be surfaced over other brands in the search results, boosting the relevance of the experience for that shopper.

AI re-ranking in Live Search takes the search experience from static to optimized for each customer searching your site.

Suggesting the perfect products with Product Recommendations, powered by Adobe Sensei

Product recommendations are the final piece of the product discovery puzzle and one of most foundational ways to introduce personalized merchandising on your site. Recommending relevant products is a great way to show customers that you understand their needs and boost cart value through cross-sell and upsell. In fact, online shoppers increased the average number of items in their carts by 68% based on tailored product recommendations. Having an AI-powered tool like Product Recommendations in Adobe Commerce provides the key, delivering personalized recommendations for any shopper, whether a first-timer or a loyal customer.

To make Product Recommendations a useful tool for merchandisers, we’ve made it super easy and accessible directly in the Adobe Commerce admin. The intuitive user experience allows any merchandiser to select the page type, select the recommendation type, and activate.

Product Recommendations also work across page types, so you can use them for a variety of use cases. Let’s take a look at a few examples:

1. Home page, "Best Sellers" — By using the "Most Purchased" recommendation type on your home page with the title "Best Sellers," you immediately showcase winning products to customers as they get to know your brand.
2. Category page, "Recommended for You" — By using the "Recommended For You" recommendation type on your category pages, Product Recommendations uses shopper affinity information to surface the most relevant products for that specific customer within that specific category. For example, if that customer is in the "headphones" category and has previously browsed Bose, they will be recommended Bose items first.
3. Product detail page, "Other Options You Might Like" — By using the "Visual Similarity" recommendation type on product detail pages with the title "Other Options You Might Like," you can take advantage of Adobe's visual technology to showcase different options in a similar aesthetic.
4. Checkout page, "Customers Also Bought" — By using the "Bought this, bought that” recommendation type on product detail pages with the title "Customers Also Bought,” you can drive cross-sells and upsells to boost the number of items in customers’ carts.

Ultimately, wherever you’d like to deploy AI-powered Product Recommendations, Adobe Commerce has you covered.

Getting started with personalized categories, search, and recommendations

All of these powerful features are included with Adobe Commerce as SaaS-based extensions — Live Search (Intelligent Category Merchandising and Intelligent Live Search Results Optimization) and Product Recommendations. Once you download and configure the extensions, your storefront pages are instantly tagged and prepared to analyze shopper behavior. There is no need to manually add code to your storefront.

Get started with Live Search and Product Recommendations in Adobe Commerce today and spend less time with manual tasks and more time on creating personalized merchandising and experiences — powered by AI.

Rohan Bhatt is a senior product marketing manager at Adobe, specializing in personalized commerce experiences. Prior to Adobe, Bhatt spent over five years in technology and payments consulting, advising some of the world’s largest technology companies on product and go-to-market strategies. Bhatt has an MBA and MS in Design Innovation from the Kellogg School of Management at Northwestern University.