by Alexander Kalashnikov, Lead Designer, Datamilk
For a modern e-commerce store, just showing a catalog of your products is not enough. Your site has to help users find products they’ll love. Otherwise, users will leave when they can’t easily find products they’re interested in.
Most stores rely heavily on recommendations to guide users, with good reason. Recommendations can be an extremely effective form of product discovery, and users expect them to help with finding products. However, if a user is just browsing or trying to make sense of your product lines, it’s harder to make a strong recommendation for a specific product.
Recommendations are perfect when customers know what they want
Product recommendations are extremely effective, but their use case is not all-encompassing. They’re great at helping users find a product they have on their minds. Recommendations work best when there’s a strong signal, based on the user’s behavior, that they are interested in a particular kind of product. Maybe the user looked into a certain category or, even better, used the search bar. They’re looking through your site—until a recommendation comes in with exactly what they needed.
But the paths of discovery are more complicated
When your user does not have a specific product in mind, and instead is just scrolling and tapping at whatever piques their interest, the classic product recommendations are useless. These users need stronger engagement to make sure they spend enough time on your site to run into a product they will want to buy eventually.
These cases are more frequent than it seems. It could be a first-time visitor from your latest ad campaign just exploring your stock, or your repeat customers planning their next purchases for the summer season. Most of these behaviors won’t lead to an immediate purchase, but a great UX will increase the chances they will come to shop again.
Tweaking recommendations to fit all use cases
-
Turn them into a solution for those who are just exploring your stock
If your store has a wide range of products, your visitors need to be seamlessly guided through exploring your catalog. So, rather than offering specific products that may be hit or miss, it’s easier for a customer to be intrigued by a category and continue exploring that direction. To ensure that they will interest the customer, use an AI solution that will suggest the categories specifically for each user.
-
Add more engagement when you feel your visitors are about to get bored
Our research found that nearly 30% of page drop-off occurs when a user scrolls a product listing page a few times (and potentially presses the Load More button). Thus, instead of one long vertical product listing page with similar products, a store must have more twists and turns to painlessly lead your user through categories and eventually get them to a potential purchase.
With the powers of AI, e-commerce can identify experiences where customers are seeing irrelevant content, and insert a carousel with recommended product categories in the middle of a long or repeated page scroll.
-
Make sure the component appears only where users need them
When a user with a clear intent of purchasing a particular product keeps needing to close annoying marketing pop-ups with random products, it’s a big UX issue and can drive away customers to your competitors.
So, instead of throwing recommendation components at random steps of the journey, add them where your customers are more likely to need guidance to seamlessly explore your catalog. There are obvious spots like dead-ends after an unsuccessful search. But the bottlenecks will be different for every user. Using various attention signals, AI can identify when someone is likely to drop off, letting you automatically show recommendations at the right time for each customer.
About DataMilk
DataMilk is an automated UX optimization layer for enterprise e-commerce stores. 5 minutes of engineering work will increase your sales by 6–12%. DataMilk AI analyzes your user behavior patterns, identifies areas for optimization, and launches Smart Components that maximize the profit for every session. You only integrate, the rest is automated. For more information on how to join and build the next generation of UX, please visit https://www.datamilk.ai/
