We just leveled up our Product Recommendations with VisualAI. From today, you can have recommendations surface products based on visual similarities to other items, including those a shopper’s shown interest in. With VisualAI now in the mix, supplementing behavioral, transactional, and imported data, we take a look at what this added sophistication means for both you and your customers.
Previously, our Recommendations have used shoppers’ behavioral and transactional data to fuel their contents. This includes historic and real-time data (for example, what shoppers have previously bought or what is currently in their cart) to surface relevant products while boosting conversions and upsells.
Now, our Product Recommendations can also be powered by VisualAI to automatically suggest items based on their visual similarity to others acknowledging attributes such as color and style.
Using VisualAI, recommendation placements can be set up to surface relevant items to the main product on a product display page (PDP), for instance. Also, behavioral-fuelled recommendations can be supplemented with VisualAI to both broaden their contents or diversify it (to encourage upsells of different but related items).
Using our Product Recommendations already?
To talk specifics, VisualAI brings with it the following new applications:
Adding VisualAI to the mix — alongside product, transactional, behavioral, and imported data — offers a more powerful and holistic bank of information (within a single platform!) to enrich Recommendations towards better conversions, average order value, and more. Here are the benefits.
Firstly, visual data quite simply enables more products to be identified as relevant, meaning you can present a higher volume of recommendations and be more likely to hit the mark.
Secondly, visual data is not dependent on shopper activity, meaning it is always available and especially useful in instances where behavioral data is insufficient.
Thirdly, having visually-similar powered Recommendations is especially helpful for shoppers looking for alternatives when the desired product is out of stock (cutting your bounce rates at the same time — win-win!).
Moreover, having artificial intelligence detect and match products based on the likes of primary color removes the need for manually tagging items with this attribute for filtering. This is particularly helpful for merchants with a broad product catalog for whom the manual alternative would be a struggle to implement and maintain.
Of course, VisualAI is going to be most helpful for merchants whose images and visual aspects of their products are significant factors in their shoppers’ purchasing decisions. Aside from this, it’s also likely to suit brands with large product catalogs in which the products have clear visual similarities between one another.
If you’re interested in learning about how VisualAI (and our Product Recommendations more broadly) can help you, you can read more about this module of our Commerce Experience Platform here, or request a demo today.
Already a Nosto customer? Feel free to reach out to your Customer Success Manager who’ll be happy to show you how you can start using this new data type to take your recommendation strategy next level.