This probably sounds like it’s from a different lifetime, but there was a time not long ago when product recommendations were made by an actual store associate in the store. So when Kelly went to her local boutique, her relationship with the store helped the associate recommend what she could buy next, what was coming, or what would go well with her current purchase. The smaller the store, the more personalized was her relationship with the associate. Today, personalization means something different. First of all, it’s mostly digital and hence has the capability to scale across multiple platforms and channels. And that’s a good thing! As more of commerce is being done online, the benchmark for relevant product recommendations is particularly high.
About two decades ago, when Amazon first came on the scene, it was fascinating to see the little “You might also like…” scroll on the product page. How does Amazon know what I might like, I wonder? With time, the recommendations became more sophisticated and now we just take them for granted. There are different approaches that brands, retailers, and commerce enablers might take to product recommendations:
The most sophisticated recommendations come from a conflation of Product Intelligence, Business Intelligence, and Customer Intelligence. That’s when the recommendation is going to be most relevant and hence effective. Shoppers have now gotten used to the generic recommendations. For example, when someone buys purple jeans, a list of suggested accessories probably in white shows up in a sidebar. But if the retailer in question knew from the customer’s shopping history that they actually never buy white, they would be able to suggest more customized recommendations that would work for one shopper but not for another. It’s replicating the in-store personalization experience for the digital age.
This wealth of digital data can be translated into more personalized recommendations in physical environments as well. Connected store associates can use it to access somebody’s information through a loyalty card, or it can also be fed into a beacon environment that then makes the relevant recommendation as soon as a shopper walks into the store. So if Kelly’s browsing history shows that she was looking at a blouse that matched a skirt she bought online, a beacon could let her know that it was right there in the store available in her size and she could try it if she wanted to.
In the future, these recommendations would also be conflated with Kelly’s other interests like what she shares on social media or other public portals. Different signals would combine to form the perfect offer. The new face of product recommendations promises to unleash a rather welcome phase of convenience and efficiency.