RichRelevance adds deep learning so new products can get recommended

The ecommerce personalization platform now generates insights from product info and other unstructured data.

Chat with MarTechBot

Many recommendation engines rely on patterns of consumer behavior — which are not very useful for new products that have no browsing or purchase histories.

That’s the view of Carl Theobald, CEO of shopping personalization platform RichRelevance, which announced Monday enhancements that are designed to overcome this “cold start” problem on retailers’ sites.

Understanding intent. The enhancements include the addition of deep learning algorithms to its AI engine, boosting its ability to glean insights from unstructured data like product descriptions, user reviews, partner data and user-generated content, in addition to shopper activity and related behavioral data.

This means, Theobald said in an interview, that his company’s natural language processing (NLP) can also better understand the intent behind search terms used by consumers.

These nuances from search and other unstructured data, he said, can include whether the consumer is looking for a “comfort fit” type of clothing, or is interested in gluten-free products. This is designed to better understand the site visitor’s needs, as well as determine when new products might be relevant.

“Our retailers don’t [currently] have that level” of understanding about their visitors, he said.

NLP-informed recommendations. As an example, he pointed to Austria-based sports gear retailer Blue Tomato, which has 60 percent of its product lineup change each year. This means that behavior-based recommendations would be guessing to suggest any of the brand new products.

RichRelevance’s new NLP can inform recommendations, he said, by a cognitive understanding of the product catalog and related info, even for items that have never been bought on that site.

This “solves the problem of longtail products and new products,” Theobald said.

Why you should care. Irrelevant product recommendations can push a would-be customer away, since it indicates the retailer doesn’t really understand the buyer.

In drawing on unstructured data, RichRelevance newly augmented NLP is designed to find and use all available clues, like a good concierge. As AI grows in power and implementation, it will increasingly be used to find those hidden hints that indicate what a visitor really wants.


Opinions expressed in this article are those of the guest author and not necessarily MarTech. Staff authors are listed here.


About the author

Barry Levine
Contributor
Barry Levine covers marketing technology for Third Door Media. Previously, he covered this space as a Senior Writer for VentureBeat, and he has written about these and other tech subjects for such publications as CMSWire and NewsFactor. He founded and led the web site/unit at PBS station Thirteen/WNET; worked as an online Senior Producer/writer for Viacom; created a successful interactive game, PLAY IT BY EAR: The First CD Game; founded and led an independent film showcase, CENTER SCREEN, based at Harvard and M.I.T.; and served over five years as a consultant to the M.I.T. Media Lab. You can find him at LinkedIn, and on Twitter at xBarryLevine.

Get the must-read newsletter for marketers.