Coveo, a leader in relevance platforms that transform search, recommendations, and personalization within digital experiences, announced a set of new features in the Coveo Relevance Cloud™ to help ecommerce providers overcome a common obstacle in creating relevant web interactions, specifically for cold-start shoppers: lack of data.

Marketing Technology News: AtScale Delivers “Live” Connection Experience for Microsoft Power BI

Many merchandisers face two key ecommerce challenges: ranking of products and personalization for new or unknown users. The first challenge includes how to get new products to rank high, as new products that have not been viewed are often ranked at the bottom of long lists — not reaching the top unless the merchandiser physically boosts them. The second challenge is becoming increasingly prevalent as shoppers are choosing to shop anonymously, or returning to sites long after the related cookies have expired.

The Coveo Relevance Cloud uses AI, machine learning, deep learning, and data to help make each and every customer interaction more relevant. The platform collects signals from all digital interactions around each customer, finds insights using AI, and then adds relevance to search results, recommendations, and personalization. The platform uses all of this intelligence to make every digital experience more relevant to each shopper or buyer, helping to improve key metrics like cart conversion rate (CVR), average order value (AOV), and ultimately, gross merchandising value (GMV).

Solving for Cold Start

The Coveo Relevance Cloud’s new personalization-as-you-go capability leverages machine learning to help deliver personalized recommendations in real-time for anonymous shoppers without high volumes of data or segmentation rules, much like a sales assistant would look for visual cues from an in-store shopper. Coveo can learn in just a few clicks and react to current shopper intent with predictive query suggestions. Retailers can now aim to deliver personalization without the need for segments, profiles, user identification, or high volumes of data.

Personalization-as-you-go is the culmination of years of research by Coveo’s data scientists, whose work has been vetted by the research community peer group at large.

“Most ecommerce companies don’t have massive amounts of transactional data. So we wanted to see what insights we could find from sparse data. Our methodology has proven that a trove of data is no longer a prerequisite for an exceptional ecommerce experience,” said Brian McGlynn, General Manager ecommerce at Coveo. “With intent-aware machine learning and predictive query suggestions, ecommerce providers can better understand their customers after just a few clicks.”

Marketing Technology News: MarTech Interview with Armen Adjemian, Co-Founder and CEO at DISQO

New or enhanced features

Other capabilities now available in the Coveo Relevance Cloud that are salient to ecommerce practitioners, include:

• Buy again recommendations: Support customer satisfaction by simplifying product repurchases for returning customers with buy again recommendations. Automatically recommend previously purchased products by propensity to purchase again.
• Headless controller for product recommendations: Launch all recommendation strategies, from cart recommenders to frequently viewed together, faster with a dedicated headless controller for developers.
• Smart catalog indexing: Simplify the management of catalog onboarding for search with automatic optimization, and avoid complex nesting and joins.
• Enriched commerce analytics and attribution dashboards: An all-in-one home for the most critical commerce metrics and visualizations. Identify trends and quickly see which improvements could be most impactful. Track business outcomes and metrics including conversion rate, average order size, average revenue per unit, and average order value.
• Snowflake reader account access: Seamlessly integrate Coveo data into an existing Snowflake database without the need to set up an automated extraction. Directly connect to live usage analytics instead of relying on manual CSV exports.