In the portal, customers had difficulties finding the right food and beauty items, and were thwarted by a lack of product information when they did find the products they were looking for.

The U.S.-based wholesale distributor had managed to improve product information by sourcing product attributes and inputting update--the distributor improved product listings for 20,000 different items over the course of 18 months but a faster pace was required if this was to make a dent in the loss of competitiveness. AI offered a game-changing opportunity to turbocharge the company’s merchandizing and marketing.

To help find a solution, AlixPartners implemented a multi-staged, layered approach to source product attributes and images. This involved integrating data with third-party syndicated data sources, employing natural language generation (NLG), and using a workflow app. ChatGPT was deployed as part of NLG to address data quality issues brought over from the collected data, including overly long descriptions and other formatting issues.

The system we put in place enriched around 90,000 products over 10 weeks, presenting core attributes for each in the marketplace. This will save the distributor around 2,500 hours of manual curation and opens the door to further enriching the product data and metadata of an additional 50,000 products.


portal channel revenue



Related case studies