SALOMON: Achieving +83%
E-Commerce Revenue by Optimizing Product Feed Data with Generative AI

Optimizing E-Commerce Visibility by Perfecting Product Feed Data

Like many retailers with a significant online presence, Salomon, a major sporting goods brand, spends most of its media budget on Search Shopping campaigns. These campaigns use Product Feed Data to determine what information appears and which brand is showcased most prominently. This data is often lifted from a retailer’s website and not optimized for this use. However, manually improving this data would be highly time-consuming.

To overcome this challenge, Salomon contacted fifty-five to help its teams enhance its product feed through Generative AI.

Designing and Testing a Product Feed Expert Prompt

Having determined that improving Salomon’s Product Feed titles and descriptions would yield the greatest performance impact, fifty-five designed and iterated a GenAI prompt that would adapt these attributes to specific formats following best practices.

The POC was tested on nearly 700 products within the U.K. market. fifty-five implemented A/B testing with FeedX. To measure the POC’s success, fifty-five monitored various KPIs, including Click-Through Rate (CTR), conversions, and revenue.

Vast, Measureable Improvements for the GenAI-Augmented Catalog

Results from A/B testing more than validated the technology’s impact on performance: Salomon’s GenAI-powered catalog induced a 43% increase in CTR, with traffic appearing roughly twice as qualified compared to «regular» traffic, which in turn led to 81% more conversions.

By aligning with platform guidelines, this project contributed not only to promoting existing products but also to putting forward products that were seldom pushed before. On the 700 products tested on the U.K. market, Salomon saw its ROAS multiplied by x2.4.