Reprogramming Retail Visibility

Client: Matalan

Matalan competes in an intensely challenging UK retail landscape where larger players dominate visibility through scale and spend. With no ability to increase budgets, the brand needed a smarter way to grow. Google Shopping - its most important ecommerce channel - was underperforming, not due to weak media optimisation but because Matalan’s product feeds lacked the structured clarity Google’s algorithms require. Thousands of SKUs featured inconsistent, human‑led descriptions that omitted essential attributes such as colour, material, and fit, suppressing impressions and limiting conversion potential. Manual optimisation across tens of thousands of constantly changing items was impossible.

CSA, part of Havas Media Network, built a bespoke generative‑AI‑driven solution: the Shopping Feed Modifier AI Tool. Rather than using GenAI for copywriting, the platform was engineered as a performance technology that ingests raw product data and automatically enriches, standardises, and structures titles and descriptions to match Google Shopping’s ranking logic. It adds missing attributes, enforces category‑level formatting, and ensures human oversight. Ultimately, the tool changed how Google understood Matalan’s entire product inventory. A time‑based regression causal analysis isolated true incremental impact. In just four weeks, the tool delivered incremental revenue in a single category, driven by major impression gains: Kids’ Dresses, Men’s Suiting, and Women’s Nightwear, all without increasing media spend. With projected annual incremental revenue, the work demonstrates how operational AI can unlock scalable, commercially meaningful performance advantages.