Coppel’s AI Strategy Is Changing the Future of Footwear Retail

Coppel’s $4.6 billion transformation shows how seriously big retailers now treat AI in footwear.

Instead of using AI only to streamline operations, the Mexican giant is turning shoes into a live testing ground for tools that decide what gets made, how it’s priced and which styles ever reach the sales floor.

Why footwear sits at the center

Coppel has used AI for years in e‑commerce search, WhatsApp commerce and supply chain work. Those systems focus on optimization: making existing journeys smoother and faster. Predictive AI is different. It works on what does not exist yet by forecasting how shoppers will respond to future products before factories cut a single pair.

The retailer first tested this approach in women’s private-label apparel, running 462 styles through First Insight’s platform. Results were strong enough that Coppel is now expanding into men’s apparel and, crucially, men’s and women’s private-label footwear.

For shoes, where each wrong call multiplies across sizes and colors, that forward-looking signal can reshape the entire merchandising process.

How predictive AI works for shoes

The workflow starts with direct consumer input rather than old sales data. Coppel shows targeted shoppers images or digital designs of potential styles.

Those consumers answer structured questions about purchase intent, perceived value and price, and they often leave open comments on details such as silhouette, materials or color.

AI then validates and weights each response, using methods built from years of comparing feedback to real market outcomes. The output is a set of clear metrics: a value score that ranks full-price potential, a model price, plus demand and price-elasticity curves for every style.

Merchants can see which shoes deserve production slots, which need a different price point and which should be dropped before they eat budget.

Why footwear is ideal for AI testing

Shoes carry more economic risk than most categories. Return rates in direct‑to‑consumer footwear sit at some of the highest levels in retail, while online conversion lags behind apparel. When a bet misses, markdowns hurt more because inventory is spread across sizes and colorways.

Because lead times run 8 to 16 months, decisions on silhouettes, materials and volumes lock in long before launch. Predictive AI gives teams a view into likely demand while there is still time to adjust.

Killing weak concepts early, tightening assortments and backing likely winners all show up quickly in cleaner sell‑through and fewer returns. That makes footwear a perfect sandbox for tools that blend data with design judgment.

Trend signals and pricing thresholds

These platforms also reveal patterns at scale. When thousands of consumers keep reacting positively to certain attributes like a new dressy shape or a specific material direction those clusters stand out in the data. Merchants can spot which concepts are gaining momentum and which ones lack demand before that split appears in sales.

Pricing becomes more precise as well. By testing willingness to pay and watching how intent changes across price bands, predictive AI can identify the sweet spot where demand peaks. That insight helps retailers set launch prices, plan markdown ladders and understand margin potential with more confidence than waiting to see what moves on clearance.

What this means for retail’s future

Coppel’s decision to move predictive AI from apparel into footwear underlines a wider shift. The shoe wall is no longer just another rack; it is a proving ground for how AI can inform product creation, assortment and pricing in real time.

The stakes are high enough, and the variables complex enough, that even small gains in accuracy drive real profit and reduce waste.

For shoppers, that should translate into more pairs that feel right, at prices that make sense, and fewer boxes going back in return mail. For retailers, lessons learned in footwear will likely spill into other categories.

In that sense, the next generation of retail AI will be tested first on what people wear on their feet long before those ideas shape the rest of the store.