Retailers turn loyalty files into fuel for AI shopping tools

Retailers and payments companies are reworking loyalty programs into data engines for artificial intelligence, using verified purchase histories and customer preferences to personalize search, offers and advertising. The shift is gaining urgency as third-party tracking becomes less dependable and AI assistants begin to influence how consumers discover products.

Ulta Beauty has tied its loyalty base, which numbers more than 46 million members and accounts for about 95% of company sales, to new AI shopping channels. In April, the beauty retailer announced a partnership with Google to make its product range available inside Google’s AI Mode in Search and the Gemini app.

Ulta is also using the same loyalty information in Ulta AI, its own shopping assistant on Ulta.com, built on Gemini Enterprise for Customer Experience. The structure shows how loyalty databases can function as an identity layer for AI tools, giving a shopping agent a basis to recognize a customer and tailor recommendations to past behavior.

Sephora has moved in a similar direction. In March 2026, the beauty retailer said it launched an app inside ChatGPT that allows users to connect with their Beauty Insider loyalty profiles through conversational AI.

First-party records move to the center

The commercial logic rests on first-party data: information gathered directly from a customer’s transactions, account activity and stated preferences. For a retailer, loyalty records can show what a customer bought, when the purchase occurred, the price paid and the frequency of repeat buying. That gives AI personalization systems a stronger base than broad demographic targeting.

Fanatics has also linked customer signals to AI-powered product discovery. PYMNTS reported that the sports commerce company partnered with Google Cloud to use behavioral data from its loyalty base across digital storefronts, with the aim of presenting products when purchase intent is high.

Technology vendors serving loyalty programs are building similar infrastructure. Eagle Eye, which works with grocery, retail and hospitality brands, developed its AIR platform as a real-time loyalty and promotions engine that brings customer data, digital offers and transaction systems into one architecture, according to PYMNTS.

CVS is applying its first-party data to retail media, where retailers sell advertising tied to their own customer and transaction records. PYMNTS reported that CVS connects loyalty, transaction and behavioral insights to advertising performance across channels. CVS Media Exchange says it has built its advertising business on 90 million customer relationships across beauty, health, wellness and personal care.

Parbinder Dhariwal, vice president and general manager at CVS Media Exchange, told PYMNTS CEO Karen Webster that the combination of customer relationships, loyalty and transaction data distinguishes retail media from other channels. “Not even social media is able to piece all of those constituents together,” Dhariwal said.

Payments groups extend the model

The use of loyalty data in AI is not limited to merchants. Jill Moser, senior vice president of offers and loyalty at Mastercard, told PYMNTS that Mastercard is using generative AI to tailor offers based on a combined view of transactional, zero-party and first-party data.

PYMNTS has also reported that Avery Miller, Visa’s vice president and head of global loyalty for value-added services, and Kipp Johnson, senior director of AI solutions at Braze, described personalization as “table stakes” in loyalty. They pointed to context, relevance and AI-led orchestration of customer journeys as the areas where firms are trying to distinguish their programs.

Consumer adoption is adding pressure. More than six in 10 U.S. consumers used dedicated AI platforms such as ChatGPT, Claude and Gemini in the past year, according to PYMNTS. For retailers, the loyalty program is shifting from a mechanism for rewarding previous spending to a system designed to influence the next purchase decision.

This story draws on original reporting from PYMNTS.