Retail Intelligence: What AI Can and Can’t Do Today

At ShopTalk Fall, retail leaders from Lowe’s, ThredUp, and Tanger offered a pragmatic view of artificial intelligence, not as a futuristic promise, but as a set of tools already reshaping customer journeys, operations, and workforce roles. Moderated by Matthew Adam Smith, CEO of Two Isaacs, the session underscored that AI is most effective when rooted in customer problems, not technology hype.

Lowe’s: From inspiration to installation

Lowe’s has embedded AI across three domains: how people shop, how the company sells, and how its associates work. Tools like Style Your Space use generative AI to transform photos of living spaces into design options, while semantic search interprets intent-driven requests such as “patio furniture for a small balcony.” The Milo assistant brings a century of Lowe’s expertise into a conversational tool for customers and a companion app for associates.

The emphasis is on reliability as much as creativity. Inventory accuracy must be deterministic; shoppers need certainty on whether an item is available online or in-store. For staff, AI is positioned as augmentation turning sales teams into “super associates” who can provide project-specific advice with confidence.

ThredUp: Managing the scale of uniqueness

With more than five million unique resale items listed at any given time, discovery has always been ThredUp’s biggest challenge. The company overhauled its search and merchandising engines to be AI-driven, resulting in a 60 per cent year-on-year improvement in conversion among new customers and a 10 per cent uplift in lifetime value.

But success has been uneven. Some AI-powered features generated buzz before engagement quickly fell away, reminding the company that customers are not looking for “AI” as a feature - they want simpler discovery and relevance. ThredUp now embeds AI cautiously, with upcoming seller-side initiatives designed to reduce friction rather than showcase technology.

Tanger: Loyalty as the foundation

Tanger, which operates 40 outlet centres across the US, uses AI to strengthen a loyalty programme directly linked to customer credit cards. This allows the company to personalise offers and messaging with precision. The goal is not to automate the shopping trip but to drive foot traffic by making marketing more relevant and timely.

CEO Stephen Yalof emphasised that the physical experience remains paramount. AI can optimise communication, but it cannot replace the face-to-face interaction that defines bricks-and-mortar retail. Tanger has even created an internal steering committee to test use cases carefully without overwhelming its 450-person workforce.

A measured consensus

Across the three models - home improvement, resale, and shopping centres - the consensus was clear. AI is not a stand-alone programme, nor a panacea. It is another lever in the business toolkit, valuable when applied to specific problems: search relevance, inventory certainty, personalisation at scale. Retailers that present AI as a feature risk alienating consumers who care only about results.

The thread linking all three perspectives was restraint. As Lowe’s framed it: AI is best seen as a new tool for solving old problems - more efficient, faster, and potentially more meaningful, but never a replacement for the human connection that underpins retail.

Independently Created. Not affiliated with Shoptalk.

Unofficially Shoptalk is a ClickZ Media publication in the Events division