Omnichannel's next era is about restraint, incentives, and the LLM

The operators on this Commerce Next panel run very different businesses, from a 560-store western wear chain to an AI-driven styling service. All five kept returning to the same two ideas. Understand the customer before adding new technology, and fix incentives before adding new tools.

Moderator Jon Kosoff of Boot Barn opened with the scale behind Boot Barn’s omnichannel operation. “Boot Barn went from 273 stores in FY21 to over 560 nationwide, the largest western and workwear retailer in the US,” he said. “30% of orders are picked up in store, 40% ship from store, and the majority of returns come back to the store. We’re a true omnichannel retailer.” He closed the panel on incentives. “Don’t ding your stores or employees for online returns processed in store,” he said. “The moment you do, the teams stop logging in. Incentive structure matters, and integrating eCommerce and stores is what makes it work.”

Tapestry’s Mandeep Bhatia set a clear rule for using AI at all. “Don’t start with technology. Just because AI is there doesn’t mean you have to use it,” he said. “Root everything in a deep understanding of the customer and the associate, and associates aren’t just store staff, it’s the people working next to you.” He added, “You’re either solving a pain point or offering a vitamin, painkillers or chocolates. Figure out the balance, but it only comes from understanding what the customer needs.”

Stitch Fix CMO Debbie Woloshin said, “Our north star is our client. Our motto is help, not hype. We don’t do something just because it’s a shiny new toy. If we’re not solving a real client problem, we’re not doing our jobs.” “Instead of showing up everywhere, we show up in the right places to support their journey,” she added. “It’s a very emotional journey; people feel insecure about what to wear.”

Ulta Beauty’s Josh Friedman grounded everything in loyalty data. Ulta has 47 million loyalty members, and “95% of our sales go through our loyalty program, so we know not only what we’re selling but who we’re selling it to,” he said. “Turning that data into gold for the guest is what matters.” The company just launched Ulta AI, an on-site conversational tool Friedman said will get more personalized to a shopper’s beauty profile over time. On measurement, he said, “Definitely know how you’re going to measure it: citations, traffic, sales, influenced sales, exposure. Pick the ones you’ll focus on.”

Novi’s Kimberly Shenk reframed loyalty for the AI age. “You can build loyalty with the model, not just the consumer,” she said. “Every iteration of a model is a chance to build loyalty and be chosen in its recommendations. The way an LLM does its fan-out research for face wash is very different from peanut butter, and we just announced a patent-pending methodology that tells you exactly what the LLM prefers by product category.” “Choosing not to be part of the discovery process means missing out on being learned by the models,” she said.