J. Crew launches AI stylist to guide online shoppers

J. Crew's four-decade style legacy

Over four decades, J. Crew built a reputation for cashmere cardigans, tweed jackets, and preppy staples that attracted high-profile fans including Michelle Obama, Demi Moore, and Meghan Markle.

Style decisions were always human-made

Throughout J. Crew's history as a catalog, physical retailer, and omnichannel business, styling decisions were made by humans. That changed on Wednesday.

Style Advisor launches today

J. Crew announced Style Advisor, a consumer-facing AI tool running on Anthropic's models. It is trained on data from thousands of outfit pairings from J. Crew's photo shoots and styling principles.

Beyond the item itself

Danielle Schmelkin, chief information officer at J. Crew, says the tool replicates what stylists do: recommend how to style a piece, not just the item itself, creating that experience online.

“It's not just about the piece that you're looking for, but how to style it.”

AI shopping is growing fast

More than fifty percent of global consumers used AI to buy goods within the past three months. New tools have proliferated from Meta Platforms, Amazon, Lowe's, Ulta Beauty, and Macy's.

Retailer AI tools lag behind general AI

Fifty-six percent of consumers use AI chat or search tools including Anthropic's Claude or OpenAI's ChatGPT. Only twenty-seven percent use retailer-created AI shopping assistants, according to McKinsey.

Visual understanding is the edge

Schmelkin says the visual piece is what gives Style Advisor its distinctive edge. The tool will eventually pull in data from a consumer's past purchase history for more personalized recommendations.

“I think the visual piece is really important.”

Contextual questions guide recommendations

Style Advisor asks questions to understand a shopper's intent. If a customer says they need an outfit for a wedding in Miami that is not too formal, the tool makes recommendations based on these nuances.

Rolling out to desktop first

Style Advisor will debut on J. Crew's desktop website and is expected to later be added to the retailer's mobile app.

Multiple Claude models power the tool

Style Advisor runs on Claude's Fable five point one for complex styling judgment, Sonnet five point five and Haiku four point five for speed and efficiency, and Opus five point five was used to build…

Mapping models to the right experience

Lauren Collett, Anthropic's enterprise go-to-market leader, says the key is understanding the desired shopper experience and then mapping the right models to deliver it.

Monitoring costs with observation layer

Schmelkin says J. Crew built a back-end observation layer to monitor costs closely. The retailer wants to refine model use for customer outputs and efficiency to make the tool scalable.

Building AI fluency across the company

J. Crew created an AI fluency program starting with basics and advancing through vibe coding. An internal AI lab called Second Thread began with the tech team but now includes all corporate employees.

Tech summit democratized for all employees

For four years, J. Crew has run a two-day tech summit with workshops, hackathons, guest speakers, and vendor presentations. Initially for the tech team only, access has been democratized to all employees.

Bottom-up innovation, not top-down mandate

A J. Crew digital product designer conceptualized Style Advisor and developed a working prototype before engineers joined. Schmelkin encourages employees to chart their own path on AI rather than mandating from the C-suite.

Finding your AI a-ha moment

Schmelkin says employees must find their own a-ha moment with AI. Once they do, it can change the way they think about their own life, work, and team productivity.

“You have to find your a-ha moment and then all of a sudden…this could really change the way you think about your own life, work, and team productivity.”