They tried to build their own AI analyst. Then they found Zenlytic. Two iconic American brands turned a week-long reporting queue into five-minute answers.

Two iconic American brands rethinking how heritage retail competes in the AI era.

• J.Crew Group: Parent company of J.Crew and Madewell, two iconic American fashion brands with a loyal, multi-decade customer base across retail and e-commerce.
• A centralized BI team supports every business area, from marketing and store operations to merchandise planning and allocation.
• Leadership: Tyler Knapp (SVP Data Analytics & Technology Strategy) and Amanda Yan (Director of BI & Analytics) are reshaping how heritage brands compete in an AI-driven retail landscape.

Self-service that wasn’t. A DIY AI agent that hit a wall at 80%.

• Self-service was never truly self-service. Power BI was the enterprise standard, but users needed to understand data models, dimensions, and filters. The refrain from stakeholders was: “You do it for me.”
• Ad hoc requests sat in a week-long queue. Strategy, marketing, and merchandising questions sat in line for an analyst to prioritize. Work that should have taken minutes took a full week.
• Monday executive reviews were consumed by data consumption. Teams spent the first half of the meeting reading reports and assembling narratives, leaving too little time for strategy.
• The DIY AI agent experiment hit a wall. 4 weeks of development, 3 rounds of training, 80% accuracy ceiling. Visualization fell short. Data storytelling was missing. Not reliable enough to ship as a business product.

A semantic layer, zero training, and a partner that ships.

• The semantic layer solved the accuracy problem: Governed business logic, verified definitions, and full data lineage grounded every Zoë response. Zenlytic had been building this for years before competitors started.
• Zero training to get value: Anyone asks a question in plain English. No dimensions, no filters, no data model knowledge required.
• Explainability builds business trust: Zoë shows how she arrived at an answer, states her limitations, and even surfaces “what questions am I not asking?”
• Built for retail, not just dashboards: Product images render directly in Zoë’s responses. Critical for fashion retail where merchants need to see the product, not just the numbers.
• A partner that ships: Amanda made a small product suggestion in her first demo. Weeks later it shipped. Zoë kept getting smarter daily based on J.Crew’s feedback.

Five-minute answers, Monday meetings transformed, and insights no analyst would have thought to pull.

• Week-long ad hoc queue eliminated: Complex questions answered in 5 minutes by the team; simple ones answered instantly by any stakeholder.
• Monday business reviews transformed: Scheduled insights deliver executive summaries before Monday meetings. Tyler queries Zoë live in front of the CEO and COO with full trust in the answer.
• Zoë surfaces insights no analyst would have thought to pull: Analyzing product preferences across 8 customer personas, Zoë independently inferred attributes like fabric, buttons, and pant cut from product hierarchy data alone.
• Analyst-grade work compressed to minutes: Cross-selling, budget reallocation, and multi-attribute customer acquisition analyses that took half a day now complete in minutes, with richer output.