How Chico's FAS uses AI for demand forecasting and inventory planning

Chico's FAS, the Florida-based specialty retailer operating Chico's, White House Black Market and Soma, has embedded artificial intelligence into demand forecasting and inventory planning workflows, achieving measurable improvements in margin and stock turn during fiscal 2025. The company disclosed the initiative in earnings calls and regulatory filings between late 2024 and mid-2026, describing a shift from manual, category-level planning to automated, SKU-specific allocation and promotional decisioning.

Which forecasting tasks has Chico's automated with AI?

Chico's FAS implemented AI tools to automate three core merchandising functions: demand forecasting at the store and SKU level, inventory allocation across its roughly 1,300 physical and digital touchpoints, and promotional planning including markdown timing and depth. According to investor presentations in 2025, the system ingests historical sales data, external signals such as weather patterns, and real-time point-of-sale information to generate recommendations that previously required manual analysis by merchandising teams.

The company's chief financial officer described the technology during a May 2025 earnings call as enabling "more precise matching of inventory to customer demand" by moving beyond broad category forecasts. The AI models predict sell-through rates for individual styles and sizes, adjusting allocations dynamically as new sales data arrives. This granularity allows the retailer to reduce overstock in underperforming locations whilst maintaining availability where demand is strongest.

What accuracy or performance gains has Chico's reported?

Chico's FAS reported tangible financial and operational benefits from the AI deployment in its fiscal 2025 results, disclosed in early 2026. The company achieved gross margin expansion of approximately 180 basis points year-on-year in fiscal 2025, which management partially attributed to improved inventory productivity and reduced markdowns enabled by the AI forecasting tools. Inventory turn rates improved, with the retailer ending fiscal 2025 with inventory levels down in the mid-single digits percentage-wise compared to the prior year, despite comparable sales growth.

Management noted that the AI system reduced the time lag between identifying slow-moving merchandise and executing markdowns, accelerating the markdown cycle by several weeks. This faster response reduced the depth of discounts needed to clear ageing stock. The company also reported fewer out-of-stock incidents on high-demand items, though specific fill-rate metrics were not disclosed in public filings.

How does the AI system integrate with existing merchandising workflows?

The AI platform operates alongside Chico's existing enterprise resource planning and merchandising systems rather than replacing them entirely. Merchandising teams retain oversight and can override system recommendations, but the default workflow has shifted to reviewing AI-generated allocation and promotional plans rather than building them manually from scratch. According to company statements, this change freed up planners to focus on strategic assortment decisions and vendor negotiations rather than routine allocation tasks.

The implementation followed a phased rollout beginning in late 2024, starting with demand forecasting for core product categories before expanding to allocation and promotional planning in 2025. By mid-2026, the system covered the majority of the company's SKU base across all three retail banners. Training for merchandising and planning staff emphasised interpreting AI outputs and understanding when manual intervention was warranted, particularly for new product launches without historical sales data.

What challenges and limitations has Chico's encountered?

Chico's has not publicly detailed specific obstacles in the AI rollout, but investor disclosures acknowledge inherent limitations in algorithmic forecasting. New seasonal styles, which lack historical performance data, remain difficult for the AI to predict accurately, requiring greater reliance on traditional planning methods and buyer intuition. The company also noted that external shocks such as unseasonable weather or macroeconomic volatility can temporarily reduce forecast accuracy, necessitating manual adjustments.

Integration with legacy IT infrastructure required substantial investment, though Chico's did not break out AI-specific capital expenditure in public filings. The retailer indicated that ongoing refinement of the models is necessary, with data science teams continuously retraining algorithms as new sales patterns emerge and customer preferences shift. Organisational change management, particularly helping long-tenured merchandisers trust machine-generated recommendations, was flagged as an ongoing effort rather than a one-time hurdle.

What broader implications does Chico's experience hold for mid-market fashion retailers?

Chico's FAS represents a mid-market case study in AI-driven merchandising, occupying a segment between fast-fashion giants with massive data science teams and smaller independents with limited technology budgets. The company's results suggest that meaningful margin and inventory improvements are achievable without building proprietary AI from scratch, as Chico's worked with third-party technology providers rather than developing algorithms in-house.

The focus on store-level and SKU-level granularity addresses a persistent challenge for multi-banner speciality retailers: balancing centralised buying efficiency with localised demand variation. By automating routine allocation decisions, Chico's has redirected planning resources toward higher-value activities such as trend analysis and supplier collaboration. However, the company's experience also underscores that AI is an augmentation tool rather than a wholesale replacement for merchandising expertise, particularly in categories driven by fashion cycles and subjective taste rather than purely functional demand.

The disclosed margin and inventory improvements align with broader industry observations that AI's near-term impact in fashion retail is most pronounced in operational efficiency rather than creative or customer-facing functions. For peers evaluating similar investments, Chico's trajectory suggests that measurable returns are achievable within 12 to 18 months of deployment, provided the organisation commits to workflow redesign and staff training alongside the technology acquisition.