Retail cloud heavyweights are turning AI into a frontline operating system for commerce, with SAP, Microsoft, and Workday each using National Retail Federation’s NRF 2026 event to pitch end‑to‑end, agent-driven retail stacks that connect planning, operations, and customer engagement. Together, the three launches show retail AI shifting from isolated pilots to embedded “commerce anywhere” infrastructure spanning merchandising, promotions, scheduling, fulfillment, and agentic storefronts.
SAP: AI-Embedded Retail Operating System
SAP on January 8 positioned its retail portfolio as a closed-loop, AI-enhanced “retail operating system” that ties planning, execution, and engagement together. SAP Business Data Cloud’s Retail Intelligence solution, launching in the first half of 2026, reportedly harmonizes real-time data from sales, inventory, customers, and suppliers, using AI-generated simulations for demand and inventory planning to improve forecasts, cut manual planning, lower stock costs, and support omnichannel experiences that drive loyalty.
On the operations side, SAP is rolling out AI-assisted assortment management so planners can create, adjust, or retire assortments through natural language using the Joule copilot, which should ease bottlenecks on expert merchandisers. Integration of SAP Omnichannel Promotion Pricing with SAP S/4HANA Cloud Public Edition for retail and fashion enables advanced promotions such as bonus buys to be applied consistently in store and online. SAP is also deepening merchandising, segmentation, and manufacturing features for fashion wholesalers and manufacturers, and adding an Order Reliability Agent in SAP Order Management Services (planned for Q2 2026) to proactively flag and resolve order issues.
In customer engagement, SAP is introducing a storefront Model Context Protocol (MCP) server as part of SAP Commerce Cloud to make storefronts intelligible to AI assistants, enabling “agentic commerce” where products, pricing, inventory, and promotions are exposed to AI-powered shopping journeys, including on platforms like ChatGPT. The goal is a channel-less commerce experience where human and AI touchpoints share the same context.
Microsoft: Commerce Anywhere via Agentic Protocols
Also on January 8, Microsoft framed “Retail Frontier Firms” as those rearchitecting operations around agentic AI, with Dynamics 365 and its MCP at the core. MCP provides AI agents with a shared, enterprise-grade view of products, inventory, pricing, policies, and customer intent, while Agent Communication Protocol (ACP) lets agents across merchandising, supply chain, store operations, and service collaborate end to end. Payment and transaction protocols extend this to checkout and settlement, supporting trusted transactions across stores, digital channels, and conversational interfaces.
The new Dynamics 365 Commerce MCP Server (preview expected in February 2026) will expose core retail logic—catalog, pricing, promotions, inventory, carts, orders, and fulfillment—as MCP-enabled capabilities so agents can securely “discover, decide, and execute” workflows across digital, physical, and conversational channels. Microsoft points to embedded agents such as the Supplier Communications Agent and retail-specific agents like Catalog Enrichment and Personalized Shopping, custom agents built with Copilot Studio, and partner-built agents from firms like Amicis, Evenica, Argano, Sunrise, and Visionet as three practical entry paths into agentic commerce.
This approach is aimed at dissolving channel boundaries so AI agents can orchestrate journeys from social discovery to mobile checkout, in-store pickup, curbside fulfillment, and voice reordering. Agents are expected to continuously optimize inventory, pricing, promotions, and supply chain decisions behind the scenes while humans set strategy and guardrails.
Workday: AI for Frontline Operations
Workday’s January 8 announcement emphasized AI for frontline workforce management in retail and hospitality, backed by a customer base of more than 1,800 companies in those sectors. By unifying HR and finance on one platform, Workday aims to give operators a real-time view of schedules, labor costs, and staffing gaps across locations and to replace manual processes with a single experience that reduces time spent on spreadsheets and schedule fixes.
Demand forecasting in Workday Scheduling and Labor Optimization uses AI alongside historical sales, traffic, and staffing data to build more accurate schedules; early customers report up to a 67% reduction in time needed to create or update weekly schedules. The Workday Frontline Agent, expected in spring 2026, will reportedly handle last-minute shift swaps and hour limits, with early adopters seeing up to a 90% reduction in manager time spent on staffing changes. Workday research cited in the release notes that 56% of organizations report higher-than-normal frontline turnover, with nearly half expecting it to rise, sharpening the focus on retention and scheduling quality.
On hiring, Workday is integrating Paradox’s conversational tools—Workday Paradox Candidate Experience Agent and Paradox Conversational ATS—to accelerate frontline recruitment. Customers like 7‑Eleven and Ace Hardware are using these agents to automate up to 90% of hiring tasks; some have seen conversion rates above 70% and time-to-hire fall to about 3.5 days, with Ace Hardware reporting an 86% “conversation” rate after automating scheduling and screening.
What This Means for ERP Insiders
AI-first retail stacks are converging on agentic, protocol-based operating models. SAP’s Retail Intelligence and Order Reliability Agent, Microsoft’s MCP/ACP stack and Commerce MCP Server, and Workday’s agents for scheduling and hiring all point to a future where retail ERP and commerce platforms are wired for continuous, AI-driven decision-making across planning, execution, and experience. This raises expectations that core retail systems must expose business logic and data to agents via standardized protocols while keeping governance and compliance intact.
Frontline labor and merchandising decisions are prime AI orchestration surfaces. SAP’s Joule-driven assortment management, Microsoft’s MCP-connected merchandising and supply agents, and Workday’s demand forecasting and Frontline Agent illustrate how AI is being aimed squarely at SKU-level decisions and shift-level scheduling rather than just analytics dashboards. For enterprise architects and transformation leads, this implies architectures where retail ERP, workforce management, and commerce services are designed to support agent workflows in real time without fragmenting data or control.
Retailers and partners will compete on how quickly they operationalize agentic capabilities. Microsoft’s three-step adoption model (embedded, custom, partner agents), SAP’s embedded AI in suite-wide processes, and Workday’s packaged AI for scheduling and hiring show vendors trying to lower the barrier to entry while still promoting deeper operating-model change. The opportunity for system integrators and software vendors is in assembling retail-specific blueprints that bring these agentic components together into workable “commerce anywhere” patterns for segments like grocery, fashion, QSR, and big-box retail.
