What Is AI Merchandising and How Are Retailers Using It?

What Is AI Merchandising?

AI merchandising is the application of AI tools to automate, optimize and improve retail decisions related to product selection, content, presentation, location, pricing and related operations.

The term covers a collection of capabilities rather than a single application. AI merchandising is not an industry term, nor a single function provided by one software vendor.

Retailers should therefore begin with specific business decisions they want to improve, not a broad search for an all-purpose AI product.

DISCOVER: Learn why it’s so difficult to measure the return on your artificial intelligence investments.

The Technology Stack Behind AI Merchandising

The foundation of AI merchandising is made up of predictive analytics and machine learning.

“A myriad of technologies power AI merchandising, but the baseline is predictive or conventional AI capabilities tied to demand forecasting as a baseline,” says Ananda Chakravarty, research vice president for IDC Retail Insights.

Forecasts can support assortment planning, pricing, SKU rationalization and planograms, while data storage and delivery systems make their output usable by planners.

The broader stack typically draws inventory and pricing information from an order management system, customer information from a customer data platform, and product attributes from a product information management system.

Where AI Merchandising Delivers the Most Measurable ROI

The best starting point depends on a retailer’s model, maturity and existing bottlenecks.

“AI-fueled personalization, inventory, supply chain optimization and price strategy are the top three returns on capabilities and topline revenue by retailers,” Chakravarty says.

Those areas also offer relatively clear performance measures, allowing retailers to:

• Track conversion, average order value and engagement for personalization
• Forecast accuracy, stockouts, inventory turns and markdowns for inventory optimization
• Monitor margin, sell-through and promotional lift for pricing

Three examples of valuable AI merchandising use cases that serve these areas include the following:

AI-Powered Search and Product Discovery

AI is transforming onsite product search capabilities for consumers, moving shoppers beyond literal keyword matching on a retailer-s e-commerce storefront.

“AI automates search and product discovery by moving from keyword-based search through a catalog to context-aware ranking that considers the language and conditions of the buyer," says Chakravarty. The result is more relevant results, fewer dead-end searches and better discovery of long-tail products that traditional search would miss.

That shift is also merging capabilities retailers once managed as separate tools. Emily Pfeiffer, principal analyst at Forrester, says search, recommendations, personalization and merchandising have converged into a single technology category. Newer systems use external large language models “to better understand a shopper’s query, interpret natural language phrases instead of just keywords, and support back-end users with GenAI assistants.”

This convergence of functions brings retailers closer to conversational and agentic search experiences, where a shopper can describe what they need in natural language, such as “a gift for a 10-year-old who likes science.” An AI agent would then navigate the catalog, filter those results and surface recommendations without the shopper ever typing a traditional search query.

Dynamic Pricing and Promotion Optimization

Pricing applications generally begin with demand forecasts and machine learning models that estimate elasticity. Retailers then optimize against a defined objective, such as profit, revenue, sell-through or inventory clearance.

“Optimization is accomplished through adjusting pricing to match key objectives outlined by a retailer, such as profit or sell-through,” Chakravarty says.

Generative and agentic tools can make pricing systems easier for employees to use and enable them to consider more parameters. But executives still need clear rules, approval thresholds and monitoring.

A faster recommendation engine only creates value when its objectives reflect the retailer’s margin strategy, brand position and customer commitments.

CHECK OUT: These are the tech trends shaping retail in 2026.

Omnichannel AI Merchandising Across Digital and Physical Storefronts

The larger opportunity is to connect decisions across channels and functions. Chakravarty calls the transformation “a holistic one, where AI is manipulating and optimizing results not just for specific silos but across multiple functions that are derivatives of merchandising and planning.”

Those functions include space and assortment planning, pricing, product discovery, visual placement, inventory, allocation, replenishment, and merchandise financial planning.

Linking them can help retailers avoid optimizing an online experience without considering store inventory or promoting products that the supply chain cannot support. The objective is a shared decision system, not separate AI pilots for e-commerce and stores.

How To Evaluate AI Merchandising Technology and Vendors

Retailers should evaluate platforms against business fit, data readiness and operational control.

Key questions to ask when considering an AI merchandising platform include the following:

• Does it integrate with existing OMS, CDP and PIM environments?
• Does it support your highest-value use cases?
• Does it explain recommendations and allow for human intervention?
• Does it measure outcomes against agreed key performance indicators?

Leaders should also examine implementation requirements, data governance, model monitoring and the cost of expanding beyond an initial deployment.

“Evaluation should be done based on strongest fit and long-term partnership goals,” Chakravarty says.

He adds that the winning platform is not necessarily the one with the longest AI feature list but the one that can turn the retailer’s data into repeatable, governed decisions with results the business can verify.