Retail is a business built on decisions. Every day, teams make choices about which products to stock, how to arrange aisles and how to allocate inventory across stores. These decisions are at the heart of product assortment planning and the creation of an effective retail assortment plan. Each of these decisions influences sales, customer satisfaction and operational efficiency. Traditionally, these decisions were made in separate silos. Layout teams focused on aisle designs, merchandising teams focused on what products to offer and allocation teams managed supply. This separation can limit the ability to fully understand how choices in one area affect another.

I’ve spent much of my career studying how decisions in retail operations interact with each other. As a Professor of Operations, Information and Technology at IESE Business School and a member of invent.ai’s Scientific Advisory Board, I’ve worked closely with retailers and technology providers to explore how data and AI can guide decisions.

My recent research, “Store Layout Optimization with Endogenous Consumer Search” (Martínez-de Albéniz and Wagner, 2026) and “Store-Specific Assortments in the Presence of Product Constraints” (Çetin and Martínez-de Albéniz, 2025), focuses on translating complex models of shopper behavior into actionable insights that improve store performance and inform assortment optimization metrics for better merchandising decisions.

Understanding shopper behavior

One of the most important insights from my years of research is that shoppers’ movement through a store plays a central role in sales. Shoppers weigh the value of products against the effort required to find them, making the design of retail merchandising systems and optimized layouts critical to assortment optimization. Long or crowded aisles can discourage exploration, while well-structured layouts can guide shoppers to products they might not have considered.

In studies with European hypermarkets, we’ve observed that rearranging a few categories could significantly change traffic patterns. Grouping products that attract attention with complementary items increased the likelihood that shoppers would explore the aisle.

Conversely, isolating high-demand products or placing them in less accessible areas reduced visits and potential sales. Understanding these dynamics allows retailers to design layouts that encourage exploration without frustrating shoppers.

Rethinking product allocation

Assortment decisions are just as critical. Many retailers face constraints that prevent them from stocking every product in every store. This is where the best assortment planning and retail optimization software can guide better product assortment planning decisions across locations. Traditional approaches often rely on simple rules, such as expected sales or overall store market share, but these methods can miss important variations in demand between locations.

Our work in store-specific assortments introduced a network-wide approach that considers each store individually while keeping the overall supply in mind. Products are allocated where they’re most likely to sell, taking into account limited inventory, store capacity and customer preferences.

Working with large apparel retailers, we’ve seen that aligning product allocation with store-level demand patterns could increase revenue by up to 30% compared to existing methods.

Connecting layout and assortment

Layout and assortment decisions are closely connected. Where products are placed influences how shoppers interact with them, and the assortment can affect how well a layout works. For example, placing high-demand items together in a long aisle encourages shoppers to explore neighboring products. Conversely, placing low-demand products in hard-to-reach locations can mean they are never noticed.

AI platforms like invent.ai allow retailers to understand these connections.

By analyzing data on shopper paths, sales and store characteristics, AI can recommend aisle arrangements and product allocations that complement each other. Using merchandising software, AI aligns planogram designs with assortment optimization metrics, ensuring every product placement contributes to overall store performance. This approach helps create store environments that guide shoppers naturally while also supporting the retailer’s business goals.

Adapting to local store conditions

However, no two stores are alike. Shopper behavior, store layout and local demand patterns all vary. What works well in one store may not work in another.

AI makes it possible for planning teams to customize decisions for each location while still maintaining a coherent network strategy.

Aisle lengths, product groupings and inventory allocations can be tailored to local conditions, ensuring each store meets the needs of its customers and operates efficiently. AI platforms allow teams to customize retail assortment plans and product catalog strategies for each location, leveraging retail merchandising systems to meet local demand patterns.

Seeing the connections in action

For many years, retail decisions were based on experience and intuition. Today, retailers can combine that knowledge with data-driven analysis to make better decisions across stores and departments. Having worked closely with retailers and technology providers, I’ve seen firsthand that solutions differ widely in their ability to capture these interactions.

Platforms like invent.ai stand out because they make these connections visible in a way that supports, not replaces, human expertise. Teams can see how adjusting an aisle, reallocating inventory or shifting an assortment affects the overall network, while still making decisions based on their deep understanding of local shoppers. The AI highlights opportunities, flags potential trade-offs and allows teams to explore “what-if” scenarios that were previously too complex to consider.

I’ve known the invent.ai team for years and I’m confident that if you have questions about applying AI to assortment planning or store layouts, they’re happy to share insights and examples from real-world retail operations. AI is a tool, but it’s the people who know their stores best who ultimately turn insights into results.

Victor Martínez de Albéniz is Professor of Operations, Information and Technology at IESE Business School and a member of invent.ai’s Scientific Advisory Board.