Digital shoppers expect retailers to understand and anticipate their needs – all the time, at every touchpoint, across their entire buying journey. They also want to expend as little energy as possible to find what they need. In Coveo’s 2025 Commerce Relevance Report, 45% of respondents said the biggest impact on brand perception is “finding what I’m looking for in just a few clicks.”
One of the most important levers that retailers use to meet these expectations is artificial intelligence. AI capabilities like natural language processing (NLP), machine learning, and predictive analytics make it possible to adapt to customers’ more exploratory and project-based shopping behaviors. And the area where AI is having the most significant impact? Merchandising.
According to Gartner, retailers who don’t adopt AI-driven merchandising approaches to support customer-centric strategies won’t survive. This isn’t hyperbole — it’s the new reality of retail.
AI fundamentally changes how merchandisers work, automating rote tasks, monitoring performance metrics, and freeing time for merchandisers to focus on creative initiatives. For shoppers, this translates to experiences that guide and resonate. For retailers, it means staying competitive in a marketplace where customer expectations continually rise.
The Evolving Customer Journey Starts with Ideas
The way customers approach digital shopping has shifted dramatically from the early days of online buying. Instead of starting with a specific need, shoppers increasingly begin their research with an idea or a project in mind.
A remote worker might explore how to create a more ergonomic work-from-home setup versus searching for one item, like a comfortable chair or ergonomic keyboard. This shift from need-based to idea-based shopping requires that site search adapt from old-school keyword search to product discovery and guidance.
The online buying journey is a largely circular process that puts the customer – not the product – at its center. At the top of the circle is the idea that drives a shopper to your website and the start of their search. From there, they may type a conversational query into your search bar, be presented with a series of products that support their idea, and get relevant context-appropriate products presented as search results, product suggestions, and other guided selling tools.
AI is at the heart of this new idea-focused search process. It’s what makes conversational and intent-driven product discovery possible. Conversational commerce allows customers to interact with your website in a more natural, dialogue-based way.
AI turns the search box into an “intent box” that goes beyond traditional search. When customers to ask open-ended questions like, “How do I build an outdoor barbecue?” the intent box synthesizes content, recommends relevant products, and guides the customer through their project.
It’s an experience that enhances customer trust by providing transparency around why recommendations are relevant. Customers want to know why certain products are suggested, with explanations grounded in their preferences and context.
The above results are achieved by combining multiple AI models to produce personalized results for each shopper. Here’s how.
Moving From Search to Intent With AI
The traditional search box is evolving into an “intent box” – a utility that serves as a conversational interface for product discovery. Search, once focused on matching keywords and phrases to exact content or products, becomes a much more powerful tool when it understands intent.
In our 2025 Commerce Relevance Report, 43% of consumers said they go straight to the search bar when they arrive at a website, and they do so with a specific intent in mind.
To understand and respond to this intent, search tools need to process natural language queries and understand these queries in the context of a given user and where they are within their shopping journey.
The Power of Multiple AI Models
For search to understand a complex query, match that query with the correct product or products, and still behave in a way that users expect search to behave, you need multiple AI models. The meta-model approach is seamless from the searcher’s perspective. It sits on top of the search experience, enabling specialized models to work together to consider:
- Customer behavior patterns
- Inventory levels
- Product attractiveness
- Margin requirements
- Business objectives
It’s a coordinated system, supported by AI, that makes it possible for the intent box to deliver truly personalized, contextual responses. When a merchandiser sets a goal like increasing profits, the meta-model fine-tunes all the underlying models to deliver that outcome.
For shoppers, the meta-model approach means they can start their journey with an idea rather than a specific product or need in mind. The intent box works to connect shoppers to appropriate products by synthesizing relevant content and product recommendations, grounded in the retailer’s actual inventory.
The intent box responds to a user query by displaying results, just as a search feature should. But it can also suggest next-best questions to help customers explore further. It maintains familiar options to filter and refine a search, while also allowing customers to dig deeper using conversational queries (e.g., ask a follow-up question). It recommends relevant categories and products too. In short, it makes it as simple as possible for customers to find exactly what they need.
AI and the Merchandiser’s Changing Role
There are literally no merchandising tasks that aren’t subject to automation and change thanks to AI. This is good news for merchandisers who stand to gain valuable time by outsourcing aspects of the merchandising process to different AI models.
According to McKinsey, Generative AI alone is expected to transform retail operations, adding between $400-660 billion per year in value for retail and consumer packaged goods companies. Here are some of the areas where AI is currently being focused in merchandising:
- Product Catalog Management: Updating product information, descriptions, and catalog data enrichment – tasks that previously could take months can now be completed in days.
- Pricing and Promotion Management: Setting up and launching promotional campaigns quickly and automatically adjusting pricing based on inventory levels, demand patterns, and margin requirements.
- Performance Analytics: Monitoring performance metrics like revenue, conversion rates, and click-through rates in real-time, providing actionable insights for optimization.
- Search Result Optimization: Automatically ranking and organizing search results to maximize revenue while reducing the hours traditionally spent on manual curation.
- Inventory Management: Analyzing stock levels and customer demand patterns to help prevent overstock situations and optimize product placement.
- Model optimization: Managing multiple specialized models so they work together to deliver personalized experiences based on customer behavior, preferences, and context.
There’s no future where we won’t have multiple AI models working together to improve and refine merchandising, which is why embracing a flexible approach to AI implementation is important. Each new model can be specialized for specific tasks in a coordinated system that lets merchandisers adapt quickly as needs emerge and new AI capabilities become available.
Skills for Tomorrow’s Merchandisers
The transition to AI-powered merchandising requires new technical capabilities and an evolved mindset. About 57% of new skills required in the AI era will come from upskilling existing employees. To thrive with AI, merchandisers should focus on developing technical literacy, for example, by taking “prompt engineering” courses and learning how AI models work. This is the best way to integrate AI tools into merchandising workflows and understand the potential application for AI in day-to-day work.
Merchandisers should also use data to understand the journey from the shopper’s perspective. You can do this by accessing and analyzing behavioral segmentation data to inform merchandising decisions and create more personalized experiences. As AI takes over routine tasks like spreadsheet administration and catalog updates, merchandisers should focus on strengthening strategic planning and creative problem-solving skills to drive innovation in customer experiences.
Another important skill will be data-driven decision making, with human experts learning to effectively interpret AI-generated insights and metrics that let them optimize merchandising strategies. And, finally, adaptability is an important quality for merchandisers to embrace in an AI-driven future. AI is here to stay, but it’s a tool (not a replacement) that lets merchandisers focus on more impactful work.
The Path Forward for AI in Merchandising
The implementation of AI in merchandising requires a clear vision of desired outcomes and value creation. Retailers must carefully consider how they’ll adopt and integrate AI capabilities – whether through off-the-shelf solutions, customized implementations, or building from scratch.
A customized approach often provides the best balance, allowing you to tailor solutions to specific needs while avoiding the significant technical overhead and skill requirements of maintaining complex AI systems in-house.
As AI continues to evolve with new models and capabilities, it’s important to remain adaptable. A composable commerce approach, where technology components can be easily integrated, modified, or replaced, offers the flexibility needed to future-proof merchandising operations.
This adaptability, combined with the right tools and a customer-centric focus, positions merchandising teams to realize AI’s full potential in creating personalized, engaging shopping experiences that drive both customer satisfaction and business growth.