Category Management – On retail shelves, there are engines that are not attractive for what price, what quantity, for years and how much. An outdated planning cycle. Scattered data. Manually run. The decision to be slower than the shopper retailer is trying to provide services.
On paper, these sound like they're easy to fix. In fact, they were nuts that were difficult to crack. The long sales cycle, clunky technical integration, and less-than-smacking issues of training for already overworked teams continue to submit progress under “someday, maybe.”
As AI becomes better, faster and easier to use, the natural problem is: Is there a possibility that this time it's different?
The short answer is: Potentially…but only when appropriate use cases are combined with humans and habits. BCG Retailer Surveydrives 70% of the value of AI.
AI is best when the rules are clear, the data is clean and success is well defined. So some of the category management (with all spreadsheets, plantgrams, and pattern-rich logic) has been fine-tuned from the cruise control.
Let's take a look where AI is already creating dents.
Shelf: Assortment, Merchandising, and Perform
- Optimizing localized assortmentAI can help you determine which products are where by calculating demand signals, margin data, and shopper needs.
- Impact analysis, It raised $40 million in 2024, and helps CPG analyze shoppers and sales data to understand the ideal product mix and pricing for each retail account, not region.
- Arabitaa Brazil-based startup supported by Qualcomm Ventures, helps supermarkets fine-tune the corrupt order quantity using data from weather, demand patterns, inventory and shelf life.
- Digital visual merchandising and store execution.
- flagshipIt raises $5.5 million in 2024 and uses computer vision to build a digital twin retailer. This means that teams can play with layout ideas, try out different product setups and see what actually drives sales before making any real changes.
- Retailers such as The low uses aI, computer vision, digital twins test layouts, adjust placement based on seasons and weather, and respond quickly to virus trends
Signal: Forecast, Trends, Price
- Predictions to see what will come: Avoid empty shelves using real-time signals, and demand shakes before it appears at Nielsen.
- Black Swan Data (Acquisitioned by Mintel in June) has used AI to sift through social chatter and consumer behavior to predict which flavors, ingredients or products claims could be the next big thing, creating names that use AI to help food brands find an upward trend.
- Smarter pricing and promotional planningAI learns from past sales and resilience models, suggesting the best time to lower prices or run promotions, and ultimately fine-tune them to grab the flashy French cheese that shoppers have been circling for weeks.
Shoppers: Personalization and Relevance
- AI that understands shoppers.
- constructor, Expanding Series B by $25 million in 2023, personalizing discovery and search in real-time using clickstream-based AI. Target Australia reported a 9% lift and bounce rate of search revenue fell 93%.
So… why isn't everyone doing that?
Because Tech is not the only part of the AI equation. And, according to BCG, that may not be the most difficult. Data preparation, people, and processes are just as important. And that's where most teams get stuck.
in 2025 BCG Survey:
- Three data points used by merchants were incorrect
- Approximately 40% of the available technologies will be unused
- 50% of merchants were not trained in analytical tools.
Top blocker? Organizational resistance, legacy systems, and lack of time.
In short, AI models, clean data, and good UX don't move the needle unless you're working on habits, skills, and workflows that actually generate (and stick) decisions.
Continue the conversation
July 30th, Leader of BCG, Deloitte, Daisy Intelligence, AGFunder Portfolio Company Lumi ai Agfunder News and Research Director Louisa Burwood Taylor will have a live discussion of how AI is used in category management and merchandising, and what it takes to move from pilot to practice.
Webinar: AI Data Driven Category Management
July 30th | 12pm ET | Online
Find more and sign up here
