What Is Category Management? A Guide to AI-Powered Growth in Retail

TL;DR: Category management is the process retailers and consumer goods companies use to decide which products to sell, at what price, with what promotions, and in what quantities. Most companies still plan these decisions manually, which means they end up disconnected from each other and slower to adapt than the market demands. AI changes this by predicting outcomes before decisions are made, connecting pricing, promotions, and assortment into one view, and eventually handling routine decisions automatically. This guide explains how.

Mentions of top-line growth on fourth-quarter 2025 earnings calls rose nearly 12% globally compared to the same period in 2024, according to o9's research. Revenue is back at the top of the agenda for senior leadership across retail and consumer goods. Yet the planning process most directly responsible for driving that revenue, category management, is one of the last major commercial disciplines still running largely on spreadsheets.

That gap has a measurable cost. Companies that have scaled AI across their commercial planning operations report 1.7 times higher revenue growth than those that haven't, according to o9 research. The AI in retail market was valued at $11.61 billion in 2024 and is growing at 23% per year. The organizations that modernize their category management now will build a compounding advantage over those that wait.

What Is Category Management?

Category management is the practice of treating a group of related products as a single business unit and making coordinated decisions about which products to include, how to price them, how to promote them, and how to present them on shelves.

Rather than managing each product individually, category managers look at the whole picture: which combinations of products drive the most value, how changes in price affect what customers buy, and how a promotion on one product ripples through demand across the rest of the range. In retail, this covers four interconnected levers. Assortment planning decides which products to stock and in what depth across different store formats. Pricing sets the right price for each product given competitive positioning, customer sensitivity, and margin requirements. Promotions plans the discounts and deals that drive short-term demand. And planograms (the detailed shelf layout maps that show exactly where each product sits in each store type) translate all of those decisions into physical reality.

Done well, category management directly drives revenue and profit. Managed manually across disconnected systems, it leaves significant value on the table.

Why Category Management Has Fallen Behind

Category management is one of the last major commercial planning processes that most companies still run primarily through spreadsheets and manual coordination, despite having rich data available. The problem isn't a lack of information. It's that the data, the decisions, and the people responsible for them all sit in separate places.

Demand planning and supply chain operations have benefited from digital platforms for years. Commercial planning, the part of the business most directly responsible for what gets sold and at what price, has largely missed that wave.

Category managers typically spend a disproportionate share of their time building scenarios in spreadsheets, aligning offline with finance and sales, and preparing presentations for approval meetings. By the time a recommendation has been agreed on and communicated, market conditions may already have shifted. Research from dunnhumby found that 9 in 10 shoppers changed their purchasing behaviour in response to price increases, meaning consumers are now moving faster than most manual planning cycles can follow.

Analytics teams compound the problem. Insights get generated but rarely make it into committed plans. The gap between a data team identifying an opportunity and a category manager acting on it can run to weeks, during which the window may have closed entirely.This fragmentation also affects the supply chain planning process directly. Commercial decisions made without visibility into supply reality produce forecasts that the supply chain simply can't fulfil.

The 3 AI Technologies Reshaping Category Management

Three distinct types of AI are changing how category management works, and each addresses a different part of the problem.

Machine learning (ML) is the foundation. ML analyses large volumes of sales history, competitor pricing, promotional performance, and external signals like weather or economic indicators to build predictive models. In a category management context, this means a system can estimate how much demand will shift if a price drops by 5% (a concept called price elasticity modelling, which measures how sensitive customers are to price changes), forecast how much uplift a promotion will generate before it runs, and group products into optimal assortment combinations based on how customers actually shop. These models update continuously as new data comes in, so the recommendations improve over time.

Large language models (LLMs) add a conversational layer on top. Rather than requiring a data analyst to pull a report, a category manager can ask a question in plain English, such as "why did the soft drinks category underperform last month?" and get an answer drawn from live data across pricing, promotions, inventory, and competitor activity. In the o9 platform, LLMs make planning accessible to people who aren't data scientists and can synthesise information across all four levers simultaneously, showing how a change in one area affects the others. The o9 planning learning hub covers how connected AI planning works across commercial functions.

Agentic AI is the most advanced layer. Agentic AI systems don't just analyse data or answer questions: they observe what's happening, form a plan, take action, and learn from the outcome. In category management, this means an AI agent can spot a competitor pricing move, identify which products in the assortment are most at risk, recalibrate the promotional plan accordingly, and route exceptions for human review, without a category manager initiating each step. In o9's APEX operating model, this reaches 90-95% touchless execution for routine planning decisions, freeing teams to focus on the commercial calls that genuinely require human judgement.

Why Assortment, Pricing, and Promotions Must Be Planned Together

Assortment, pricing, and promotions are not three separate decisions. They form a feedback loop where each one directly shapes the others, and planning them in isolation reliably produces outcomes that undermine the overall commercial strategy.

A promotion on a product boosts short-term demand, which clears inventory faster than the supply plan anticipated. That stock pressure may force a price increase, which reduces the product's appeal and leads the category manager to cut its assortment depth at the next review. Each individual decision made sense in isolation. Together, they weakened the category.

BCG's research on AI-powered pricing found that the most effective approaches optimise individual product decisions rather than applying category-level averages, and that doing so requires simultaneous visibility into demand, inventory, and competitor positioning. A collection of spreadsheets can't provide that. A connected platform can.

The o9 Digital Brain connects all four levers in a single environment, simulating the cross-effects of any change across revenue, profit, supply chain capacity, and the overall profit and loss (P&L) statement before a decision is committed. A category manager can see not just what a promotion will do to sales this week, but what it will do to pricing power next quarter and assortment viability next year. How supplier collaboration feeds into commercial planning decisions is also directly connected within the same platform, so supply constraints inform commercial choices in real time rather than arriving as a surprise after the fact.

What AI Category Management Actually Delivers

NVIDIA's 2024 State of AI in Retail report found that 69% of retailers using AI reported an increase in annual revenue, with 72% also reporting lower operating costs. Retailers deploying AI across pricing, assortment, and promotions report 5-15% annual revenue growth and 10-30% cost reductions, according to AllAboutAI's 2025 retail AI statistics research. In live deployments, SymphonyAI reports that AI-powered shelf planning has delivered 5% category growth and a 25% average reduction in out-of-stocks for retailers using their solutions.

The mechanism behind these results is consistent: AI shrinks the time between insight and action. In a manual process, days can pass between an analyst identifying an opportunity and a manager committing to a plan. In a connected AI platform, a promotion underperforming against target triggers an automatic flag, the system surfaces the likely cause, proposes a corrective action, and routes it for approval within hours.

For large enterprises, the scale of potential value is significant. McKinsey has estimated that generative AI could unlock between $240 billion and $390 billion in economic value for the global retail sector, with commercial planning among the highest-impact areas.

o9's expanded next-generation Category Management solution for Grocery Retail, launched in January 2026, integrates these capabilities into a single platform covering assortment, pricing, and promotion planning, with the APEX model enabling autonomous execution for routine decisions across all three levers. For a closer look at how demand signals connect into the broader commercial picture, see how o9's demand and supply planning solutions sit alongside category management in the platform.

Conclusion

Category management has been the last major commercial planning process to benefit from digitisation. That's changing, and the pace is picking up. The companies that move first will build a planning capability that compounds: better data leads to better predictions, better predictions lead to better decisions, and better decisions widen the gap with competitors still working from spreadsheets.

The shift puts better tools in the hands of category managers, handling the routine so teams can spend their time on the parts of the job that genuinely require commercial judgement: understanding customers, reading market signals, and making bold calls that a model alone can't make.

Frequently Asked Questions

What is category management in retail?

Category management is the practice of treating a group of related products as a single business unit and making coordinated decisions about which products to stock, how to price them, how to promote them, and where to place them on shelves. Rather than managing each product independently, category managers optimise the whole range for revenue, profit, and customer satisfaction, looking at how products interact with each other and how changes to one affect the others.

What is the difference between category management and demand planning?

Demand planning forecasts how much of a product will be sold so the supply chain can prepare accordingly. Category management sits upstream of that: it decides which products should be in the range, at what price, and with what promotional support. The two are closely connected because category decisions directly shape the demand signals that supply chain teams plan against. When they run in separate systems, the result is forecasts that don't reflect the actual commercial plan.

How does AI improve category management?

AI improves category management in three main ways. Machine learning builds predictive models that forecast how demand will respond to price changes, promotions, or assortment shifts before those decisions are made. Large language models allow planners to ask plain-English questions and get answers from live data across the business. Agentic AI monitors plans, identifies exceptions, and handles routine decisions automatically, so category managers can focus their time on complex commercial judgment rather than administrative coordination.

What is assortment planning and why does it matter?

Assortment planning is the process of deciding which products to stock, in which locations, and in what depth. A retailer with hundreds of stores won't carry every product everywhere: assortment planning determines the right product mix for each store format or region based on local demand, shelf space, and commercial objectives. Getting it right drives revenue by matching supply to what customers actually want. Getting it wrong creates out-of-stocks on popular lines and excess inventory on slow movers.

What does "touchless execution" mean in category management?

Touchless execution means routine planning decisions are handled automatically by AI, without a human needing to review each one individually. In a category management context, this might mean the system adjusting a replenishment order when a promotion generates more demand than planned, or resolving a pricing deviation within pre-agreed rules without involving a planner. In o9's APEX model, 90-95% of routine planning decisions reach this level of automation, freeing teams for the smaller proportion of decisions that need human input.