How AI Is Transforming Category Management in Practice: The Lorenz Experience
Wednesday, 31 December 2025 - 10:30 AM
Topic: Company Update
In FMCG, shelf space isn't made of rubber. Every SKU fights for inches and attention. That pressure forces Category Managers to move faster with more data, fewer people, and tighter timelines.
At Lorenz, Sandra LemaΕska, Category Management Specialist, puts it plainly: "Today, Category Management is not just about shelf arrangement and what should be in the store. It's something more, because it's often an increasingly strategic role, as we base our decisions on knowledge from data."
Category Management in the Data Era
Category Management has evolved from planograms and gut feel to a data-led function. The job now blends analytics, operations, and cross-functional coordination. You make calls on assortment, shelf, pricing, and promo - and you need those calls backed by facts, not opinions.
The KPIs That Matter
- Distribution: weighted and numerical availability by channel
- Rotation: sales per store, category, brand, and SKU
- Margin: the non-negotiable benchmark for profitability
- Market share: constant competitive tracking
And that's before you add promo calendars, shopper marketing, and cross-team alignment to the mix.
The Real-World Challenge
Salty snacks is a fragmented market - discounters, small shops, supermarkets, hypermarkets. Each has different rules. Each needs a tuned assortment. "There's always a challenge in assortment management from the manufacturer's perspective," Sandra notes. "Our recommendations go to retail chains, which may or may not implement them."
From BI to AI: Why Lorenz Partnered with DS STREAM
Lorenz reached a ceiling with traditional BI. Reports explained the past but didn't guide the next move. Data volume grew, while time to analyze shrank. Scaling decisions across channels and SKUs demanded more.
That's where the partnership with DS STREAM came in - bringing advanced analytics, machine learning, and a stronger data backbone.
What Changed in Practice
- Single source of truth: a shared Data Lake brought sales, marketing, supply chain, and external market data together.
- Predictive over reactive: forecasting models for demand, promo outcomes, and trend shifts.
- Automated insights: patterns, anomalies, and opportunities surfaced without manual digging.
- Scalable setup: solutions that grow with new categories, channels, and data sources.
Why AI - and Why Now
- BI dashboards topped out at "what happened," not "what to do next."
- Data kept multiplying; analyst hours did not.
- Teams needed faster decisions than headcount could support.
- Retail partners wanted instant, defensible recommendations.
As Sandra recalls: "We were already processing a lot of data across departments. We created a Data Lake and dashboards together, but everyone wanted more."
AI in the Day-to-Day: Jobs Don't Disappear - They Shift
Worried about AI taking jobs? Sandra's view is clear: "We definitely don't take away work. If someone was worried that artificial intelligence would take away work, I have the impression that it adds a bit to me, but completely different work."
What Managers Actually Gain
- Time saved
- Automatic report updates
- No more manual collection across systems
- Instant analysis summaries
- Better insight
- Clear read on promo mechanics
- Smarter customer and store segmentation
- Assortment choices backed by data
- Faster decisions
- Recommendations for retailers in minutes
- Adjustments to market shifts in near real time
- Support for bigger strategic bets
What DS STREAM Delivered for Lorenz
- Data foundation: a unified Data Lake built from multiple departments and external feeds.
- Advanced models: beyond static reporting to predictive and strategic analytics.
- Dashboards that matter: real-time access to the right KPIs and insights.
- Machine learning in category workflows: demand forecasting, assortment optimization, promo effectiveness.
This work sits on deep FMCG experience and tech fluency (Google Cloud, Microsoft Azure, Databricks). More importantly, it fits how Category Managers actually work - which is why adoption sticks.
Where AI Adds the Most Value in Category Management
- Analysis and segmentation: shopper behavior, store clusters, mission-based baskets
- Assortment optimization: store-level line-ups by channel constraints and roles
- Promotion management: uplift prediction, cannibalization, and ROI
- Partner collaboration: automated reporting and easier retailer conversations
- Demand forecasting: smoother production and distribution planning
Practical Takeaways for Managers
- 1. Start with the data. Get your house in order: sources, quality, governance. A Data Lake and basic dashboards are often the first step.
- 2. Treat AI as a support tool. It augments judgment. You keep the steering wheel.
- 3. Solve specific problems. Pick clear use cases like assortment or promo. Prove value. Then scale.
- 4. Expect role shifts. Less manual reporting, more scenario thinking and decision-making.
The Bottom Line
Lorenz shows that bringing AI into Category Management is an evolution. You add capabilities step by step, focused on real business needs. As Sandra puts it: "Tools that support us in this work - these tools are key for us."
With experts like Kuba from DS STREAM working closely with industry specialists like Sandra at Lorenz, AI becomes a practical lever for better decisions, faster execution, and a stronger position with retail partners.
Next step: build team capability
If you're planning your team's upskilling roadmap, explore curated role-based learning paths such as the AI Learning Path for Business Analysts, the AI Learning Path for Supply Chain Analysts, or the AI Learning Path for Production Planners.
Your membership also unlocks:
