Nestlé Ukraine: From Visibility to Store Execution by Agent

Retail Execution has historically suffered from a fundamental limitation: a lack of real visibility into what is actually happening in-store.

For global organizations like Nestlé, operating across multiple categories, thousands of SKUs, and millions of store interactions, this challenge is amplified. Decisions are often made based on delayed, incomplete, or unverified data.

Nestlé Ukraine set out to change that not by improving reporting, but by redefining execution itself.

From Fragmented Execution to a Digital Twin

Nestlé’s journey with effie.ai began in 2019, initially focused on digitizing merchandising and improving data quality across operations.

Over time, this evolved into something much more powerful:

A digital twin of retail execution where every store, SKU, and action is captured, verified, and available in near real time. 

Today, the system operates at scale:

• ~45,000 store visits per month
• ~1,500 SKUs across 7 categories
• Millions of data points generated and processed continuously 

This foundation enabled the next step: moving from visibility to action.

The Foundation: AI-Ready Data

The transformation started with data — but not just more data.

AI-ready data.

Nestlé Ukraine built a structured data foundation where:

• Every shelf observation is captured with image recognition, GPS, and timestamps
• Shelf conditions are converted into a structured digital map of SKUs and positions
• Store agreements, planograms, and promo rules are unified into machine-readable requirements
• Data is synchronized in near real time with core systems 

This turns in-store chaos into something actionable:

A system that doesn’t just collect data — but can operate on it instantly.

Real-Time Shelf Intelligence at Scale

A critical enabler in this transformation was AI Image Recognition.

Deployed across Nestlé’s operations in Ukraine, it delivered measurable impact:

• 97% data accuracy (vs ~75% before)
• 60% reduction in shelf audit time
• Higher promotional availability (95% vs 84%)
• Reduced operational costs

But the real breakthrough was not accuracy.

It was speed.

Insights are generated fast enough to be used during the store visit — not after.

The Shift to Agentic Execution

With structured data and real-time shelf understanding in place, Nestlé Ukraine introduced the next layer:

AI Agents.

These are not analytics tools or dashboards.

They are systems that:

• Understand store context
• Interpret requirements
• Generate store-specific actions
• Guide execution
• Verify results automatically

This marks a clear transition: from AI as a detector to AI as an operator.

How Execution Changes on the Ground

The impact becomes most visible at the store level. A typical merchandising visit is transformed:

Before:

• Manual data collection
• Interpreting requirements
• Creating action plans
• Completing reports

With AI Agent:

• Data captured instantly via image
• Requirements automatically consolidated
• Actions generated in real time
• Execution guided step by step
• Results verified automatically

The result:

• Higher consistency across teams
• Immediate issue resolution at the shelf 

Role Transformation Across the Organization

The shift to Agentic Retail Execution is not just technological it fundamentally changes how teams operate.

Merchandiser: From Manual Executor to Guided Operator

Merchandisers no longer need to interpret complex requirements or decide what to do next.

The system:

• Provides clear, store-specific actions
• Guides execution in real time
• Ensures consistency across all visits

Their role becomes focused, efficient, and standardized —
every merchandiser performs like the best one.

Supervisor: From Control to Performance Leadership

Previously, supervisors spent their time:

• Reviewing reports
• Checking compliance
• Following up on issues after visits

With AI Agents:

• Execution is verified automatically
• Visibility is real time
• Issues are resolved during the visit

Supervisor workload is reduced by over 50%, shifting focus toward:

• Coaching
• Optimization
• Driving performance 

HQ: From Reporting to Real-Time Control

At HQ level, the transformation is even more profound.

The model shifts:

• From reporting → controlling execution
• From hindsight → real-time action
• From hierarchies → field-driven decisions
• From firefighting → prevention

Decision-making time is reduced by up to 90%.

HQ no longer analyzes what happened.

It acts on what is happening — instantly.

From Visibility to Agentic Retail Execution

Nestlé’s journey reflects a broader evolution:

2019: Simple Basic automationzation
2020–2022: Verified data and image recognition
2026: AI Agents
Today and beyond: Agentic Retail Execution 

This progression moves the organization through three stages:

1. AI-ready data
2. Real-time shelf analytics
3. Agentic automation

What This Case Really Proves

This is not just a technology deployment. It is a new operating model.

A model where:

• Data is verified at the source
• Decisions are made in real time
• Execution is guided and validated instantly
• Every level of the organization operates with full transparency

Takeaway

The most important takeaway from Nestlé’s transformation is simple: AI is no longer just helping us understand execution. It is starting to run execution.

And in Retail Execution, that changes everything. Because value is not created in dashboards.

It is created at the shelf in the moment when action happens.