[Customer Testimonial] How a Global Consumer Goods Company Improved Shelf Share Visibility with AI

As competition intensifies in the logistics and retail sectors, shelf management has become a direct driver of profitability. In today’s consumer-centric era, shelf visibility directly influences brands’ sales performances. According to a report by Dataintelo, the AI-based shelf monitoring market is valued at USD 2.1 billion in 2024 and is projected to grow at a CAGR of 21.8%, reaching USD 15.1 billion by 2033. Similarly, Grand View Research estimates that the smart shelf (which uses RFID and computer vision to track product presence) market will grow from USD 3.3 billion in 2023 to USD 14.6 billion by 2030, with a CAGR of 23.7%.

Behind this growth lies a critical issue: stockouts and display errors. Global retailers lose USD 1.775 trillion annually—about 8.3% of total sales—due to stock shortages. Furthermore, 91% of customers who experience a product stockout choose not to return to the same store. This highlights the competitive advantage of real-time shelf visibility and instant issue resolution in retail operations.

Challenge – Managing Shelf Share Across Thousands of Stores

A leading global consumer goods company, referred to here as Company B, sells a diverse range of products through domestic and international retail channels. Despite significant efforts by sales and merchandising teams to improve shelf share, they faced several persistent challenges:

• Inefficient Manual Audits: Field representatives photographed store shelves using cameras or smartphones, manually counted products in the offices, and compiled reports in Excel. Inconsistent photo quality and varying display layouts across stores resulted in frequent errors.
• Lack of Data Reliability: Manually recorded shelf share data often reflected subjective judgments, creating discrepancies in numbers between the reports from the headquarters and stores. It was also difficult to verify planogram compliance quantitatively.
• Slow Feedback Cycle: Teams were unable to detect competitor product dominance or stockouts in real time, leading to delayed responses. Stockouts led to a decline in customer experience and sales.
• Scalability Limitations: Managing tens of thousands of stores globally required a system capable of quickly consolidating and analyzing shelf data at scale.

Solution – Building a Mobile-Based Shelf Analytics AI with Superb Platform

To modernize and automate shelf management, Company B implemented an AI-powered, mobile-based AI shelf analytics system built on Superb AI’s MLOps platform, Superb Platform. The project followed four key stages:

1. Data Collection and Curation
   • Field sales representatives captured photos of the shelves during store visits using mobile devices and uploaded them directly to Superb Platform.
   • The platform automatically stored metadata such as location, store ID, and timestamp, while providing features to conveniently filter out duplicate or low-quality images.
2. Accurate Labeling and Quality Management
   • Using the platform’s AI-assisted labeling features, the system automatically identified and pre-labeled both in-house and competitor products. Human labelers then verified and adjusted labels as needed, increasing the overall labeling accuracy.
   • Quality assurance features of the platform ensured labeling consistency and high annotation quality across multiple users.
3. Custom AI Model Development
   • Leveraging Superb Platform’s integrated MLOps pipeline, the team developed and trained a product recognition model using a relatively small dataset while achieving high accuracy.
   • Model training, evaluation, and deployment were fully automated within the platform, with real-time performance monitoring through dashboards for continuous enhancement.

(Product recognition and automated labeling through Superb Platform)

1. Mobile App Integration and Real-Time Analysis
   • The trained model was deployed to a mobile app used by field staff. By simply taking a shelf photo, the app instantly calculated product counts and shelf share percentages.
   • The results were transmitted in real time to the headquarters dashboard, enabling sales and merchandising teams to monitor performance and take necessary actions.

Benefit – Enhancing Data-Driven Decision-Making and Sales Efficiency

Following the implementation, Company B achieved the following measurable and strategic improvements:

• Time and Cost Savings: Shelf audits that previously took half an hour now required only a few seconds. Sales teams could visit more stores per day and dedicate more time to customer relationship management.
• Accurate Shelf Share Insights: The AI system provided objective, product-level shelf share data, allowing direct comparisons with competitors and data-backed negotiations with retailers.
• Faster Stockout and Planogram Violation Detection: The solution automatically detected out-of-stock items and planogram (shelf display) violations, preventing both sales losses due to stockouts and damage to brand reputation caused by poor shelf execution. A planogram (short for Plan of Program) is an optimized store display layout plan drafted based on sales data. Research shows that most retailers are losing more than 5% of their profit due to non-compliance, underscoring the competitive advantage that AI-based shelf analysis can offer.
• Data-Driven Campaign Management: Real-time shelf visibility during promotions helped managers evaluate campaign performance by store and allocate resources efficiently.
• Strengthened In-House AI Capabilities: Through collaboration between the sales and data teams, Company B gained hands-on experience in data collection, labeling, and model validation—laying the foundation for future AI initiatives.

Scalability – Expanding AI Capabilities Beyond Shelf Analytics

The data and technology developed through this project open doors to new AI applications across retail operations:

• Inventory Shortage and Stockout Prediction: Detect disappearing products in real time and predict future stockouts based on sales velocity.
• Price and Promotion Monitoring: Implement a price tag recognition model to automatically detect pricing errors or unexecuted promotions.
• Omnichannel Integration: Extend analysis from convenience stores to supermarkets, discount stores, and even e-commerce product images for unified inventory visibility.
• Integration with Other Vision Models: Leverage additional AI vision models on Superb Platform—such as safety detection, customer behavior analytics, or ad performance measurement—to transform retail environments into intelligent spaces.

Conclusion – Why AI Shelf Analytics Is Essential for Brand Competitiveness

The accuracy and speed of shelf management are now critical to maintaining brand competitiveness. By adopting Superb AI’s mobile-based shelf analytics system, Company B dramatically enhanced data-driven decision-making and sales efficiency. With global losses mounting from stockouts and planogram non-compliance—and the rapid growth of the AI shelf monitoring and smart shelf markets—this innovation represents a crucial step for all consumer goods and retail companies.

Through Superb Platform, enterprises can cover all steps of AI development—from data collection and labeling to model training and deployment—in a single space, reducing barriers to AI adoption while advancing sales and marketing strategy to the next level. This case clearly demonstrates how AI can revolutionize shelf management and retail execution.