Background

Notino, Europe’s leading online beauty retailer, prided itself on efficient inventory management and accurate performance metrics. However, a discrepancy in product availability across different language versions of the website hinted at a potential issue with their stock-out evaluation metric.

The Challenge

As an analyst at Notino, I noticed inconsistencies in product availability across different country-specific versions of the website. This observation led me to suspect that the metric used for evaluating stock-outs might be flawed, potentially affecting various critical business processes.

My Approach

1. Data Collection Script:

• Developed a Python script to monitor suspected products every 5 minutes.
• The script checked product availability across all language versions of the website, each corresponding to different warehouses.

1. Extended Monitoring:

• Ran the script continuously for several days to gather a comprehensive dataset.

1. Data Analysis:

• Compared the collected data against values in Notino’s data platform.
• Conducted a thorough analysis to identify discrepancies and patterns.

1. Validation and Reporting:

• Confirmed the existence of significant discrepancies between actual availability and reported metrics.
• Prepared a detailed report of findings for stakeholders.

Key Findings

1. Metric Inaccuracy: The stock-out evaluation metric in the data platform was indeed flawed, not accurately reflecting real-time product availability.
2. Wide-ranging Impact: The inaccurate metric was affecting multiple critical business areas:

• Demand Forecasting: Leading to inventory surpluses or shortages.
• Purchasing Decisions: Causing inefficiencies in stock replenishment.
• Performance Evaluation: Resulting in incorrect KPI calculations for buyers.

1. Financial Implications: The metric flaw had potential significant financial impacts due to inventory mismanagement and misaligned incentives.

Results and Impact

1. Metric Correction: My findings led to an immediate revision of the stock-out evaluation metric.
2. Improved Inventory Management:

• More accurate demand predictions.
• Optimized stock levels across warehouses.

1. Enhanced Performance Evaluation:

• Corrected KPI calculations for the purchasing team.
• More fair and accurate quarterly performance assessments.

1. Cost Savings: Potential for significant cost savings through improved inventory management and reduced overstock/understock situations.
2. Data Integrity: Increased trust in data-driven decision-making processes across the organization.

Key Learnings

1. Continuous Monitoring: Regular checks of key metrics against real-world data are crucial for maintaining data integrity.
2. Cross-functional Impact: A single metric flaw can have far-reaching consequences across multiple departments and processes.
3. Proactive Problem-Solving: Taking initiative to investigate suspicions can lead to significant improvements in business operations.
4. Technical Skills Matter: Proficiency in programming (Python in this case) enables data professionals to conduct independent investigations and validations.
5. Data Visualization Importance: Effectively presenting findings to stakeholders is crucial for driving change.

This case study demonstrates how a data-driven approach and technical skills can uncover critical issues in large-scale e-commerce operations, leading to significant improvements in business processes and decision-making.