Optimizing Fuel Analytics and Operational Integrity for a Leading National Retailer

The Business Case for Self-Driving Data

Given the substantial revenue and operational complexities of fuel, this organization saw opportunities for autonomous data quality in the following areas:

• Inventory Management: Tracking fuel in thousands of underground storage tanks.
• Dynamic Pricing: Reacting to rapid pump price changes across competitive local markets.
• Financial Reporting: Having an accurate fuel reconciliation report in-hand every month.

Fuel division stakeholders recognized that with constantly fluctuating fuel prices and demand, any errors in the daily movement of fuel could carry enormous financial consequences. It was clear that to maintain their competitive edge and remain pricing and operationally compliant, they needed to completely trust their fuel data.

Unseen Data Errors and Blind Spots

Looking back at the limitations of manual oversight, it became apparent why certain data issues persisted undetected:

• Overlooking New Types of Errors: New SQL rules had to be hardcoded for every new issue, and were not able to detect anomalies in incoming sensor data or delivery logs.
• Alert Fatigue: Early efforts to automate monitoring generated excessive false positives, distracting the team and making it hard to uncover and resolve real issues.
• Delayed Discovery of Blind Spots: Once the team happened upon one of these blind spots in the form of an audit, regular bug, or transformation-based issue, the business would often report the issue before the data team was able to catch it.

Due to limited visibility, business stakeholders couldn’t easily tell which data was most reliable for planning and reporting. Delays in comparing data prevented the service techs from remediating pump issues in a timely manner.

Implementation: Rapidly Identifying Data Drift

The fuel analytics team implemented Anomalo to automate data quality monitoring across their Databricks environment. The implementation focused on:

• Autonomous Fuel Observability: The team monitors critical tables, including fuel transactions, inventory, and loyalty data. By cross-referencing fuel haulers’ data with internal tank level data, the platform ensures that the fuel truck dispatch scheduling is based on reliable data, reducing the likelihood of running out of fuel or over-ordering.
• Agentic Data Quality: Custom validation rules tailored to specific business requirements, such as store-level transaction monitoring and margin analysis. If a region or fuel grade has an impossible spike or dip, the issue can be addressed at the source.
• Integration with Power BI: Business users get real-time data quality status indicators.
• Tiered Alerting Strategy: Different alert types are routed to the appropriate teams through Microsoft Teams, email, and PagerDuty.

Key Results and Business Impact

The shift to self-driving data has provided several strategic advantages for the fuel analytics program:

• Reduced Alert Noise and Improved Signal Quality: The team minimized false positives while maintaining comprehensive coverage.
• Proactive Issue Detection: Anomalo enabled the retailer to catch data quality issues before they impacted business operations. The system successfully identified problems such as missing store data from specific locations, pricing anomalies in fuel transactions during volatile market periods, and schema changes from upstream systems that would have broken downstream pipelines.
• Expanded Coverage and Adoption: The team scaled from initial pilot tables to monitoring hundreds of critical data assets.
• Business User Empowerment: The Power BI integration transformed data quality from a technical concern into visible business value. Business partners now receive targeted alerts for their specific domains.

Innovative Operational Use Cases

Beyond traditional data quality monitoring, the retailer leveraged Anomalo for operational improvements:

• Fuel Pump Outage Detection: The team built checks to identify fuel pumps with no transactions for extended periods, alerting service technicians to potential hardware failures.
• Fuel Supply Monitoring: Custom checks monitor tank levels and predict when stores might run out of fuel, enabling proactive supply chain management.

Strategic Outcomes

The shift to automated data has provided several strategic advantages for the fuel analytics program:

The fuel leadership team stated that the platform has effectively become a foundational part of their operations. The confidence gained from knowing their fuel data is clean allows them to directly improve the company’s bottom line.

Conclusion

This retailer has turned the risk of bad data undermining their critical fuel analytics program into a major competitive advantage. Spotting and correcting bad data before it corrupts the insights fueling their core business ensures they manage the product with the greatest precision and peace of mind.

High-stakes industries like fuel and retail can’t afford to make decisions on stale or inaccurate data. See how Anomalo’s self-driving data enables your team to focus on driving growth, not fixing feeds.