Improving business efficiency through data automation and BI analytics

The client is one of the largest networks of electronics and household appliances stores and has been successfully operating in the market for over 25 years.

The client approached us to solve several problems:

• Lack of clear methodologies for defining and calculating financial indicators that reflect business performance.
• Ambiguous financial reporting within the company.
• Fifteen poorly integrated primary systems with financial data.

Financial data was spread across multiple unsynchronized systems. In addition, a specific indicator was calculated using the data convenient to a particular manager to demonstrate the department's performance.

All analytical reports were generated in a system supported by third-party vendors and did not have documentation. The system did not enable company specialists to manage data sources and formulas without formal change requests to the vendor. This caused delays and additional costs.

This misled senior management, who did not receive an objective picture of the company's key business parameters. Our task was to develop a comprehensive solution that would include:

• Revision of the set of key financial indicators and developing the methodology and rules of their calculation.
• Revision of data sources and the consolidation of ‘true data’ into a single data warehouse.
• Development of analytical reports related to business intelligence (BI) tools.
• Deployment of the whole system and post-production support.

Initially, we engaged our senior financial consultant to elaborate a unified set of key financial indicators. During the business consulting phase, he came to understand the existing reporting system and business demands and emphasized the importance of good data management.

To ensure mutual understanding, we developed a glossary that defined all the concepts and terms used in the reports.

We then performed a formalization and specification of the parameters. Each one received a ‘passport’ that contained information about how it should be calculated, what data should be used for its calculation, and how to interpret the results. Finally, our team created a comprehensive ‘scorecard tree’ that incorporated 200 input parameters. We created passports for 65 of these parameters and collected and calculated data for 50 of them. In addition, we successfully integrated 15 primary systems into a single information space.

During the data engineering phase and the implementation of the IT solution, we set up a data warehouse based on the open-source software Greenplum. This also enabled the client to replace previously used proprietary applications and reduce the license cost. We also discovered a vendor lock in the existing BI system and eliminated significant data discrepancies, which helped the client reduce material losses.