Together with the client, we built a list of requirements (including systems and warehouses where the data lived), metrics, dimensions, segments, and filters that were being used to create the report.
Since the client wanted to work with Excel, we adopted the ODBC software to exchange data between different proprietary databases.
With data accuracy and integrity being our number one concern, both business and technical expertise were used to identify “dirty data” within the data sources, set them aside, and clean them.
Implementation of the clean data involved creating the queries to the data source with our client, applying filters and corrections previously defined, creating KPIs, and creating the final report where users could see the data and charts in a clean and accurate way.
Together, we identified ways to connect the data and understand it, built the queries, linked the results of the queries to the output sheets, and did quality assurance.
We helped our client test the consistency of data through time and helped verify the variance between the data sources and the report. Since data is often sampled for several systems, we made sure the calculation of variance was done using unsampled reports.