How many customers are there?
Over six million customers in databases, three million real customers. How to transfer customer information to the new system? How to recognize electricity supply points assigned to the same customer and payer? How to use migration to improve data quality and then make it remain high? Customer data deduplication is a typical problem for electricity suppliers. Our client, one of the key players on the market, decided during the system migration to prepare for new challenges related to the optimization of customer retention costs.
Data cleaning is a continuous process
Initially, one-time cleaning of migrated data was assumed. At the beginning of the analysis, it turned out that the process of information exchange between systems will be continuous. The new system did not replace the old systems, but expanded their capabilities. In this situation, one-time data cleansing could not give the expected results. It was necessary to change assumptions. It consisted in the creation and implementation of a repetitive process and its maximum automation. It was necessary to develop and implement tools and procedures to ensure proper data synchronization.
Tool for customer data deduplication
After the first stage, which consisted of one-time cleansing of data from source systems, Sanmargar consultants developed a solution enabling regular and automatic data cleansing during data synchronization between systems. The solution was developed using Sanmargar DQS and Metastudio DRM technologies. The customer’s employees were given the opportunity to start the process of data cleansing on demand, without the need for the supplier’s participation. The ability to continuously monitor quality and improve data migrated to the new system was also obtained.
Data quality improvement
The problem with customer data deduplication was not only when combining data from different systems. Duplicate data was also identified in the data from one source system. This caused a number of problems with correctly identifying customers with multiple power points. The implemented solution allows not only to indicate duplicate clients in data sets subject to deduplication, but also automatically marks groups of clients in cases where different power points actually corresponded to the same client. This allows you to improve the performance of not only new but also old systems.