SAS Visual Analytics

Client

Our Client is a data analytics company based in India providing customer value management solutions. The organization specializes in telecom data analytics and serves many telecom service providers worldwide.

Objective

The analytics framework should be able to handle huge data volumes and should also be customizable for different telecom providers. Every telecom provider uses their own data mart and even within a given telecom company, the CDR data format could vary from tower to tower. The data also need to be analyzed almost in real time, so as to identify trends that need to be arrested or promoted. Thus, the analytics framework should be capable of supporting heavy volumes, unstructured data and also work super-fast. 

Challenges

The analytics framework should be able to handle huge data volumes and should also be customizable for different telecom providers. Every telecom provider uses their own data mart and even within a given telecom company, the CDR data format could vary from tower to tower. The data also need to be analyzed almost in real time, so as to identify trends that need to be arrested or promoted. Thus, the analytics framework should be capable of supporting heavy volumes, unstructured data and also work super-fast.

Solution

Congruent developed the following solutions to address the customer needs:
• Data extraction tools
• To extract the call detail records (CDRs) in both ASCII and binary formats
• To extract customer information from telecom company’s data mart

• Advanced Analytics tools to identify revenue distribution patterns and trends

• High-performance data visualization tools to create meaningful visuals of the results
A sample list of analytics developed includes:
• Revenue Bell Curves
• COHORT Deep Dives
• User Vs Revenue variance and
• Revenue Influencer tree

Benefits

Leveraging the analytics framework, the customer was able to
• Significantly reduce the lead time by 80% for analysis and identification of insights – it was
brought down from 30 days to one week
• Identify the top 10% of subscribers who contribute to 50% of revenue and defend that
revenue

Technology Used

SAS VA, Oracle Scripts