Big Data, although more specifically, Big Data Analytics, help solve business problems. These business problems include advanced customer analytics:
- Customer segmentation for targeted marketing
- Root-cause analysis of network problems
- Data correlation for improved health care outcomes
- Customer churn management
- Advanced risk management
These are all problems that can be solved today with traditional data warehouse & business intelligence techniques. But advanced forms of these analyses with additional complex & streaming data sources provide additional business benefit that lift the already improved outcomes and marketing lift. This is the value that Big Data brings to your business.
And this is why I tend to focus on Big Data Analytics and why it is a clearly an extension of business intelligence and data warehousing, not a replacement. Analytics provides root cause, correlation and data discovery that you cannot achieve with KPI-based balanced scorecards on a dashboard.
But, you need to beginning playing with an experimenting with Big Data tools to break through the DW/BI barrier where you are currently boxed in with 10-20% organizational data asset reach and 8-hour ETL windows:
- Hadoop for storing large & complex data files across distributed nodes
- MapReduce to process those files on Hadoop with data locality and divide & conquer
- NoSQL databases like Cassandra & Hbase to write data into clusters quickly, beyond RBMS boundaries
- In-memory analytics for real-time drill-down and data discovery
- Columnar data storage for max compression and analytical capabilities