Would you rather have Data, Knowledge or Intelligence?
Automation has enabled companies to collect massive amounts of data across the value chain, but they often don’t have either the disciple or the capability to systematically transform the raw data into information, information into knowledge and knowledge into actions that produce tangible results. Data and information are primitive forms of knowledge. For e.g a customer disconnecting their service is piece of Data, which is not as interesting as the Information that a customer disconnected service because he was dis-satisfied. The Knowledge that customer had a service outage and had called Customer Service regarding a billing issue several times before dropping the service is much more meaningful. These nuggets of data and information is stored across silo-ed systems, with no real time mechanism to pull together the full story. Without a systematic framework for collecting and synthesizing these bread crumbs of data, and asking the right questions, such insights would be lost in the sea of data. Yet acting on these insights might prove critical in reducing churn and to maintain customer satisfaction.
Analytics involves coming up with critical hypotheses or questions that need to be answered and then mining the operational data to discover patterns, anomalies and common themes that provide insights to enhance business decisions. Data warehousing, business intelligence and OLAP technologies have made it easier to process, cleanse, filter, summarize and de-normalize data. Often the harder task is creating a culture and discipline that encourages fact based decisions, and minimize “gut based” decisions. This means that the Analytics infrastructure must be fast, reliable, accurate and comprehensive for the subject area under study.
Coming up with the right hypotheses or questions to ask is part Art part Science. A company’s strategic objectives often prompt and define the hypotheses. In the above stated example, the relevant questions might be, which customers are in danger of churning out? What can be done to keep those customers? Are we better of giving the bad customers to the competition?
Predictive Analytics is fast becoming a tool that is being used to fundamentally change the how decisions are made and the quality of those decisions directly impacts business and operational performance.