Predictive Analytics: The New Crystal Ball
By Rousseau Kleuver, BI Engagement Manager, Decision Inc.
Data has taken over the world. It is in every device, business and action. The amount of data produced in the past two years has already exceeded what was created in the entire history of the human race; according to the IDC 1.7MB of new information will be created every second for every human by the year 2020. Aside from these staggering statistics, there is that all-important business angle. A Baseline report found that a 10% increase in data accessibility can result in around $US65.7 million in net income for the average Fortune 100 company. Data done well redefines the meaning of opportunity.
The insurance industry is situated on top of a data gold mine. The nature of the business means that these organisations have access to significant quantities of personal and historic data. Globally, insurers are capitalising on technology to sharpen their market segmentation strategies, reduce claims fraud and strengthen underwriting and risk management.
“We are seeing significant shifts in analytics trends in the insurance industry,” says Rousseau Kluever, Engagement Manager, Decision Inc. “Advanced analytics and general analytics are increasingly integrating with one another as insurance companies become more sophisticated and ask more complex questions. In the past, predictive analytics were something that people would say they ‘should’ get round to doing. Now it has become an urgent trend. The significant competitive advantage of leveraging data has been realised and the race is on as insurance companies understand how best to use available technology.”
Now vast amounts of data can be mined to uncover insight, transform best practice, increase market footprint and change the way an organisation invests into the future. It also needs careful handling and an awareness of its potential. Tapping into predictive analytics isn’t something just any corporate can do, to harness its potential and capability the business must be ready and capable.
“You can’t just dive into predictive analytics and implement it in an organisation,” explains Kluever. “There is a need to understand the maturity of the business and to gain insight into its overall strategy and processes before embarking on any predictive analytics strategy and solution. In order to truly benefit, the organisation has to be able to get hold of the data and understand what needs to be done with it. Otherwise it ends up implementing a solution with no idea on how to use it or what to do with the information it generates.”
That said, if the business is ready, then the opportunities presented by predictive analytics are impossibly endless and the potential extraordinary. Using data science as its tool, the business can take calculated risks using data-driven insights. It can use this information to catch more than just a glimpse into the future. It hands the organisation an invaluable tool which can be used to minimise potential risks and capture unexpected trends.
“Done well, analytics can give the business a feeling for what is going to happen, allowing them to recognise trends and make plans around them,” says Kluever. “Knowledge makes such a difference, especially of future shifts in market and landscape, and is far more powerful than historical data. The latter is a lagging indicator, not a leading indicator and is a picture of what is done and dusted, not what trends and influences could happen tomorrow.”
Predictive analytics gives an insurance organisation a significant edge. The ability to see what’s coming and use this data to correct a course of action or take advantage of an unexpected situation. Traditionally, organisations have had to rely on historical data to make decisions and this has often meant that changes have come too slowly or have not been in line with market or customer.
“When we look at historical data we realise that any actions taken based on old trends should have been done a month ago, not today,” says Kluever. “With this level of predictive technology, the business can look forward at what is going to happen, forecasting trends and finding out who to sell to, and the results can be astonishing.”
“As an example, companies can use classification analysis to create a predictive model to identify customer contacts that are most likely to respond positively to a marketing campaign. By using classification, we can predict which customers are most likely to respond to a phone call from a telesales agent for a specific product. Profiling customers for specific products will decrease marketing costs and increase sales effectiveness and efficiency.”
Predictive analytics has the ability to fundamentally change the business, but it has to be mature enough, have the right internal capabilities and understand exactly how data can be used to drive growth and development from today and into tomorrow. Once those boxes are ticked, however, the potential business insight that can be released is limitless.