By Dr Corine Van Erkom Schurink,
While predictive analytics is a hot topic, many local businesses are hesitant to embrace it. However, given the significance of providing valuable insight on the current and future performance of a company, implementation has to happen sooner rather than later.
Some decision-makers are concerned about the complexity of integrating with business functions that are seemingly quite disparate – think marketing and IT for example. However, once the value of predictive analytics is understood, as well as the associated processes and data requirements, then the choice becomes clear.
While it makes sense to utilise predictive analytics in sectors like financial services and insurance, it really is something that can benefit any organisation in this age of connectedness and Big Data.
While predictive analytics requires deep data, insurers today, who often work with silos of information, need to recognise the fact that a good analytics dataset can typically consist of a mix of integrated data (and not separate information) – being customer data, historical policies or product data, and intermediary/agent performance records. Through recognising this, they will be able to reap the benefits predictive analytics can bring to an insurance business.
For insurers, predictive analytics can aid in ‘flagging’ customers who are most likely to commit fraud at an early stage of their life cycle, for example. It can also help predict the performance of a company representative or selling agent and their likelihood to ‘drop out’. Just think about the cost-savings that could be achieved.
Similarly, in the case of predictive fraud, actuaries and underwriters can now quickly reduce the risk exposure of the company by adjusting the rules supporting decisions and the algorithms that determine new policy premiums or claim settlements – which will reduce expenses even further at individual customer levels.
So why are South African companies so slow in embracing this technology?
For one, local universities have been slow to adopt a curriculum that places data scientists in the market. Another reason is that companies fail to understand what is needed to effectively implement predictive analytics. Far too often it becomes one of the ‘functions’ of the IT department, which has no real sense of the business directives of the company and of the strategic business implementation of the advanced analytics outputs. There has to be inter-functional consultation, shared resources, and shared budgets.
Cynics might argue that if all insurers implement predictive analytics there will be no real differentiation. However, the ones that are able to go back to their ‘roots’ and use traditional marketing tools such as pricing, quality of service, innovative products, and effective service, will be the companies that are able to supercharge their offerings. The level of effectiveness in how predictive analytics is leveraged and actioned across the organisation will also contribute to the gain of competitive advantage.
Insurers adopting predictive analytics will not only reduce their risks and related costs, creating larger profits for reinvestment, but also free a large amount of resources that can be redirected to new projects. Those who adopt predictive analytics will become slicker and less ‘cash strapped’ than the insurance companies who will have missed the boat.
So, for those who are still undecided about the merits of embracing predictive analytics, think about the lost revenue opportunities and whether that can be written off in such a competitive economy.