Insurers need to be more proactive in using data to innovate themselves before they are overtaken by external forces, said speakers at the Asia Conference on Big Data and Analytics for Insurance. But how should they do it? Where should they begin? We bring you highlights of the strategic tips offered and the pitfalls to avoid.
The way insurance is sold and engaged is ripe for disruption, evident in the trends such as the emergence of online marketplaces and new ecosystems; the increase in adoption of internet and mobile-based channels; and the ubiquity of connected devices, among others.
Hence, insurers must be prepared to ride the digital wave, said Ms Sophia Van, Vice President of Swiss Re’s Big Data & Smart Analytics Centre. She noted that startups are invading insurance tech, with a threefold jump in funding to this sector between 2014 and 2015, and multiple tech giants such as Alibaba, Baidu and Tencent moving into insurance.
“If you don’t revolutionise from within, external companies (from other industries) will help you disrupt,” said Ms Na Jia, Managing Director of ReMark Asia and CMO of ReMark International, responding to a question on the likes of Google stepping into the insurance space.
“The clock is ticking,” she added, noting the “loads of innovation that are going on daily” in markets like China, which recently saw the formation of online property insurer Zhong An.
Many lack good grasp of smart analytics
So far however, many insurers have not got a good grasp of smart analytics, which is the key to harnessing Big Data, said SAS’ Global Insurance Practice Director Stuart Rose. “Insurance companies are drowning in data but starving for information,” he said. “Big Data is useless unless you can add value to it.”
He added that data analytics is not new to the insurance industry; the game-changer rather, are the new techniques out there. “It can be argued that insurance companies have been using analytics for decades and centuries. Mortality tables are a form of analytics as insurers took historical data to predict the mortality rates. What the game-changer is, are the new data and analytical techniques that are available to insurers to better assess and price risks,” he said.
There may not be a clear-cut or best method to turn data into insight since the science of extracting insight from data is constantly evolving. But regardless of the amount of data one has, one of the best ways to understand and sieve out information that is important – and what is not – is through “exploratory data analysis”, said Mr Rose.
“Begin with the end in mind”
Implementing a successful data strategy is both an art and a science. Mr David Hardoon, Executive Director at Azendian, said that to do so, companies need to begin with the end goal in mind. One of the most crucial elements of a data strategy is its alignment with business objectives – where it is going to go. “Start thinking of the final outcome and understand where it will be of value to you. With this, it will impact how you go about deciding the strategy.”
A data strategy also needs to be measurable and be part of the KPI, otherwise “it’s not going to happen,” he added.
Notwithstanding, Mr Hardoon also shared a few overall considerations: a data analytics strategy is neither about accuracy nor research, rather, it is about transformation; insurers will need to assess the relevance of the strategy, whether analytics is “really needed”; and third, it is not only important to ascertain the usability of the “outcome, relationships and predictions” but also to differentiate between what is actionable and measureable – and what is not.
Ownership and management
Another key issue that invariably came up during the two-day discussion was the ownership and management of data. Throughout the course of discussions, it was evident that data has evolved from what was originally considered an IT function and responsibility to an asset that must be owned and driven by businesses and its top management.
Ms Catherine Khaw, Chief of Intelligent System Practice at the Institute of Systems Science, National University of Singapore (NUS), said that data is an asset that requires a strategy and execution plan at the enterprise level, and how to manage the asset will depend on its owner. IT, she added, facilitates its management.
“Companies need to take ownership, step up and manage their data. Invest in a proper management system or it will come back to haunt you when you most need the information.”
She outlined the pyramid that makes up an optimal enterprise data management framework: data stewardship (foundation) made up of robust architecture and delivered according to strategy; data management (mid-tier) where data management vision and strategy are executed alongside and alignment and communication plan; and data governance (apex) where data is sponsored or owned by the business and data policies and guidelines are set and enforced.
With data comes cyber risks
With the growing pursuit of data analytics and digitisation today, Mr David Piesse, Chief Risk Officer, Guardtime, noted that cybersecurity and the integrity of data were key issues that must not be overlooked.
Cybersecurity, he said is an equal opportunity risk that does not respect borders. While “data in motion” is covered by public key infrastructure (PKI) protocol, more data in motion is “moving to ‘data at rest’” and current security models are insufficient in ensuring data integrity or prevent breaches.
As such, there is a need for mathematical certainty, an independent audit trail for all human and machine activity in digital society to mitigate such risks, he said. He advocated the use of Keyless Signature Infrastructure (KSI) blockchains, which can “enable the properties of data to be verified without the need for trusted third parties, keys or credentials that can be compromised”.
“Insurers have vulnerabilities too – sign crown jewel digital assets before it is too late,” he said.
The conference, which was organised by Asia Insurance Review and sponsored by ReMark, was held in Singapore and attracted about 120 delegates.