NEWS & EVENTS
Data Strategy and The Modernisation Puzzle
Check out the article our CEO wrote for Insurance Day below
Frank Perkins, Chief Executive, Inari
The insurance sector collects, processes and analyses a veritable feast of data, a situation that is becoming increasingly complicated as the use of generative AI accelerates and establishes itself as mainstream.
Huge advances in digitisation mean the insurance industry, historically labelled as a laggard compared with other sectors such as banking, now operates with a plethora of apps and application programming interfaces (APIs).
The growing adoption of innovative technologies we are witnessing is attracting highly skilled tech specialists from outside the sector with deep expertise in solving complex data problems. However, the flipside of this rapid progress is too many companies have joined the tech revolution without a coherent data strategy or robust data governance.
Companies’ experimentation with artificial intelligence (AI) is adding to the complexity. Re/insurers have generally embraced the technology since the launch of ChatGPT little more than a year ago. Recent research by EY in its 2024 Global Insurance Outlook found 52% of insurance chief executives are planning significant investments in AI this year.
However, the huge quantity of data AI can create at speed risks becoming a jumbled mess. Another major problem here is the data that models are working with is likely to be inherently biased in the first place.
This matters because insurers are essentially knowledge suppliers, taking information and creating an opinion using that information. If data is flawed, or if data sets from multiple sources are not integrated, we are heading for trouble. At best, this could impede an insurer’s ability to arrive at a decision effectively; at worst, it could lead to flawed, unethical or even catastrophic decision-making in functions including underwriting, reserving, claims management, product design and business development. Reputational damage would most likely accompany any of these undesirable outcomes.
Right tools
It is certainly not a question of the sector having “too much tech”. Each company must use the right tools to execute their own strategy, so if that involves 150 data suppliers and pieces of technology, so be it. However, problems occur when these solutions are introduced without an overarching plan.
An article by Deloitte outlined many insurers are still struggling to maximise the full value of their data, with research the consultancy conducted among insurance analytics specialists pointing to almost a dozen significant challenges preventing them from doing so. Significantly, this included a “leadership mindset that tends to think of data spending as short-term expenditures for isolated initiatives”, Deloitte said.
That deficit can afflict both small companies and extremely large ones. In some cases, start-ups are well ahead of the game in both data strategy and governance, being digital natives with chief technology officers often a vital part of their founding team. Bigger organisations may lack this type of discipline. However, I have known even insurtechs to be operating without a coherent data strategy.
Companies looking to determine their data strategy need first to step back and ask themselves what their overall business objectives are. Once they are clear about that, they need to understand what information they are gathering and how it is being used – including the shape of their data stack, how external sources of data feed into it and whether parallel systems can be integrated.
Although initially daunting, implementing a data strategy becomes less intimidating if broken down into its component parts, with the steps necessary to achieve the end goal determined at the outset. Data security, the General Data Protection Regulation and any other relevant data protection frameworks must feature in the process, but data governance is just part of a data strategy, not a substitute for it.
Companies may be worried about the expense of this type of bottom-up review and the necessary modifications it will entail. They may also be tempted to rush it through to move on to something more immediately lucrative – but this would be a false economy. A considered approach underpinned by careful planning is essential.
Failure to adapt
A common failing is a company’s data strategy reflects its data acquisition and use perspective as of five or 10 years ago, whereas its situation will have changed radically in the interim. A mid-sized Lloyd’s carrier may have undergone a thorough overhaul and established a data lake and warehousing at the cost of many millions of pounds. But the subsequent introduction of various pieces of low-code software means its data has once again become impossible to track.
The challenges of a given data strategy will change with an individual company’s development and with technological advancements in general, which means it is important for any data strategy to adapt to such changes.
At Lloyd’s, the standardisation entailed in the creation of the Core Data Record under Blueprint Two is a move in the right direction, although elements of the programme inevitably have their detractors. Overall, it will facilitate the development of coherent data strategies by providing more certainty.
However, each company individually must still be clear about where it is going and why. It is also vital senior managers without particular tech expertise are involved in the process and have a clear sightline of how data can help them achieve their overall aims. Ultimately, having the right data strategy – including sound data governance – will separate the re/insurers who thrive from those who will struggle to achieve a competitive edge.
