The Underwriting Decision at the Right Time: Implied Knowledge
I had my leaking shower replaced recently. Thought you might like to know that. It was an easy and efficient process facilitated by information on the internet, comparison and review sites and online collaboration. Wind back a few decades and this would have been an entirely different experience.
The Yellow Pages and Thomson Local were the Google of their time, that and word of mouth. With a phone in one hand, one that would have had a cord attached to it, and the big yellow book in the other, we would have relied on the wordings in those small post-it sized adverts, whether they had certain language that provided assurances and our gut feel. For a large piece of work, we would arrange a day off work from home to invite tradesmen to inspect the job and provide a quote, that we would then compare and hopefully choose who shall be rewarded with the contract.
Access to information, however sparse, and our ability to assess and decide was a key determining factor to success, and still is.
Within a similar timeframe to the phone with a cord and the yellow and blue books, the insurance industry would have relied on, amongst other sources, printed publications, commercial data partners and information packs from the broker or the insured to assist in the underwriting assessment. The underwriter’s experience and intuition also played, and still does, an important and key role in the decision process.
I was reminded of a point of view recently, by an underwriter, which I found intriguing: We are challenged with facilitating the onboarding process of new talent within the emerging era of intelligent automation and intelligence-based assistance.
The point was intriguing to me as I am in a similar position with attempting to transfer tacit knowledge to bright, energetic and keen to learn staff.
Underwriters 2.0
Insurance is a data rich and data led industry. Printed publications are now available as large data sets that can be ingested for further analysis. Trusted data source providers such as IHS Markit provide online access to analyst researched and sourced data packs. Connected devices, through IoT, provide access to real-time information opening opportunities for new business models such as Telematics. Google maps is relied upon as a trusted data source. The street view capability lets you plonk the orange man down at street level and take a 360 view of the location, as if you were there inspecting the area or property. It won’t be too long before access to remote drones provides a real-time view.
The proliferation of digital data sets, online access to near real-time analytics and bringing the customer and their asset information closer to the underwriter has meant that cataloguing, orchestration, and assessment tools are needed to bring sense and insights from the underlying core and technical data. The need for easy ingestion, sorting and assessment will continue to grow.
Insurance industry analysts are pencilling their view on the future of underwriting with EY stating four roles that future underwriters will often need to act: Sales executive, decision scientist, customer advocate and innovator. We can agree today that the underwriting teams play these roles already and the point being made is that data, automation and innovation, will provide the underwriter the ability to enact these roles more easily and more often.
The Charters Insurance Institute (CII) is looking back at the predictions it made on the underwriter of the future. The key themes of data and technology driving the future still hold true.
Willis Towers Watson is reimagining the role of insurance underwriters and stating a move towards intelligent automation to deliver efficiencies and effectiveness with better risk selection consistent decisions and improving loss ratios.
How do these views of intelligent automation, innovation and transformation impact the underwriters view and need to nurture and grow talent?
Implied knowledge
The role of a junior underwriter and the changes in the onboarding process over time do mean that they are still required to fulfil assistance type activities as they learn the ropes and aspire to become class underwriters. However, there is no shying away from the fact that they are faced with training schedules, soaking up the knowledge that comes from a senior underwriter’s wealth of experience, intuition and sound judgement whilst also now having to make sense of and learn the plethora of tools and data sets that are available to them.
Intelligent automation and intelligence-based assistance do not mean codifying tacit knowledge as in its very nature, tacit knowledge is difficult to write down and share as it’s based on experiences, insights and intuition. The intelligence is in providing contextual and relevant information from the varied data sources and providers that I have talked about, at the right time in the underwriting process to validate the tacit knowledge.
For a new customer and possibly new risk, the underwriter and underwriting team will perform their due diligence and research the customer, insurance risk, gather historical information and review the information that the broker and/or insured has provided. The intelligent automation can assist with pulling information from trusted sources such as Bloomberg, Experian, Companies House and augmenting it with information stored with an organisation’s data repository.
For a new risk, if it is location based such as with property and marine, having access to hazard information, peril scores and cat risks assists with deciding the risk factors that could be used in pricing models, probabilistic models and contractual inclusions and exclusions. Many data providers of hazard and peril information use their arbitrary rating system, which an underwriting team would manually do, so they neatly fit within their risk assessment and pricing models.
Intelligence-based assistance processes could automate the homogeneous process and present the underwriter with a holistic view of the risk factors and peril scores. Non-location based risks, such as casualty insurance would provide risk assessments from a global view with some assessments being relevant at country level such as political, terrorist and incidental risks.
Rating solutions, whether based from the underwriter’s pricing sheet or rating sets in a software solution, use in part the technical information about the asset, amongst other factors such as claim history, benchmarks, hazards and probabilistic models, which could be more readily sourced and fed through automation.
Intelligent automation and intelligence-based assistance are then automation of menial tasks, of connecting apparently disparate information together to bring sense and insights to the underwriting process. Where once this would have been a manual activity to source, digest, analyse and report, this activity can now operate in the background of the underwriting process allowing the underwriters to concentrate fully on the foreground thought, assessment and decision-making process.
My view on implied knowledge and intelligent automation is that they are complementary and should be embraced in the underwriting process to validate experience and intuition. The training and onboarding of underwriting personnel would include the underwriting guidelines and governance and the process of underwriting that provides a carrier with their competitive edge. They would also introduce the intelligent automation, models and process that now occur in the background which supplements the guidelines and governance. Intelligent automation and intelligence-based decision support tools can be incorporated in the underwriting process and used for validation but shouldn’t be a replacement for an underwriter’s tacit knowledge.
Where do we go from here?
As underwriters’ roles move forward toward market-facing relationships and with the rise in the volume of data to analyse, AdvantageGo will soon announce news on how it plans to support underwriters to maintain underwriting discipline and their digital future.
Watch this space!
