Transforming Insurance Underwriting with Google Cloud AI— Part 1
Understanding the key building blocks that form the foundation for striking an optimal balance between fully automated, low touch and full touch insurance underwriting.
In the most simplistic of terms, insurance underwriting is the process of assessing an individual or entity’s risk in insuring a property, automobile, driver, health, life or activities like travel. It aims to determine if it would be profitable for an insurer to take a chance on providing coverage to the individual or entity.
But the simplicity of insurance as a critical function of underwriting ends there. An Insurance Insider survey showed some of these industry sentiments on underwriting excellence.
The simple complexity of Underwriting
There is a general uncertainty on how much of an underwriter’s workload could be automated. About 29% of respondents in the survey said that while some, if not all, parts of the workload could be automated, negotiation, relationship management, and adaptive decisions will still need a human touch.
Impact of underwriting tools towards improving underwriting quality has been indifferent to low. Over 36% of respondents declared that the impact towards underwriting excellence has been neutral, while 24% replied the impact has been low, 9% said very low.
Data is abundant, but vast swathes of it is still unstructured and a challenge to organize. Around two-thirds of respondents said the market had enough relevant data at its disposal to produce actionable insights to help underwriters, but much of that data is unusable.
Another common theme I hear is that insurance companies, as well as several technology vendors, have developed automated underwriting platforms in which the underwriting manual is embedded as automated rules. However, these platforms have only superficially helped insurers build a truly end-to-end, automated process. One insurance executive commented “The data is out there, just not easily accessible for the underwriters to evaluate. It’s like wrestling an octopus.”
The building blocks of achieving underwriting excellence
Before we start discussing how Google Cloud’s Artificial Intelligence and Data Analytics platforms can accelerate an insurer’s path to underwriting excellence, it’s important to understand the key building blocks that we should seek to address in the overall underwriting journey.
- Portfolio Steering: Forward-looking analysis to understand how risk exposures and coverage needs change, assess portfolio plan deviations to profitably underwrite risk.
- Profitable Pricing: Elasticity modeling and performing price optimization by analyzing movement of pricing within a portfolio of common exposures over time.
- Risk Selection: Combine data and geolocation driven analytics with human judgement for consistency and accuracy in risk selection.
- Capacity Optimization: Optimize underwriting capacity to generate premiums that exceed losses and expenses, increasing the policyholder surplus and the capacity to issue more policies.
- Rationalized Coverage: Implement controls around the breadth of coverage offered, translating qualitative wording into quantitative parameters.
- Submission to Quote Efficiency: Analytics-driven workflow to enable underwriters to offer a competitive quote to customers all the while keeping an eye on the prevailing market dynamics.
In this introductory post, we discussed the challenges insurers and underwriters face as they attempt to consume enormous data and gain critical insights from it. We also covered the key building blocks of underwriting that we will use in subsequent posts as a foundation over which Google’s AI and analytics solutions would help tap the untappable data.
In the next post, we’ll discuss deeper the limitations of each of these building blocks that impact effective underwriting and how we can solve it with Google AI and Analytics. (Update: Read the next post of the series here.)