Underwriting is key
The economic impact of the COVID-19 pandemic is visible in every sector, and insurance is no different.
As economic activity nosedived, property and casualty (P&C) insurance too decelerated. To be resilient and chart a quick path to rebound, ramping up digital transformation initiatives is key. However, it has not been easy for enterprises to leverage digital assets to overcome the COVID-19 crisis.
For P&C insurers, underwriting was one of the critical business functions impacted in 2020. As it influences both profitability and competitiveness, insurers need to re-focus key functions such as underwriting data, risk assessment and decision-making with cloud-managed services offered by cloud service providers such as AWS.
Leading insurers are looking to leverage cloud-native services and artificial intelligence (AI) tools to design an IT strategy and work culture for a future-ready insurance paradigm. They must quickly build customer-centric integrated insurance ecosystems to create a differentiated experience. Customer-centricity often involves using data from various sources and calibrating faster insurance decisions via a rich analytics engine.
Three key aspects of underwriting
Effectively managing the key aspects of P&C underwriting can offer insurers superior outcomes and benefit their end customers.
Underwriting data: Timely and accurate data is fundamental for effective underwriting. P&C insurers face challenges in aggregating data across sources in different formats. Existing P&C systems can’t process unstructured data, which then has to be done manually. Plus, integrating traditional rule-based underwriting systems poses a complex challenge. Curated data sources provide valuable insights to enhance underwriting outcomes, and deriving insights from unstructured data is becoming an emerging industry trend.
Risk assessment: Post COVID-19, virtual interaction has gained popularity for the convenience and safety it offers. The adoption of internet of things (IoT) is trending as smart IoT devices (sensors, cameras, security systems, etc.) give customers more control of their home or vehicle, generate timely insights and offer usage-based incentives. IoT devices enable industries to monitor their equipment and workflows. Digital capabilities can help convert the data gathered for risk management in P&C underwriting.
Decision making: Automated underwriting is a leading P&C insurance trend. Manual underwriting is costly, time-consuming, and prone to cognitive bias, and it could result in losing customers to competition in the long run. However, for underwriting complex or high-value P&C risks, human oversight is required to ensure accuracy. Developing customized workflows for different combinations of human and machine interventions is a challenge that requires significant IT development effort or costly third-party solutions. Underwriting models can be continuously enhanced using insights from unstructured data.
Harnessing cloud solutions to enhance P&C underwriting
Cloud service providers(CSPs) provide futuristic platform capabilities which can enable P&C insurers overcome legacy challenges and capitalize on emerging industry opportunities, with better cost efficiency, flexibility, scalability and low upfront investment.
AI and machine learning (ML): Leading CSPs offer fully managed AI-ML services to build, train and deploy ML models that can be applied to various insurance business functions. ML technologies can help with seamless data aggregation, ensuring data quality and integrity. From complex unstructured data, ML-based natural language processing (NLP) service can identify contextual insights and relationships (keywords, places, specific phrases, and so on). Trained with P&C insurer’s contextual data, underwriting ML models can perform customized risk profiling.
ML-based audio and video processing services can be leveraged for offering virtual contactless risk assessments to customers. Trained ML models can identify underwriting risks (such as a wood-burning fireplace, unsafe windows, and so on) specific to P&C insurer’s requirements.
Automated underwriting ML models can be developed to overcome cognitive bias and provide consistently accurate underwriting outcomes in a short time. Where required, human oversight can also be easily added to an ML workflow. Leading CSPs provide services that can help build training datasets to continuously enhance ML models for underwriting new risks. For example, new construction type, roof type, and so on, can be underwritten by training existing models with images of new types of property.