Insurers need to strategize around client stability
Given current headwinds, insurance carriers need to focus on retaining market share and providing assistance to customers.
Holding onto the best customers is of paramount importance for all businesses, regardless of what products they sell. Insurance companies are constantly trying to retain their best policyholders, typically needing to fend off competitors who attempt to swoop in and entice the customer’s loyalties.
But with COVID-19, new challenges present themselves when it comes to customer retention. With most businesses shuttered and people stuck inside, a number of insurance products have become either unnecessary or underutilized. Take, for example, car insurance. Why should someone who uses their car once a week to drive to the grocery store be obligated to pay the full premium cost? What about restaurant owners? Should they pay for business owners’ insurance (BOP) if they’re going to be closed for two to three months?
Besides insurance under-utilization, there’s also the fact that most households are earning less or have been impacted by unemployment. Stimulus checks will help, but will they be enough for every family in the country? At a certain point, households need to start cutting expenses, and non-essential products, subscriptions and insurance policies will likely be first on the chopping block.
Given these headwinds, insurance companies need to focus on retaining their market share and providing assistance to customers in need. A few notable U.S. insurance companies have already taken the lead, putting customers’ best interests first by providing policy refunds. But now is the time to think about what comes next. The insurance environment is changing, and refunding premiums isn’t a sustainable business strategy. Instead, insurers need to start thinking about short-term and long-term strategies for customer retention. Automation can help.
Short-term strategies
Customer service requests: Nearly all service providers are experiencing massive call center and online request volume. Orderly request routing is paramount to keeping customers happy and maintaining day-to-day operations. Automated routing models can be built immediately to assist with this task, and online and call data from the past month can be used to build models right away.
A simple approach to get started would entail building a binary routing flow between COVID-19 requests and regular insurance inquiries, such as sales, claims handling, and renewals. A more sophisticated approach could involve predicting the customer’s request and determining the best company representative to service each customer.
Claims handling: Depending on the insurance line of business, claim volume is either stressed or near a standstill, but the slowdown is really the calm before the storm. Automating the claim intake and decision-making process is crucial for effectively handling elevated claim activity.
Simple models to get started could be case reserve predictions upon first notice of loss or straight-through acceptance models for minor claims. Implementing these types of models will free up the time of claim handlers, allowing them to focus on more complex claim matters and reduce the number of manual errors.
Long-term strategies
Churn prevention: Certain insurance customers will be hurting more than others over the next couple of months. Taking a proactive approach in reaching out and offering solutions to these policyholders has massive potential to bolster service reputation and stabilize retention. A simple approach to this would involve outreach programs to companies and individuals working in business segments most impacted by COVID-19. These would include restaurants, travel and hospitality, retail, and real estate.
The Bureau of Economic Analysis provides resources for tracking which industries have been most impacted. A more sophisticated approach to pinpoint churn could involve AI modeling, which would include data elements associated with the impact of COVID-19. Developing a plan to get ahead of inevitable cancellation requests will safeguard a company’s valuable customer base.
Product differentiation: COVID-19 has caused massive shifts in the way industries communicate and conduct business. These process changes are very likely to persist into the foreseeable future.
This provides a critical window for insurers to augment their product offerings. Starting with insurance coverages provided today, the risk profile of every insurer’s book has changed since the underlying exposure is suddenly different. Pricing models need to be overhauled to reflect these changes or insurers risk under or overcharging the insureds. Acting quickly to update and retrain existing models will also provide competitive advantages.
Thinking ahead to new insurance offerings, customers will be seeking protection from the new risks popping up in the business environment. Cyber-insurance will gain even more traction as companies have more and more people working from home. Traditional personal insurance policies may need to incorporate additional endorsements, such as internet disruption protection or virus-specific policy language.
Implementation
Uplift modeling: Each COVID-19 decision-making strategy should be evaluated against each other, typically using representative control groups. Measuring the incremental impact for each strategy provides direction as to which ones an insurance company should pursue. A simple way to start could be evaluating business strategies via A/B testing. A more sophisticated approach could incorporate uplift modeling or response modeling.
There are many strategies insurers need to consider in order to survive this pandemic, especially with high expectations of a second wave in the fall. Using AI paired with strategic decision-making to address short-term and long-term areas of opportunity will differentiate the winners and losers from COVID-19.
Nicholas Hamwey is an actuarial data scientist at DataRobot, the insurance technology provider. He has more than six years of insurance and predictive modeling experience. He’s a member of the Casualty Actuarial Society and actively researches machine learning capabilities for insurance pricing and loss reserve modeling. These opinions are his own.
Related: