The impact of the COVID-19 outbreak has been devastating for the global economy, and with many countries in extended lockdown, there has been a seismic shift in customer behaviour and business operations.

Insurance is likely to see significant changes to business models in the coming months, with significant impact on revenues across the industry.1

The current modus operandi is likely to undergo significant changes as insurers find themselves focusing on these three areas:

Projected COVID’19 impact index on Insurance

  • Sourcing new data and projecting recovery: Tracking and using new external/internal data sources to generate insights into behavioral and financial changes of customers will be a key priority for insurers.
  • Customer Segmentation: Categorisation of population into separate clusters based on parameters such as occupation, age group, employment and financial status, and devising relevant marketing strategies for each segment.
  • Pricing Strategy: Devising the right pricing strategies for each segment will be critical for success and we will see a shift from traditional pricing, to usage-based or discounted pricing for some segments.

This white paper focuses on the impact of COVID-19 on motor insurance premiums, analysing how insurers may need to evolve business models to continue achieving growth.

Impact on Motor Insurance Premiums

As of June 4, 2020, there have been roughly 6.6 million confirmed cases of COVID-19 globally, resulting in 388,000 deaths. The United States, United Kingdom, and Spain have faced the brunt of the epidemic, with cases in the US and UK numbering 1.9 million and 279,000, respectively.2

To study the impact of COVID-19 on motor Insurance, we can draw parallel trends to other epidemics like SARS.

COVID-19 will have multiple impacts on the revenue coming from motor insurance.

Impact 1: Decrease in premium volume: COVID-19 has caused major disruption to global travel and business activities. Coupled with government or selfimposed lockdown and social distancing norms, this has led to a significant drop in vehicle usage. The decrease in automobile usage has led people to re-consider the perceived value of motor insurance and thus people are demanding a reduction or discount in the insurance premiums they paid previously. Insurance companies must now determine the optimal discounted rates to retain customers while maintaining profitability.

Impact 2: Delay in payments: Global regulators have been urging insurance providers to accept delayed payments without a penalty. This will put a major strain on their cash flow.

Impact 3: Change in valuation and loss recognition systems: Market volatility has led to an unclear stance on the valuation of certain insurance products. Due to economic uncertainty, present assumptions on loss recognition assumptions may be insufficient. Insurance companies may now have to revisit their outlook on the various parameters and KPIs they have been using in the past for loss regulation and premium calculations.

Major factors contributing to decline in revenue:

Factor 1: Sharp decline in automobile sales leading to a fall in number of new motor insurance policies: Insurance companies will see a dramatic change in customer behaviour. With partial or full loss of income, people will be more hesitant to spend, leading to a sharp decline in automobile sales. This will negatively impact the number of new insurance contracts.

Factor 2: Sharp decline in use of automobiles in lockdown: With the world in lockdown and most offices adopting work from home policies, people are going out only for essentials. This has drastically reduced travel and automobile usage

Factor 3: Relaxation period provided by insurance regulators for the payment of premium installments, leading to more defaults in payment: Numerous governments have requested that insurance companies provide discounts or grace periods and lift charges for repayments. This has in turn reduced the amount of premiums received for motor Insurance.

Global Automotive Sales, M units

How can Insurers acquire and retain the right customers in a post COVID-19 world?

To tackle the economic impact of the pandemic, insurers need to adopt a robust framework that optimises the customer experience, resulting in higher retention while maintaining profitability. Some key objectives are:

  • Defending and growing market share by retaining and acquiring customers
  • Offering superior customer service
  • Improving loss ratios
  • Lowering cost of acquisition and servicing

Insurers can meet these objectives by following a four-step process driven by effective use of data and analytics across the value chain of the organisation.

A sample framework: Addressing customer acquisition and retention challenges by effective use of data and analytics


It is critical to bring in the right data sources (both internal and external) so that insurers can make timely and strategic decisions based on key recovery indicators. COVID-19-specific insights can be extracted using internal data points like telematics data and driving behaviour, customer interactions, and MTAs during the pandemic, and by categorising customers into sub-groups like essential/non-essential workers. External data sources such as mobility data, macroeconomic indicators, consumer behaviour data, business activity, and recovery trends can be used to enrich the existing models for better business insights.

  • Mobility Data: As movement of people is highlycorrelated with vehicle usage, mobility data can be a leading indicator of economic and business activity recovery. Google Mobility data (UK) clearly shows how the movement of people has shifted from the workplace to the home in line with the lockdown measures. As lockdown measures are gradually eased, the trend is expected to reverse and we should be able to judge the path to recovery based on the trajectory of the mobility data.
  • Macroeconomic Indicators: As per the trends seen for the UK, macroeconomic indicators like GDP, Employment Rate, Household Disposable Income and Import are highly correlated to the Motor Insurance Gross Written Premium (GWP). Since forecasts for these major macroeconomic indicators are widely tracked, insurers are able to use them to predict how Motor Insurance premiums may behave over the years. Employment rate has the highest correlation with Motor Insurance GWP, but it is a lagging indicator. Household Disposable Income mimics the GWP numbers very closely.

Google Mobility

Apple Mobility

Macroeconomic Indicators Vs GWP

  • Business Activities: Analysis of the business activity data in the UK suggests that over the years Motor Insurance GWP has seen a strong direct correlation with the indicators like Business Confidence Index, Import and Export numbers. Based on how businesses perceive the economic threat of the COVID-19 crisis, there will be a direct impact on the volume of export and import, which will in turn impact the motor insurance business as well. The Business Confidence Index has the highest correlation with Motor Insurance GWP and its behaviour over the years has closely resembled GWP trends.
  • Consumer Behaviour: Customer behaviour and sentiment can be tracked using indexes like Consumer Confidence Index (CCI) and Consumer Price Index (CPI), which show the degree of optimism consumers feel surrounding the country’s economy.
    Analysing consumer behaviour and sentiment can help insurers price accordingly and improve loss recognition models. In particular, CCI is a good indicator to feed into pricing models as it has a strong correlation with GWP numbers and closely resembles consumer’s behavioural changes due to the pandemic.
  • Other External Indicators: A lower employment rate will have a direct impact on the risk appetite and buying behaviour of customers. A reversal of trends, with household savings going up and spending going down, may also be expected Household savings have a strong negative correlation with Motor Insurance GWP.

Vehicle sales have also nosedived, and this will have a huge impact on the number of automobile registrations in the short term, and consequently on new business premiums received by insurers.

word of caution

Business Activities Indicators Vs GWP

Consumer Behaviour Indicators Vs GWP

Other External Indicators Vs GWP

  • Unstructured Data from Internal Sources: Better use of unstructured data such as call and chat data from customer interactions could provide key insights into changes in consumer behaviour during COVID-19. Requests for MTAs regarding change in mileage, removal of additional drivers, and reduction of add-on coverages are all indicators of changed circumstances. Unstructured data can be converted using speech and text analytics and may be used for further analysis and modelling activities. Use of NLP and Machine Learning in speech analytics can result in transcriptions with ahigh degree of accuracy, helping to generate key insight.

Key Consideration

With the economy trending much worse than the financial crisis of 2008, bringing these indicators in the preview of analysis can add more depth to all the analytical models.


In the new world order, it is critical that insurers identify the right customers for their products to match the risk appetite and offer the best combination of channel + proposition + risk coverage + customer experience. Insurers should be looking at their existing and prospective customer base and re-evaluating their target segments. It’s essential to evaluate the needs and preferences of customers at each stage of their journey, and correlate these across marketing, sales & service, underwriting, policy administration and claims.

Customer Segmentation: The fundamentals of segmentation for motor insurance customers have not changed, but several parameters have now become key to the segmentation process in a post COVID-19 world.

These key parameters will be governed by requirements for travel and vehicle usage, relative risk of infection and financial strength of certain groups of customers.

Some of the key indicators for the new segmentation strategy will be:

1. Employment Status & Type14: With a large number of people furloughed or out of jobs, employment status and type has become a key categorical variable for segmentation. The changing employment landscape in the UK is captured below:

The unemployment rate for Jan-Mar 2020 is 3.9%, but it is expected to go much higher (close to 10%) as more and more companies are facing the brunt of COVID-19.

2. Occupation: A distinction between essential versus non-essential workers has become very critical with respect to future motor insurance requirements.

  • Increased Mobility – Frontline Workers: Frontline workers, including medical professionals, care home workers, retail and delivery service professionals, emergency service, postal, transport and utility workers have continued to work through the pandemic and have higher driving requirements
  • Quick Return to Normal – Outdoor Activities: Workers in construction, manufacturing, real estate, public administration, wholesale trade, education and training have started returning to work as lockdowns are being eased, with their driving requirements also returning to normal
  • Slow Recovery – Hospitality & Travel: Hospitality, restaurant, travel industry workers will probably see the most downtime from their jobs, and their driving requirements will decline in the short term
  • Work From Home – IT & Professional Services: Some professionals in sectors like Information Technology, Financial Services, Insurance, Administrative and Support Services, Scientific and Technical advisory may see a complete elimination of driving requirements as they continue to work from home in the medium and long term

3. Age: Since the impact of the virus is quite disproportionate based on different age groups, certain age groups (mostly above 50) are expected to stay in isolation longer and hence will have significantly reduced demand for motor insurance, which will also impact claim scenarios

Unemployment Rate

Customer Segmentation by Occupation

Customer Segmentation

4. Financial Stress Score: A combined score created for the financial stress of individuals can be a key indicator for segmentation. It takes into account different parameters such as affluence in the area of home address, primary source of income, job profile, stability of the employment sector, previous payment history (including any deferred payments during COVID19 crisis.) This score can also help determine the propensity to buy add-on coverages and premium products.

  • Pricing Strategy: These new segmenting attributes could drive different pricing strategies for different groups of customers based on their incremental or reduced driving needs, propensity to buy different coverages and their unique risk factors. It is critical to align these segments to the right pricing strategy. A three-tier pricing strategy will align with most customer needs.

1. Usage Based Pricing: A low fixed premium with a variable usage based premium calculated every cycle

2. Traditional Pricing: The standard pricing strategy with a fixed premium for standard coverage allocated over the policy period

3. Discounted Pricing: A fixed premium pricing with a lower premium providing only basic coverage

Financial Stress Score


  • Price Sensitivity Analysis: Once a new customer segment is created, it is important to perform price sensitivity analysis for each segment to position different products and coverages appropriately surrounding that sensitivity. For example, a segment that is made up of furloughed employees in the travel industry would be extremely price sensitive and might need motor insurance only to comply with the regulations, with insurance premiums at the minimum. This customer segment could be targeted with a no-frills motor insurance product with basic coverage, at a highly competitive price point. To perform the price sensitivity analysis, historical data regrouped into new segments with additional new data points captured through aggregators and direct channels could be used.


  • Recovery Projection: In the past, motor insurance premiums have shown strong co-relation with a number of macroeconomic indicators, business activities and customer indicators. Today, insurers must track these indicators in real time to project what the future holds in their road to recovery. The data can be used for projecting premiums, recovery trends, and market sentiments. For example, it has been seen that employment rates have direct a correlation to Household Spending and an indirect correlation to Household Savings - which further translates to auto sales and impacts mew motor insurance contracts.

COVID-19 Impact & Recovery

  • CLTV Modeling: Due to changes in risk factors and the financial appetite, there will be a significant shift in the Life Time Value of different customer segments. This will require an overlay of different behavioural and financial trends over existing CLTV models in order to make them more robust and dynamic so that they are able to capture COVID-19 related uncertainties. To increase CLTV, there will also be a shift towards a bundle approach rather than singular products. The bundle approach will streamline the process for customers as they can buy y different policies from a single insurer.
  • Response and Conversion Modeling: Customers from different segments will be impacted differently by COVID-19, and their risk appetites and financial outlooks will change. This shift will have a major impact on Response and Conversion models that have been built on historical data and which are becoming obsolete. For example, the motor insurance buying behaviour of IT professionals who are more likely to work from home will be very different from those professionals working in the manufacturing sector, as they will have to return to work as soon as the lockdown ends. IT professionals would more likely respond to a “Usage Based” model rather than a constant premium, whereas those working in manufacturing may find a fixed premium insurance policy more cost effective. Insurers will need to assess the impact of the new normal and incorporate new variables and data modelling. This will see inclusion of real time data, application of telematics and use of advanced Machine Learning techniques like GANS.


In the long term, the behavioural and operational changes associated with the pandemic will impact not only the way fundamental risk assessment is done for technical pricing, but also the way retail price and premiums are optimised by insurers. The size of the motor insurance market is going to contract and insurers will be looking to defend and grow their market share. The industry will need to constantly innovate and come up with novel targeted offerings that can help acquire and retain the most profitable set of customers.

  • CLTV Modeling: Due to changes in risk factors and the financial appetite, there will be a significant shift in the Life Time Value of different customer segments. This will require an overlay of different behavioural and financial trends over existing CLTV models in order to make them more robust and dynamic so that they are able to capture COVID-19 related uncertainties. To increase CLTV, there will also be a shift towards a bundle approach rather than singular products. The bundle approach will streamline the process for customers as they can buy y different policies from a single insurer.
  • Pricing Model Refresh: Both the frequency and severity of claims are expected to go down in the next 6-8 months, which means that a large number of loss cost models will require refreshing in order to incorporate the new scenarios.

What are some of the key considerations for insurance firms in undergoing these model refresh exercises in future?

1. New Age Data: Personal line insurers need to incorporate newer data points emerging from greater use of technology, including telematics and IOT devices.

2. Machine Learning: While Machine Learning and AI have made significant progress in areas such as claims and marketing, adoption has been quite low in pricing due to challenges around definition and reinforcement of biases in Machine Learning models. This needs to change quickly, as insurers have to extract more value from better pricing, and traditional GLM models can go only so far.

3. Rapid Experimentation: Pricing teams will need to adopt a rapid experimentation approach for getting results faster, while also experimenting with different methodologies in an agile manner.

4. Frequency of Refresh: Typically core pricing models are not refreshed frequently and only marginal changes are done, but in a rapidly changing insurance landscape, it is imperative to reduce the frequency between model refreshes.

5. Deployment Strategies: For Machine Learning based pricing model deployment, it is critical to set up a flexible and scalable technology architecture that can ingest data and provide outputs in near real time in order to satisfy the complex requirements of personal line insurers.


  • Discount / Offer Calculations: We have already seen some of the impacts of COVID-19 on motor insurance premiums, refunds and discounting. What is clear is how insurance firms treat their customers now is likely to create a long- lasting perception in the market. Insurers that understand the plight of the people and implement fair and balanced reforms in their pricing and premium practices will be rewarded in the long run. However, there are multiple methods of calculating and offering the discounts to the end customer. Each method has its own pros and cons and it is important to evaluate them before implementation. Some of the common methodologies are:
    • Flat Discounts: Estimate the overall indemnity and cost savings from the drop in claims and then distribute part of the savings uniformly among all customers (for example, a fixed discount value like £50.)
    • Flat Personalised Discounts: Estimate the indemnity and cost savings from the drop in claims based on frequency/severity models and pass on a part of the savings to each individual customer (i.e. a discount value £40, £50 or another sum offered to each customer or customer groups based on profile).
    • Premium Proportionate Discounts: Distribute the claim savings amongst the customers as a percentage of their premiums without any cap (i.e. a fixed percentage value, such as 20% of the premium.)
    • Premium & Claim Proportionate Discounts: Distribute the claim savings as a percentage of the premiums but personalise for each customer based on their claim history and propensity to claim (for example, a percentage value such as 10%, 15% or 20% for each customer).

Discount / Offer Calculations

  • Additional Value Delivery: Insurers can take other proactive steps to create additional value for their customers, such as utilising any spare capacity in their sales, service and claim contact centres to keep in touch with tenured and elderly customers in these difficult times to enquire about their well-being. Making the necessary changes in customer journeys to make them more streamlined and digital is also an additional value add.


COVID-19 is unlike any challenge this generation has faced in the recent past, and the economic impact has been unprecedented. Like all sectors, insurance recognises the need to be agile and to use new data to project recovery so that it can understand the behavioural and financial changes customers are experiencing. Devising the best pricing strategies for customers will be critical for success particularly in regards to motor insurance premiums, and insurers will have to apply sharply focused customer segmentation in order to develop relevant motor insurance marketing strategies that resonate on an individual level.

The transformation journey analysed in this paper covers the pre to post COVID-19 era, emphasising the fundamental changes insurers will have to make in order to acquire and retain the right set of customers. Through the use of applied technology and analytics, motor insurance specialists will be able to track key economic indicators and project recovery, providing an excellent opportunity to optimise the customer experience while boosting retention and profitability amongst new and existing customers.


















Swarnava Ghosh
Senior Engagement Manager – Analytics

Assistant Project Manager – Analytics

Contact US