Data Analytics: Improve customer service and gain sustainable growth
Dutch Insurance Outlook 2017
Data is at the heart of an insurer, since the servicing of clients and risk management is based on data. Deloitte believes it is essential to strengthen an insurer’s data and analytics capabilities, since it can improve an insurer’s services, underwriting, sales, risk profiling and personalisation of offers. In this article we will specifically focus on the possibilities of using data analytics in client services, and provide three recommendations that help insurers know where to start when building their data analytics capabilities.
Data analytics in the insurance marketplace
In today’s rapidly changing world we see that innovative, technology-oriented companies grow faster than their competitors because they collect, analyse, and use data in an intelligent way. Insurers have collected a lot of data over time but the analysis and usage of that data has been relatively limited. However, insurers seem to have reached a turning point: data and data analytics can be the hidden power of insurers, opening up endless opportunities.
Current computing power and technological developments have increases the focus on the value of data analytics over a wide spectrum of functions, including: improving the customer experience, enhancing risk assessment in underwriting, reducing the cost of claims, and identifying new sources of sustainable growth. But a recent Deloitte study1 on data analytics among 68 EMEA insurers showed that they are still struggling to get value out of data analytics, even though they have historically managed, gathered, stored, and interpreted data across various different departments.
The key observations from the market study are:
• The analytics strategy is not in line with the business strategy, making it difficult to obtain and monitor the value from analytics.
• The focus is on short-term tactical initiatives as opposed to long-term strategically aligned projects, stating that insurance companies might not be focusing on the biggest opportunities.
• That analytics is not embedded into strategic decision-making and (senior) management focus, making it difficult to change the decision-making culture.
• The added valued of analytics cannot be identified or linked to the investment, leading to a discussion about the added value.
• The focus is on building an in-house capability, making insurers at risk of comprising on the agility and quality of analytics solutions.
Do these struggles mean insurers should stop investing in data analytics? On the contrary. Since 2011, insurance tech companies have raised large amounts of money. These organisations often use data analytics to cost-effectively serve tailored products to customers. We also see that incumbents are interested in adapting to new technologies, such as RPA, and making data analytics a core capability. Therefore, we think now is the time to take an extra step forward and put the spotlight on a data analytics strategy in order to defend market share in the (near) future and find opportunities for sustainable growth.
Three data analytics applications in the customer service domain
We believe, and have seen, that data analytics can radically improve the service to customers and provide sustainable growth for stakeholders. We will elaborate on what we believe are three key data analytics applications that insurers can invest in to improve their products and services for customers and clients. They can also help to provide better insights and define best actions in order to retain and grow a sustainable and well-balanced portfolio:
1. The insurer perspective: Data-fuelled cross- and upsell approach
2. Service from insurer to client: Optimised underwriting process
3. The client perspective: Services and products
1. Data-fuelled cross- and upsell approach
The insurer needs to have a solid client base in which the clients’ risks are pooled and cross-subsidising effects create a well-balanced and overall profitable portfolio. This means that specific segments of the portfolio might be loss giving, which is then offset by other profitable segments (subsidisation). To manage the balance between revenue, costs to service, and the risks of the portfolio, it is really necessary to understand, recognise and manage the current client portfolio. As opposed to dynamic pricing (see section on Dynamic Pricing in this Outlook), which is more pulling driven by client value, client behaviour and market movements, we see data analytics also as a very good tool to identify interesting segments and target groups across the full spectrum of the (potential) portfolio (pushing).
When offering clients products, or additional products and covers, it is very important to understand which offers are relevant for the customer:
• To attract new customers, Closed Loop Marketing (CLM) techniques help the insurer to create the best customer journey. Analytics are a major component of the work within CLM. For example to present the customer with the most appealing website to his or her taste, based on the data that is known about this customer.
• After a customer has become a client, the next step is to determine the right actions (‘Next Best Action’). Determining the client’s value over a portfolio range, as opposed to just looking at the product profitability of this client, can give some fresh insights. A client might seem to create a high risk appetite for one product or even one cover, whereas the client might feel this way for other products or covers. Based on this information, the insurer can decide whether the investment of offering additional products and/or covers is interesting for both clients and insurer. This approach will need sufficient past examples of successful and unsuccessful offers to be able to make good decisions. Hence, (starting) A/B testing often gives very useful information to include in the analysis. Pricing of the products, specific discounts, and product terms and conditions also influence the risk appetite of the clients.
Implementing a data-fuelled approach for cross-and upsell will allow more relevant offers to be provided to customers, improving customer satisfaction and making a difference in the portfolio. Furthermore, feeding the predictive models with new training data every year brings a wealth of opportunities and insights to the insurer. This is exactly what can be done and what we call ‘the industrialisation of an analytical discovery’.
2. Optimised underwriting process
At the underwriting department, the output of workstreams of the risk pricing, dynamic pricing, marketing, sales and product management departments come together. The main question is: can we accept this client for this product and cover, given the risk profile, for the fair price, given the terms and conditions, and given the overall portfolio and strategy? The other workstreams already present a lot of data and analyses around this data. Combining this into a (preferably) automated and optimised process, requires the insurer to use a fair deal of data analytics as well. Using this in a good way, unlocks four areas of improvement:
• Better decision-making on risk-acceptance: If the insurer is able to collect more and better data about client risk, it can use data analytics to identify the high risks. This makes it better-positioned to decide on whether to take or mitigate this risk, or even not to accept it.
• Better connection between underwriting and the distribution strategy: Creating a data-driven underwriting platform gives the insurer the opportunity to better connect to all distribution channels (e.g. underwriting agents) and be more in control of the total risk acceptance and acclimatise the distribution strategy accordingly.
• Automated underwriting process: Using data analytics in a digital environment in the underwriting process gives the insurer additional opportunities for automating the process. This will bring cost reductions without compromising on quality.
• Preparing the underwriting process for future innovative product development: New products and services can be based on new data and technological possibilities. These developments will also have an impact on underwriting. Insights from data analytics help the insurer to set up the right processes and tooling to cope with this new environment.
Optimising the underwriting process will therefore not only improve profitability through better pricing and risk acceptance, but will also improve the efficiency of the applications process and increase customer satisfaction (i.e. a better Net Promoter Score). A much shorter and more efficient application process will be a big benefit, especially for agents and intermediaries, and in the future it may even be a requirement. But also in commercial lines insurance, optimising the underwriting process can be very valuable to insurers. This will provide vast opportunities for growing a sustainable customer base.
3. Services and products
A third application area in which data analytics can drive sustainable growth is in adapting services and products to customer needs. Customer behaviour is changing, and being able to identify new needs can clearly mean finding new ways to improve customer satisfaction and creating competitive advantage. The possibilities are along four paths: usage-based products, prevention, peer-to-peer insurance, and tailored services and policy conditions.
One of the reasons for an immense growth in data, besides the increase in unstructured data on social media and internet, is the rise of sensors and their data. Sensors can measure pressure, position and motion, vibration, temperature, humidity, chemical concentrations, radiation and many other things. Given the immense potential benefits, huge investments are made in further sensor improvements, which is likely to result in a growth in the number of sensor-equipped internet of things Units.
The rising use of sensors, combined with connectedness (‘Internet of Things’), internet capacity, and cloud computing speed have made it possible to unlock the potential of sensor data to apply usage-based-products2. Based on the data collected, an analytics platform can help to correlate aspects of e.g. driving information with accident likelihood and claim size. Intelligent scoring algorithms can then be used to calculate and adjust the premiums based on costs to service and actual risks, driving sustainable growth. This also enables a close feedback loop to customer behaviour and new value-add services, such as trip reports and driver behaviour and patterns.
Another example of the application of sensor data is the quantified self; also described as self-knowledge through self-tracking with technology. Individuals can quantify biometrics that they never knew existed. In addition to improving risk-based pricing, the rise of quantified self-application provides insurers with the opportunity to fulfil a relevant role: prevention. It is in the interests of the insurer, as well as the consumer (and society), to prevent damage, disability, unemployment, or death. With the increasing and staggering amount of data being collected, insurers could find themselves to be more in the business of prevention rather than in the business of claim payments in the near future.
3. Peer-to-peer insurance
The fact that customers also have more information available on their own risks and are able to share this with others, gives rise to the thought that pools of customers might want to share their risks together first and only insure the larger risks with insurers. This peer-to-peer insurance is also in line with other trends in the sharing economy. Bringing this service to clients has mainly received attention from InsurTechs so far, but gives the incumbents an opportunity to attract a new world of data. On the other hand, the insurers that cannot or do not provide this service themselves, will need to adapt their analytics to this new environment.
Tailored services and policy conditions
A last possible use of data analytics which we present here refers to tailored services and policy conditions. Data analytics can identify what customers might prefer as additional services to their insurance policy. Currently, insurance products (especially in private lines) normally include general services and conditions that are equal to all clients. However, analysis might indicate a customer has more interest in the insurance product if it includes certain services (e.g. prevention services on home insurance). Another example is to adjust deductibles or the maximum sum insured per customer, to be able to accept a risk that would otherwise be rejected or very costly. Adapting the product and its conditions dynamically, provide opportunities for an increased, and more satisfied, client base.
If the insurers are capable of using data analytics to find their clients’ real needs, they can offer a better range of products to more satisfied customers.
What to do and where to start
Unfortunately, there is no one size fits all approach regarding the use of data analytics at insurance companies. The size of the organisation, analytical maturity, involved business lines/functions, complexity of the process and products must determine the best approach. However, besides the abovementioned application of data analytics, we have three recommendations for increasing the added value derived from data analytics:
1. Take a holistic approach to the long term business strategy but start small. Our recommendation is to start with an organisation-wide holistic approach that keeps in mind the data analytics capability an insurer will require in the future. In addition, while keeping the bigger picture in mind, start small, with proofs of concept via an agile approach. This will help to focus on the long-term strategic goals (bigger picture), serving the organisation as a whole, but focus on agility and showing results in the short term. More information about an agile way of working can be found in the agile section of this Outlook.
2. Leverage the ecosystem. Because the marketplace changes rapidly, it is easy to be overwhelmed by the required data analytics capability to be at par, or even ahead of, the marketplace. There is also an obvious benefit in having the data and being able to make sense of it yourself. However, as both sources and volumes of data grow exponentially and there is a scarcity of talent, insurance companies need to investigate different options. Our recommendation is to investigate the possibilities of analytics eco-systems: ecosystems focusing on the data (selling, supplying of brokering data), talent management (universities, business school, innovation hubs and consultancy firms), or crowd sources.
3. Address the behavioural challenges. Besides having the data, the scientist, and insights, there is a cultural and behavioural change required to obtain maximum value from data analytics. Hence, our third and last recommendation is to change the mindset of the organisation and decision-makers. By this we mean that a loud and clear message should be communicated about analytics and its importance, and that success stories should be spread across the whole organisation. This can be using the physical environment (war rooms, laboratories, large touchscreen) but also in the shape of cultural angles, such as thinking about talent in a new way.
Improving the application of data analytics in the abovementioned areas and keeping in mind the recommendations provided will help insurance companies to use data analytics to improve customer service and achieve sustainable growth. Not only will the business benefit from more insightful analyses, decisions will be based on facts, and the culture of an organisation will be data driven. Insurers need to be prepared to accept that the process of becoming data-driven is sometimes like A/B testing: some analytics initiatives will fail. But the key is to be agile, learn and grow!