Digital Insurance: Architecting to Differentiate
In this post I will show what insurers must to do to build a differentiating digital insurance business. In my view, insurers need to invest in new digital capabilities across channels, analytics, and operations in order to compete in the new digital insurance market. This requires insurers to determine how they will differentiate, understand the value digital capabilities provide to their customers, define the future state enabling digital capabilities, and establish a real-time integration layer between these capabilities. In doing so, insurers should adopt a multi-speed business architecture that allows channels, analytics, and operational capabilities to evolve at different ‘speeds’ in order to gradually roll out new disruptive capabilities whilst retaining the performance, stability, and integrity of existing products and lines of business.
Disclaimer: the views expressed in this article are entirely my own and do not reflect those of my employer or its customers.
The State of Digital Insurance
The insurance industry is transforming significantly with the demand for better digital services being one of the key drivers of change. A recent survey of more than 9,000 insurance customers globally (see Insurance 2020: The digital prize — taking customer connections to a new level) indicated that:
- 71% used digital research before buying an insurance policy
- 68% were willing to download and use and app from their insurer in order to access services
- 65% were willing to install a sensor in their car or home in order to reduce their insurance premium
As a result, insurance carriers deploy new strategies to engage with customers and deliver services through digital channels. Key strategic responses deployed in the global insurance industry include:
- Digital self-service, which provides convenient 24/7 access to key underwriting, claims, and billing processes through multiple digital channels, i.e. tablets, mobiles, and PCs. Key benefits from digital self-service include increased customer convenience, improved customer experience, and reduced workload on existing channels such as the contact centre
- Better decisions through better data, which means leveraging both structured and unstructured data to improve the quality, timeliness, and precision across customer facing, operational, and risk and fraud related processes. Sometimes better data also means exploring new avenues of data gathering such as deploying machine to machine technologies in cars, providing insures with better data about customer behaviour and risk whilst reducing the premium of the policyholder. Key benefits include more responsive services and transparent, informed decision making based on data
- Digital insurance operations employing extensive use of analytics and automation in distribution, underwriting, claims, and billing processing which eliminates the need for manual rekeying, trivial risk assessments, and user intervention. Benefits include improved predictability, management of claims by exception, fast approval times, and lower cost to serve
The challenge insurers are facing is what specific digital capabilities to build and how to launch and sequence them across multiple product lines, lines of business, and channels. They must also take into account both direct sales and brokers whilst at the same ensuring the ongoing stability and reliability of core insurance processes. In addition, insurers face the challenge of an increasing technical debt and legacy core insurance systems, which impede the flexibility of and speed to market for new insurance products and processes. Whilst it is tempting to undertake a complete overhaul of core insurance platforms starting from scratch, this usually also involves a great deal of cost, risk, and complexity to stakeholders.
The solution: build multi-speed business capabilities
Based on my experience, there are three general modes — or speeds — of digital insurance capabilities:
These are explained in the following.
Channels capabilities provide the business capabilities for providing digital customer touch points along the insurers’ value chain and providing a starting and monitoring point for straight through processing of customer transactions. In the graphic below we show some of the key digital touch points for a large general insurer, all of which would be supported through a unified insurance channels layer. Channels are often constructed using APIs in order to support many different types of client devices and endpoints and enable quick and agile roll-out of new insurance products and services to new device types.
Above: Sample digital opportunities along the general insurance value chain
Analytics capabilities provide the business capabilities for storing, analysing, and reporting on very large volumes of core insurance data captured in the channels. As insurers are increasingly gathering more data points about policyholders, incidents, and claims in order to improve claims efficiencies, personalise policies, and prevent insurance fraud, the need for better real-time management reporting increases rapidly. Whilst the notion of “big data” and in memory/real-time data stores have certainly made it easier to consume and analyse large volumes of data, insurers are still struggling with consolidating data kept in legacy applications and mainframes. The challenge of enterprise architecture therefore becomes how to build an information model that integrates and analyses customer and operational data from a variety of different data sources, both internally and externally, in real-time and supports effective decision making in a true multi-channel insurance carrier.
Analytics and the search for better data should not become a goal in itself. The key goal of analytics capabilities for insurers is often ‘a single customer view’, e.g. a single view of and for the customer,which both delivers an end to end view of the customer’s policies and claims to the insurer and the customer across product lines, lines of business, and channels.
In my experience, the nature of the underlying technologies supporting the data & analytics layer in insurance is shifting from the classic centralised data warehouse with a canonical information model to decentralised “data lakes” where in-memory analytics engines pull and process data from various insurance platforms in real-time in order to produce a result. With the advent of flexible real-time analytics engines, it is more cost efficient to pull and combine data in real-time as opposed to constantly extracting and transforming data in bulk to a single, centralised data store with the overhead it brings. The second challenge for enterprise architecture in the age of digital business therefore becomes how to build and integrate different data lakes in order to provide better, faster management reporting and enable more efficient self-service access to and discovery of data without the need for employing expensive and complex BI tools. Deploying a variety of data analytics technology (e.g. NoSQL and deep learning engines) also assists in tackling the increasing volumes of semi- and unstructured data, which insurers can harvest to develop better data about their customers and assess insurance risk.
The operational capabilities layer contains the non-differentiating business functions that “keep the lights on” of the entire organisation, ranging from finance and procurement to HR and IT. For insurance carriers large, integrated ERP platforms no longer provide a differentiating capability in itself. Most competitors have adopted ERP platforms and run competitive operations. On the other hand, no large carrier can operate for long without a single finance system. The existence of the operational capabilities becomes the precedence for and foundation of digital capabilities built in the channels and analytics layers and more specifically:
- Building a solid business and systems architecture for delivering reliable, integrated operations across corporate functions; and
- Providing easy amalgamation of channels and analytics capabilities back to operational capabilities, i.e. integrating key transactions in the channels layer to finance and payments systems as part of a straight through processing initiative
Each tier represents the strategic imperatives and priorities of different business stakeholders. Marketing and sales wish to deploy new products and services faster via the channels layer and provide a better customer experience to attract and retain customers. Actuaries and underwriters wish to deploy new analytics capabilities to better understand the behaviour of customers and claimants in order to optimise pricing and reduce risk. Operations managers wish to ensure the continued stability and reliability of core insurance systems in order to minimise operational risk and improve inefficiencies. The crux of digital enterprise architecture therefore becomes catering for these different needs and priorities in an engaging, and coherent manner and articulate the strategic roadmap of business initiatives and supporting platforms to make it happen. I have summarised this role in the table below including typical differentiating capabilities:
The figure below shows how the three tier digital architecture applies to the high level business capability model of a global general insurer:
Above: Example digital insurance capability model for a global general insurer
In this example, customer channels, engagement, and core insurance processes across underwriting, claims, and billing enabled the digital channels tier whilst data, analytics, risk, and management of service providers enabled the analytics tier. Back-office and shared service capabilities such as IT operations, HR, and finance were necessary to provide a stable business capability foundation for the digital strategy to evolve and grow.
Thinking of their capabilities in different tiers providers insurers with the following benefits:
- Differentiating capabilities for different strategic imperatives can be evolved at different speeds, depending on budget and risk appetite of stakeholders.
- Architecture concerns are decoupled in a business oriented way. Historically, enterprise architecture has focused on the separation of different business and technology layers in order to translate business concerns into technology requirements. Instead, we argue that decoupling of architectural components should begin on the business architecture layer in order to inform multiple strategies evolving at different paces.
- Digital initiatives can be implemented without needing to undertake a complete step-change systems overhaul. Instead, each tier can be partitioned further into separate business domain architectures with separate core platforms and technology investments.
In order to realise these benefits, one particular capability that becomes critically important: enabling real-time integration between the three tiers. Enterprise architecture would typically propose a service oriented architecture to make this happen, whereas digital entrepreneurs would suggest a more granular architecture of APIs and micro services to make it happen. Regardless of the technology choice, the key question that must be solved from day one is how the different capabilities and underlying applications will exchange information in real-time in order to avoid “information stovepipes” and enable continued sharing of data and insights from the analytics tier to stakeholders. The analytics tier requires insurance data to be fed in real-time in order to provide the differentiating advantage to the business — otherwise the architecture will end up as another silo of information and poorly informed decision making as a consequence.
In my experience, there are four key steps for insurers to build a successful digital business, which are:
- Determine how your business will differentiate digitally in the market;
- Understand the distinct value proposition for each capability, ensuring that the investment delivers new value to the customer; and finally
- Define the enabling digital capabilities (people, process, and technology) required to transform and compete.
- Integrate capabilities and capability layers in real time
A successful digital insurance strategy starts with identifying the essential digital capabilities, that will enable you to differentiate in the future. These will differ between each tier, depending on the product, customer segments, and market maturity. The table below shows some of the typical differentiating capabilities and supporting technologies, which insurers invest in, given the maturity of the market or market segment in which they operate:
Once the differentiating capabilities are known, the next step is to crystallise the digital customer value proposition for each insurance product and segment that is served. The capabilities need to clearly enable each value proposition, ensuring that customers receive a uniform service across all channels. To measure success, having a single view of the customer in terms of risk, premiums, policies, and claims history becomes increasingly important in serve the customer equally well regardless of the channel and fully leverage customer analytics to cross-/up-sell and provide personalised discounts.
The third step is to define and develop the enabling digital capabilities in terms of people, process, and technology. Apart from the broad investment in digital systems and technologies (i.e. analytics to enable omni-channel customer engagement), this also requires a wider shift in organisational ways of working such as:
- Driving a digital culture and mindset such as developing working prototypes and prioritising a minimum viable product over fully defined & documented requirements
- Embracing continuous development practices with agile development teams collaborating closely with business/product owners and executives
- Adopting contemporary development and IT operational best practices such as DevOps, ensuring that an agile and scalable infrastructure is in place to enable the digital capabilities
The fourth and final step is to establish the real-time integration of information between the capabilities and capability layers. This is often (but not necessarily) done by building an API layer, which wraps all digital capabilities as accessible, reusable services that are accessible across all channels and technologies (i.e. both enterprise applications and mobile technologies). Embracing real-time data is critical in ensuring customer requests are processed instantaneously and decisions are based on up to date information.
 See PwC US Technology Forecast, Issue 1 2014: : The enterprise data lake: Better integration and deeper analytics (2014) — direct link.
 ERP is short for Enterprise Resource Planning and refers to large integrated financial software platforms (such as SAP).
Many thanks to Mark Johnson for providing thorough, value feedback to this blog post.