Skip to main content
SpringerLink
Log in

The Impact of Digitalization on the Insurance Value Chain and the Insurability of Risks

  • Published:
The Geneva Papers on Risk and Insurance - Issues and Practice Aims and scope Submit manuscript

Abstract

Based on a dataset of 84 papers and industry studies, we analyse the impact of digital transformation on the insurance sector using Porter’s value chain (The Competitive Advantage: Creating and Sustaining Superior Performance, The Free Press, New York, 1985) and Berliner’s insurability criteria (Limits of Insurability of Risks, Prentice-Hall, Englewood Cliffs, NJ, 1982). We also present future research directions from the academic and practitioner points of view. The results illustrate four major tasks the industry is facing: enhancing the customer experience, improving its business processes, offering new products, and preparing for competition with other industries. Moreover, we identify three key areas of change with respect to insurability: the effect of new and more information on information asymmetry and risk pooling, the implications of new technologies on loss frequency and severity, and the increasing dependencies of systems through connectivity.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Figure 1
Figure 2

Similar content being viewed by others

Notes

  1. See, for example, Moreau (2013) on the music industry or Chathoth (2007) on the travel industry; we also refer to Back et al. (2016) and Kane et al. (2015) for cross-industry comparisons on the importance of digitalization.

  2. Müller et al. (2015).

  3. Catlin et al. (2015).

  4. Dozens of media articles and studies analyse the impact of new technologies on customer satisfaction and loyalty (e.g. Maas and Bühler, 2015; Moneta, 2014), on the improvement of cost structure and business processes (e.g. Berger et al., 2016; Catlin et al., 2015; Chester et al., 2016), on the future workforce (e.g. Johansson and Vogelgesang, 2015), and on the insurability of new risks (e.g. Biener et al., 2015, for cyber risk). These industry studies focus on specific digitalization trends and their strategic implications; none of them offer an overview of the existing knowledge on digitalization.

  5. Porter (1985).

  6. Berliner (1982).

  7. E.g. Biener and Eling (2012); Biener et al. (2015); Eling and Schnell (2016).

  8. In our paper, we do not focus on literature on algorithms and computational methods. Regarding these topics, we refer, for example, to Salcedo-Sanz et al. (2013).

  9. We also searched for “digitization” instead of “digitalization.” The results were often the same, even though the words are typically defined differently. See “What is digitalization and which technologies will influence the industry?”.

  10. The Journal of Finance, American Economic Review, Journal of Risk and Insurance, Insurance: Mathematics and Economics, The Geneva Papers on Risk and Insurance—Issues and Practice, The Geneva Risk and Insurance Review, Journal of Insurance Regulation, and Risk Management and Insurance Review.

  11. Rahlfs (2007).

  12. Biener et al. (2015).

  13. The terms “digitization” and “digitalization” are sometimes used synonymously and sometimes not. “Digitization” is often defined in the technical context of making analogue data digitally available (e.g. Ingleton et al., 2011; Breading, 2012)—for example, scanning of paper contracts. In contrast, “digitalization” is a broader description of the transformation of the economy and society.

  14. Ingleton et al. (2011).

  15. Tischhauser et al. (2016).

  16. Also, focusing on the business, but in a more abstract way, Dörner and Edelman (2015) describe “digitalization” as a process of “creating value in a new business environment”, “creating value in the customer experience” and “building capabilities to support this structure.”

  17. Hiendlmeier and Hertting (2015).

  18. Back et al. (2016).

  19. We do not discuss virtual reality which has been mentioned by some studies but whose applications to insurance have not yet been developed.

  20. A high-level summary of the technological impact on the value chain is also presented in Appendix C (the so-called “value chain and technology matrix”).

  21. Bieck and Tjioe (2015) find that people under the age of 30 are more open to non-traditional insurance providers (e.g. auto dealers, retailers). Bieck et al. (2014) find that future customers will be less price sensitive, will seek advice, want personal multi-channel interaction, and be open to new products. Concentrating on the motor insurance segment, Barwitz et al. (2016) define four customer segments based on the interaction between customers and insurers, independent of socio-demographics: utilitarians change the interaction frequently, depending on their personal benefit; hedonists prefer a high-quality and personal interaction; cost-minimisers want to reduce money and time investments; relationalists prefer personal interaction and stay loyal to their agent. Catlin et al. (2013) define nine customer segments, depending on the preferences for price, brand, loyalty, convenience and personal advice.

  22. One important aspect in this context is that customers research online, but then purchase offline via traditional channels (ROPO). For example, 84 per cent of German consumers gather information online to buy insurance products, but the majority purchase them offline (59 per cent research online and purchase offline—ROPO); only 25 per cent are pure online customers (Zurich et al. 2016). The ROPO behaviour also depends on the product type: whereas 77 per cent of pension plans are researched online and purchased offline, this only holds for 50 per cent of motor insurance plans. One resulting challenge for insurance companies is the need to create a uniform customer journey, i.e. the customer expects to get the same information in the same quality at any time and through any channel (Pain et al., 2014). Also noteworthy are the large differences across countries when it comes to aggregators. While for example in Germany 41 per cent of insurance customers use aggregators for the evaluation of motor insurance policies, only 27 per cent of Swiss and 23 per cent of Austrians use them (Barwitz et al., 2016).

  23. Maas and Bühler (2015) find that today on average 41 per cent of processes are automated in the German, Swiss, and Austrian insurance industry, and health insurers have already automatised 47 per cent of processes; they estimate automatisation will increase by 28 per cent, leading to an average cost saving of 14 per cent. Catlin et al. (2015) note in their global study that 70 per cent of processes today are done mostly manually, 25 per cent are partially automated, and only 5 per cent are fully automated; through digitization, only 15 per cent of processes will be still be done mostly manually whereas 50 per cent will be semi-automated and 35 per cent fully automated; it is possible to save 30 to 50 per cent in non-commission costs through automatisation. Note that neither of the studies mention a time period for reaching full potential.

  24. The annual spending on big data analytics will increase globally in the next three to five years by 24 per cent in the life segment and by 27 per cent in the P&C segment (Müller et al., 2015).

  25. For more detailed information on the tools, see, for example, SAS (2017) and Fayyad et al. (1996). In addition to traditional statistical methods, data mining uses machine learning algorithms, which iteratively learn from past computations, for more efficient analyses. Usually, machine learning algorithms are trained on existing data and then automatically analyse new data sets (Hall et al., 2016). Most commonly, for the analysis classification, clustering or regression algorithms are used (Fayyad et al., 1996). Machine learning algorithms are also used for applications which cannot be programmed by hand (e.g. handwriting recognition) or self-customising problems (Amazon, Netflix product recommendations). The newly generated information can be used for applications such as risk allocation, customer segmentation, exploiting cross-selling potential, and fraud detection (Jones, 2016).

  26. Hussain and Prieto (2016).

  27. For example, the EU has reformed its data protection rules to simplify the use of big data for businesses and to set high standards of data protection (European Commission Justice, 2016). Furthermore, see Krotoszynski (2015) for a detailed comparison between the U.S. and EU legal systems regarding privacy rights.

  28. For the discussion on motor insurance, we refer, e.g. to Paefgen et al. (2013), Filipova-Neumann and Welzel (2010), or Keller and Transchel (2016). Anchen et al. (2015) present some thoughts on wearables for the life insurance market.

  29. PWC (2015).

  30. One example is the use of big data techniques for risk underwriting and analysis; Climate Corporation (US) uses climate and soil data to offer farmers insurance against losses from weather events (Müller et al., 2015). AllLife (South Africa) offers life and disability insurance to policyholders, who suffer from HIV or diabetes; in their monthly health checks, every client gets a personalised analysis and advice on managing their conditions. To assess their clients’ conditions the insurer has direct access to medical data from medical providers. If clients do not follow the check-up plan, coverage can be reduced or cancelled (Brat et al., 2014).

  31. Cant et al. (2016).

  32. E.g. Haller (1997).

  33. Maas and Bühler (2015).

  34. Maas and Bühler (2015) find that 42 per cent of IT development and 51 per cent of IT operations will be outsourced by 2020.

  35. We can also observe the same behaviour in the other direction, i.e. the increasing efforts of insurers to grow outside their core business. For example, Allianz recently bought the auto sales platform Instamotion Retail (Hegmann, 2016).

  36. The same arguments can be made for other industries, e.g. mobile phone providers.

  37. For example, technology firms offer payment solutions (e.g. Apple or Samsung Pay). Another example is Telefonica which offers a telematics motor insurance (O2 drive) in the U.K. However, it seems that Telefonica is not carrying the risk, but is rather an intermediary to other insurers. They are providing the customer with a telematics device.

  38. We also refer to the valuation of companies like Google or Facebook, whose most valuable asset is the data of their customers.

  39. Note that the return on equity consideration does not only mean investing in other industries but also working with the insurance industry with alternative business models. For example, Google has even withdrawn the insurance broker Google Compare from the U.S and U.K markets. One might suspect that the profit from pure advertisement on Google is higher than the potential profit from acting as a broker or risk carrier, all of which require substantial expertise and upfront investments (Jergler, 2016). Another example in this context is that both Amazon and Apple are working with insurance companies (London General Insurance Company Limited and AIG, respectively) for their warranty programs Amazon protect and AppleCare+.

  40. For more examples, see KPMG and H2 Ventures (2015).

  41. KPMG and CBInsights (2016).

  42. Wiener and Theis (2017).

  43. We refer to Mesropyan (2016), Noack (2016), Kottmann and Dördrechter (2016), and Braun and Schreiber (2017) for more examples. We also refer to Braun and Schreiber (2017) for a more detailed discussion on the role of insurtechs, their business model, and how they could affect the insurance industry.

  44. We note that by using GetSafe the company is also contracted as the customer’s broker. As a consequence, it is getting the trailer commission.

  45. Barwitz et al. (2016).

  46. Kottmann and Dördrechter (2016).

  47. We emphasise that the reasons are outside of today’s industry perspective. Examples from other industries (e.g. low-cost carriers in the airline industry) have shown the impact of a possible disruption that was not previously envisioned either.

  48. For example, health and life insurers could not only separate people by age and by whether they are smokers or non-smokers but, for instance, by how physically active they are. Another example is in motor insurance, where data are enriched with information about driving behaviour (acceleration, braking behaviour, speed, etc.).

  49. There is also a possibility that the customers share their information only with a technology provider (e.g. Apple tracks the usage behaviour of their iPhone customers). In this constellation, the technology provider supplies the risk calculation and prevention. In exchange for carrying the risk, the insures get a minimum margin.

  50. Hoy and Ruse (2005).

  51. Doherty and Posey (1998).

  52. Erixon and van der Marel (2011).

  53. We note that better prevention could reduce loss probability and loss amount so much that the utility gain from transferring the risk is not enough to justify the transaction costs of an insurance company. Then the idea of insurance would be in question. We see this, however, as a rather extreme scenario, which is not going to be realised for at least the next ten years. But customers prefer the accident that never happens, i.e. digitization will lead insurers to move toward mitigation and prevention.

  54. Treacy and Wiersema (1995).

  55. Müller et al. (2015) argue in the same direction by introducing four strategic pathways: advanced analyser, digital distributer, customer-centric insurer, and effective operator. Johansson and Vogelgesang (2015) predict that the digital transformation will also impact the workforce of insurance companies; insurers will need to attract new employees with knowledge in data science, analytics, and/or IT development. Moreover, there will be a significant number of layoffs in the operations department.

  56. Burgelman et al. (2008).

  57. Filipova-Neumann and Welzel (2010) develop a model for telematics motor insurance demand. Other studies focus on the impact of assistance systems or financial incentives on driving behaviour, e.g. Hummel et al. (2011) or Bolderdijk et al. (2011).

References

Download references

Author information

Authors and Affiliations

  1. Institute of Insurance Economics, University of St. Gallen, Girtannerstrasse 6, 9010, St. Gallen, Switzerland

    Martin Eling & Martin Lehmann

Authors
  1. Martin Eling
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Martin Lehmann
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Martin Eling.

Appendices

Appendix A

Table A1 Dataset of papers and industry studies
Full size table

Appendix B

Table B1 Definitions of digitalization
Full size table

Appendix C

Table C1 Value chain and technology matrix—Summary of technology impact
Full size table

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Eling, M., Lehmann, M. The Impact of Digitalization on the Insurance Value Chain and the Insurability of Risks. Geneva Pap Risk Insur Issues Pract 43, 359–396 (2018). https://doi.org/10.1057/s41288-017-0073-0

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1057/s41288-017-0073-0

Keywords

Search

Navigation