Dissecting InsurTech — Part 2: But will it scale?
Welcome to part 2 of my analysis “Dissecting InsurTech”. Let’s quickly recap: In part 1, I identified sub-categories within the insurance industry that have the potential for high value creation. I did this in two steps — by differentiating B2B and B2C companies, and applying a more detailed framework for each of the categories. In the following, I make an attempt to identify sub-categories that will allow for the creation of highly valuable technology businesses.
Step 3: Which activities can be turned into highly valuable technology based business models?
Now that we have determined the sub-categories that allow for high value creation in InsurTech, the last step is to determine those sub-categories that will allow tech startups to build large businesses and create massive (enterprise) value. The following prerequisites for this to be possible come to mind:
- Scalability: Does the sub-category face a problem or inefficiency that is repeatable for a large number of cases? This could be different things, for example scalable distribution to large customer groups or standardized processes across different countries/legislations
- Defensibility: We like to invest in companies that are able to create a “lock-in effect”, i.e. an increasing defensibility of the business as it grows. This can be for example through economies of scale or a continuous improvement of the service/product that can only be achieved through more usage (e.g. machine learning algorithms). Defensibility can also be created through high barriers to entry
- Customer inertia: Insurance is an old industry — insurance takers are locked into long-term contracts, and incumbents are large, slow moving organizations. In order to gain customers (B2B and B2C), startups have to be able to overcome their inertia. On the B2C side, P&C insurance tends to have a significantly lower customer lock-in than health and life insurance
- For B2C — Unification of customer experience: As mentioned above, the more of the customer experience is improved, the more value this provides for the customer
- For B2B — Effect on combined ratio: Looking at the CR, a startup that attacks the 70% claims costs, by simple math has a stronger lever to create value than one that attacks the 20% operational costs
Please note that there is no formula for a certain number of criteria that need to be fulfilled, or a ranking of which is the most important one. Considering the criteria above, here is my take on the sub-categories that have strong potential:
On-demand insurance: On-demand insurance aims to provide coverage only exactly when needed (e.g. I’ll only insure my expensive camera when I’m taking it outside), and thereby reduce costs for the insurance taker. The concept is relatively new and requires a high degree of automation and flexibility from the offerer to process the insurance requests. This alone requires high scalability and can then be extended across a range of different use cases. There is also little customer inertia, because in many cases there have been no insurance policies in place before. Finally, strong lock-in can be achieved through collecting a lot of customer data that can be used to continuously improve pricing, product offerings, etc., and stay ahead of the competition.
Peer-to-peer insurance: The attractive thing about P2P insurance is the creation of new risk pools. Part of the risk is distributed among a smaller risk pool, and the “better” (read: fewer losses) the pool, the cheaper the insurance gets. In turn, this makes it increasingly attractive to join such a pool, thus creating a lock-in effect through positive selection of risk. Furthermore, MGAs and carriers operating this model, will profit from the positive selection by reducing fraudulent and rightful claims, which potentially has a large effect on the combined ratio. P&C insurance seems most promising for this model as it has the lowest customer lock-in and relatively simple and standardized insurance policies.
New digital insurance carriers: Entirely new carriers are not dependent on partnering with an existing insurance, which allows them to own the entire customer experience and gives them maximum flexibility for creating new products. Defensibility is quite high due to the high barrier of obtaining an insurance license. However, in many cases we see that customer inertia can be an issue. Especially in health and life insurance, customers have to make large time and money commitments and have high switching costs, which may make it difficult to acquire customers with reasonable time and cost efforts.
Loss prevention: As shown above, loss prevention has a huge potential to improve the combined ratio for carriers. Technology can help reduce losses for example through a wide range of IoT devices in homes or cars and analyzing the resulting data. In contrast to some of the other sub-categories, this doesn’t require integration into insurance policies or processes and is easily scalable across large customer groups. Furthermore, insurers should be quick in investing in and implementing loss prevention measures, since it doesn’t require a deep integration into existing processes, but rather happens before.
Fraud prevention: As with the previous point, fraud prevention holds enormous potential to reduce claims for insurers, although here the goal is to identify the fraudulent ones. This can also be achieved through IoT devices and analyzing customer behavior. I believe that machine learning and AI has large potential here, and constantly increasing detection accuracy allows for strong lock-in effects.
Claims management efficiency: Compared to actual claims payouts, operating costs constitute a rather small part of the cost base of an insurer. For P&C insurance, the largest part thereof (approx. 25%) is caused by claims processing (see “Successfully reducing insurance operating costs”, McKinsey 2014, p. 6). However, there is still big potential for InsurTech companies to create value in this area. Claims management is a standardized process that is run in high frequencies, but still involves many manual steps. Increasing efficiency with technology can therefore result in huge cost savings, as some startups are already showing today.
Risk modeling and predictive underwriting: Also here, data plays a huge role. New data sources from IoT and mobile devices, vehicles and others can be fed into pricing and risk models in order to improve product pricing and margins. The same holds for new processing methods like image recognition.
Summary of part 1 and 2
In the two parts of this analysis, I took three steps to dissect the insurance industry and made an attempt to identify areas that are well suited for the development of highly valuable tech companies. Of course, there are many “landscapes” out there, and you can structure the industry along a number of criteria. However, I believe that this approach gives a particularly clear picture on where a lot of value can be created. Furthermore, in comparison to a mere industry landscape, it also describes the steps that led to the result.
That said, the result is a list of hypotheses rather than certainties, is in no particular ranking, and it does not claim to be exhaustive. I’m curious to see in which (other) areas we will see InsurTech unicorns, as evidently there is plenty of opportunity to create value in the massive insurance industry. I’m looking forward to further engaging in discussions — please feel free to reach out!
