Articles - LLMs and the Future of Insurance Choice

By Isabelle Williams (LCP), Catrin Townsend and Jeremy Keating (Price Writers)

PCWs have undoubtedly been a game-changer for the customer, but the system does have its flaws. Ultimately, even though PCWs have made an effort to introduce more clarity related to the true total cost and coverage of the policy, the PCW game is still mainly predicated on price, and not product features or quality.

Once a marketplace has become price-driven, the real competition is in the product cost. There is a natural tendency to start “hollowing out” products, reducing cover and increasing excesses while keeping the product looking “the same” to the customer. An example of this can be found in the “Essential” Motor products that now frequently do not include windscreen coverage, a previously standard part of all motor insurance products. 

LLMs could significantly change this game, putting more power back in the hands of both customers and insurers. But how?

The difference lies in the plain English interface. Where LLMs truly shine is their ability to allow customers to really dig into the cover they’re planning to purchase. This could reduce the risk that a customer accidentally exposes themselves to risks they would rather avoid, or purchases additional cover that offers little to no value to them.

Secondly, even after purchase, LLMs can be incredibly useful to cut through the jargon in the policy documents, ensuring that the cover the customer has purchased is what they expect. If not, there’s always the cooling off period. This output can also be provided directly to the insurer, giving them useful context on cancellations. 

Thirdly, while LLMs can make the journey through PCWs easier, thereby further increasing the probability of churn, there is a scenario where they encourage customers to renew with their current insurer.

Imagine this: a customer has set up a yearly automated process to gather insurance quotes and coverage details. After reviewing these, the LLM notes that the renewal offer received (currently sitting in the customer’s inbox), requires no click through, no subsequent actions, has all the coverage they want and is only a few pounds more than the cheapest quote. Customers then face a choice: save a few pounds but continue spending time on the insurance journey, or to allow the policy to auto-renew. Because no (human) time has been spent on the quote journey, people may be less invested in changing providers, therefore reducing customer churn.  

Conversely, less interaction with insurers directly could make customers less engaged with their current provider, which could erode the impact of brand in insurance purchases. This could make customers more likely to change providers.

These are just three examples of how LLMs could affect the PCW process, but from each example, it is clear that LLMs have the potential to change customers’ decision-making processes, putting less importance on the price alone and more on the overall appropriateness of the product for the individual in question.

When a customer’s decision making process changes, this naturally changes the nature of the PCW game. Supposing that PCWs retain their hold on the industry, the nature of the rankings on their websites and in LLM apps may change in response to the increased importance placed on product suitability. This in turn could cause insurers to put more of a focus on developing truly differentiated products that target specific customer needs, instead of focusing on branding and product tiers as a way to differentiate products. 

The capabilities of LLMs do, however, pose a question as to the PCWs purpose as a middleman between insurer and customer. While some sites have already chosen to embrace a future with LLMs via their own dedicated apps, there is a question as to whether customers will really need their services as a middleman in the future.  

Via Codex and other similar tools, with the answers to the standard personal lines question set on hand, it is possible to design and run a script that will return quotes across all potential providers, whether or not they exist on a PCW. While this won’t be a path that every customer wants or even could take, it will certainly be possible for insurers to target those customers that do, and offer products more aligned to their needs.  

This, however, reveals the true downside of LLMs. Inequality of access to LLMs, and the knowledge required to be able to create these automated pathways, is a huge issue even in developed economies such as the UK. Those who do not have access to LLMs, or those simply want to exist outside the AI pipeline, could easily end up at a disadvantage to those that do.  

If we are not careful, LLMs could introduce a two-tier system. The changes in market demand could well lead to the creation of well-differentiated product offerings, available direct on insurer websites for those who are able to use the tools to easily access them, potentially available at a better price than equivalent generic products if the insurer decides to pass the PCW fee savings on to the customer. This could directly disadvantage customers not using this method, where generic products bought via PCWs will still be the easiest way to access insurance. 

We are still at the very beginning of this journey, and there is a long way to go before we see the true effects of this market destabilisation, but one thing is certain – LLMs are changing the world, and PCWs (as well as the insurers who use them) need to be prepared for the future.