21
Examination of witnesses
Alistair Hargreaves, Jason Storah and Nick Turner.
Q94 The Chair: Welcome to today’s session, which is the eighth oral evidence session as part of the committee’s inquiry into the regulation of the consumer insurance market. I thank Mr Hargreaves, Mr Storah and Mr Turner for attending. The session is open to the public. It is broadcast live and subsequently accessible via the parliamentary website. A verbatim transcript will be taken of the evidence and put on the parliamentary website. Before we move to detailed questions, I invite each of you briefly to introduce yourselves and then I will get going with the first question.
Alistair Hargreaves: I am the UK insurance CEO of Admiral. Thank you for inviting me to give evidence. Admiral is a motor insurance company that was established about 30 years ago and has grown to serve about 9 million customers, including in home and travel insurance.
Jason Storah: I am the CEO of Aviva’s UK and Ireland general insurance business. Aviva is the leading insurance company, wealth and retirement business in the UK. We have 21 million customers in total and insure about one in five homeowners and one in five travel policyholders in the UK.
Nick Turner: I am the CEO of NFU Mutual. We have about 935,000 customers. We are a rurally focused mutual insurer, so we are owned by our members, which makes life a bit easier for us in thinking about the core purpose of our organisation. That is us—focused on the rural community.
The Chair: Excellent. I should declare that I am a long-standing policyholder with NFU Mutual, both for motor and home. Can we start with some context, part of which Mr Storah just gave? We are concentrating on the consumer home and travel insurance markets. Can you outline the total size of the UK market and the position and market share each of you has in those markets?
Alistair Hargreaves: The home insurance market is about £6 billion in terms of premium. We have been in that market for about 13 years. We have just over 2 million customers, which is about 7% market share. The travel insurance market is just over £1 billion-worth of premium. We have been in that market for about seven years and have just over a million active customers. Again, that is around 7% market share.
Jason Storah: I will not repeat the market share numbers because I agree with what my colleague here said. As I said, we insure one in five homes in the UK and about one in five travel policyholders in the UK. If you extrapolated that out for market share, based on his numbers, we would be aligned.
Nick Turner: Of our 935,000 customers, about 370,000 have home insurance. We are minuscule in the travel market. We only write travel insurance for high net worth customers as part of their high net worth policies. We do not write for new business at the moment on travel.
The Chair: What percentage of market share do you have of home insurance?
Nick Turner: It is 370,000. It would be very low, so 1% or 2%.
The Chair: One of the things we have been struck by is the high rates of claims being declined by insurers in the home and travel markets, with varying percentages, some as low as 60%. Can you say what your decline rates are for your businesses and describe what you see as the root causes of those low decline rates?
Alistair Hargreaves: To start with home insurance, our claims acceptance rate, which is the other side and is the measure that the FCA reports, is about 70%. That is different from an insurance product such as motor insurance. When we look at that, we think it is predominantly because of the customer understanding of what is covered by home insurance. In terms of the way in which we record that number, it is important to note that any queries we get we will count as a claim. If a customer is calling us about some tiles that have come off their roof, even if there has not been a storm, and we explain that that is wear and tear and it is not covered, we would still capture that because that information is really helpful for us to understand consumer understanding. We do not want in any way to put off a customer contacting us about those queries. We then use that information to improve our communication at point of sale throughout the journeys, remind the customer about that and make other changes, which I am sure we will get on to.
The Chair: And for travel?
Alistair Hargreaves: For travel insurance, our acceptance rates are about 85% for our main products. Similarly, travel insurance covers a number of aspects, but we see that it is about understanding of the product, in terms of whether there are cancellations or delays. We make sure that those are disclosed clearly up front. Again, we are really working to improve that through educating customers.
Jason Storah: It is similar for us. From a claims acceptance perspective, out of every 100 home claims we receive, we accept and pay 72. If you then take that delta between the 72 and 100, of those remaining 28 claims, just to build on what my colleague said, 13 of them would not involve an insurable peril—for example, if somebody was phoning up about a claim but there was no storm, or it was due to wear and tear. Perhaps a roof got damaged because of the lack of maintenance and wear and tear on the roof, rather than an insurable event like high winds because of a storm.
The Chair: Do you operate the same definition of claim as Admiral?
Jason Storah: For the purpose of how the FCA captures the data, yes. The FCA includes queries, as my colleague said, but we all use the same definitions. So, that is 13 of the 28 that we would not have paid. Seven of the 28 would have had inadequate cover. Perhaps the individuals bought a product that did not have certain features or perhaps they chose not to include enhancements to the product features. Maybe they would not have included cover for personal belongings or accidental damage.
For the remaining six, we would have declined those claims for a variety of reasons—or those claims would not have proceeded for a variety of reasons. It could be that the claim level was going to be below the excess, and the customer might decide, “Well, it is just not worth it because I am going to have to pay an excess”. It could have been due to faulty workmanship or a breakdown or, again, some kind of wear and tear. Or the customers may just have decided not to proceed with the claim. That is the breakdown for us. On the travel side, by comparison, we would accept 84 out of 100 claims. Of the remaining 16 in travel, the claim may not have proceeded for a number of reasons, such as there being a “no travel” government advisory or consequences that led to the cancellation of flights, or somebody was able to get another flight within a 12-hour period and was not stuck for a longer period. In some instances, with what is going on in the world, people just decide that they do not feel safe or they change their mind about travelling to a certain region. When everything happened in the Middle East earlier this year, with Iran, we saw a spike. We had 12,000 claims calls in a very short space of time. A number of those were calls where people just wanted a bit of advice. They saw what was happening on the news, they were looking at government advisories and they just wanted to talk to us about the implications if they decided to change their travel plans, but many of them did not.
Nick Turner: Across all products, our claims level is 96.7% claims accepted. On home policies, that is split between contents and buildings, and buildings is where there is more complexity. Our claims level would be around 85% accepted and, for contents, this would be higher than 96.7%. You can see why: it is much clearer to settle a claim on a contents policy than it is on a buildings policy. There will be a big range among insurers anyway, because it will depend on the kind of building stock that you insure. Imagine a rural insurer: we have many older properties and, in that scenario, wear and tear is definitely a feature. The majority of our disputes come about when people think that we should be paying a claim and it comes down to working through what is wear and tear and what is not: did the storm cause the problem, or did you need to fix your roof several years ago and not get round to it, and you are now leaning into an insurance policy to get that that repair done? That is not really the purpose of insurance, and we try to make that clear, as indeed does the entire industry. People understandably have selective memory when they are faced with a financial crisis. That is where we are. Our travel sector is small enough so as to be statistically insignificant.
Lord Turnbull: Just on a point of clarification, you have all three quoted quite high rates of acceptance: 72%, 84% and 96%. Is there a difference between a claim you pay straightaway and a claim that you refuse, followed by a bit of a battle that could go to the Financial Ombudsman, and then eventually you accept all or part of it? Is this the total of claims that you pay out on or the claims that you meet without the customer having to fight for it?
Alistair Hargreaves: This would be the total number. It would therefore include everything that we are contacted about and everything that we pay. In terms of things that were not paid, any which way they were not paid, these would be captured within that number.
Lord Turnbull: Where do you score it if the claim is refused or challenged initially, but then you eventually either change your mind or the ombudsman requires you to change your mind?
Alistair Hargreaves: If we paid the claim, I think that would be captured within the paid claims number. That would be a very small number of claims that would involve the sort of dispute that you are talking about. It might be something that would be investigated from a fraud perspective if we were worried about the risk of a fraudulent claim and we were investigating that. If we then found that was not the case and we paid the claim, then we would include it in the number. Our job is obviously about looking after the customer at the point of claim. It is also about making sure that, if there is a claim on the policy that is not a misunderstood query, we are there to look after those customers and pay those claims out.
Q95 Lord Turnbull: “Declinature” rate: I like this new word—it ought to go into the dictionary. I turn to customer duty. There is a long chain of people involved: customers, brokers, companies like you—let us call you the principals. You then subcontract various activities, most notably claims management. Does the customer duty really percolate down right to the coalface, where the real struggle is going on, or are you having lots of away days, seminars, board meetings and discussions with the regulator at the corporate level?
Jason Storah: I would say that the consumer duty obligations that we hold remain regardless of whether there is a third party involved in settling the claim. The dialogue that we have and the data that we share with the regulator holds us accountable to that. We use third parties to help to give us more coverage. Outsourcing claims to other organisations can involve more costs for us, or it can sometimes involve cost savings. We apply the same rigour to the completion, the customer journeys, the outcomes and all the data and metrics that we look at, irrespective of whether an external party has been involved in settling a claim or whether it has all been done in-house. This is particularly true for home claims: that is, buildings, people’s apartments, flats and houses. These are quite complicated. With more complex claims, it is not untypical to have anywhere from three to five different organisations, different companies and skill sets involved in assessing what has happened and assessing what remediation is required. Ultimately, however, there is somebody in Aviva who holds the consumer duty responsibility to follow that claim all the way through and track the customer outcomes.
Lord Turnbull: Earlier witnesses have referred to the consumer duty as a work in progress. It is three years on now. When does a work in progress become a point whereby the transition to a new relationship should have been completed?
Nick Turner: I could talk about our perspective on consumer duty, if that is helpful. As an organisation we are very supportive of consumer duty. The general transition towards outcome-based regulation is a positive and we are obviously in a transition towards that. We have lots of rules that the regulator looks at and examines where there is any streamlining that it can do to lean into the outcome regulations. If that is the transition state that you are describing, that will take some time to work through. I am supportive of that because, if you remove rules or streamline rules too fast, it creates a friction of work. However, I think that leaning into the outcomes-based model is good for the industry. For us, it was not a very significant piece of work; it was a relatively modest cost. I am not sure that is the case for everybody, but it was a very low cost for us. The effort was largely in bringing together the data that we already had into one single reporting environment to ensure that we could demonstrate that we were complying with consumer duty. Our mutuality helps because we focus on members: we have one customer to serve, if you like, and they own us as well. That also focuses the mind somewhat, which is perhaps why we found it a relatively straightforward exercise to demonstrate our compliance.
Lord Turnbull: Do you have a different view?
Jason Storah: I would say that consumer duty cost Aviva quite a lot of money. This is possibly because of our size and scale and the number of people who we insure across the UK, and the breadth of propositions, brands, price points and coverages. It is well embedded, though. It is not something that still has a long way to run in terms of its impact on consumers and its embeddedness in companies. The visibility that has been achieved around customer metrics at the board level, and all the way through the organisation, is very good. The type of metrics we look at, the type of dialogue we have, the clarity on what we are doing and its impact on customers is very well embedded in the business.
Alistair Hargreaves: The only thing I would add is that a consumer duty is principle-based and outcome-based, and it is about looking at the outcomes for customers. We are talking about things such as claims acceptance rate and customer understanding. Part of what consumer duty is designed to do is drive for continuous improvement across the industry. It is very well embedded, but it is still highlighting areas that we as an industry are working hard to continue to improve and help customers.
The Chair: Can you point to any areas where there is visible improvement? I think the comment has generally been made that people have not yet seen any visible change from the impact of the consumer duty.
Alistair Hargreaves: If I talk about consumer understanding, and some of the things that we have done as an organisation, we have talked about the fact that some claims get declined because of excesses. I use this as an example. We make sure that this is very clear at point of sale on the journey, and a customer is often actively selecting it. We reinforce those. For some of our excess levels, we have reduced the level of excess, which increases the claims coming through. This is us looking at some of that data and saying, “What’s happening and what are the things that we think we can do to improve it?” I would argue that it is a push that we had as an organisation before consumer duty, but consumer duty encourages the industry as a whole to look at things in that way and continue to drive for continuous improvement.
Jason Storah: To build on that, if I think about the desire to make sure that there is a breadth of product offering and price points for people, we and a number of other companies offer essentials-type products, so it is coverage that is a bit more stripped out, with lower premiums. But we also look at the data and the behaviour of those customers. I can give you a couple of statistics. If you think about the customer journey that people go when they buy on price comparison websites and they have a number of options, 60% of the people who click on an essentials product—so a more bare-bones type of insurance product—subsequently choose a higher-coverage product. I think consumer duty has encouraged and increased people’s understanding of what they are getting rather than just comparing one price against another. Price comparison websites have definitely played a role in helping people understand the different price points.
Another thing we see is that at renewal, 13% of our customers who have an essentials product renew with additional coverage. So again, these are all consequences, either direct or indirect, of consumer duty, which mean that customers are making more informed choices and are learning as they go through buying insurance products.
Q96 Baroness Bowles of Berkhamsted: I would quite like to follow on from the comparison websites point. We have been told that price competition has led to a hollowing out of insurance products so that they can score in the cheapest on offer and that that will attract some consumers. In response to that, it is sometimes suggested that there should be some minimum standards for products to ensure a baseline across the industry, so that would be a form of standardisation. What do you think about that, and what other issues do you think it might raise?
Nick Turner: NFU Mutual is not on price comparison websites.
Jason Storah: I will jump in, then.
Baroness Bowles of Berkhamsted: You can still have an opinion.
Nick Turner: Yes, I have an opinion. Standardisation of products is not a good thing, but standardisation of reporting might well be. The FCA is already working, as are we with a number of other insurers through the ABI and whatnot, to try to improve the consistency of reporting to enable consumers to better compare some of their products and services.
That is a good thing but, for us, price comparison websites are not an environment that properly reflects the value proposition that we have built for our members. If that were to change, that would be a distribution channel that we could consider, but we distribute through 300 agents that are all NFU Mutual-branded. People can walk into those shops, for want of a better word, sit down and discuss their insurance properly, and understand the value of the different component parts that make up an insurance product. They can talk about the claims service, what happens when your house burns down and what alternative accommodation means—all those kind of things—so that you walk away understanding the risk that you are running as an individual, but also matching the insurance and the service better to what you are buying. We believe it is difficult to truly get a comparison of service quality and cover value through a price comparison website in its current form, so we are not on them.
Jason Storah: Some 75% of our sales come through price comparison websites, so obviously we have quite a different business model to NFU. I do not think that product standardisation is the answer, as my colleague was saying. Comparability, being able to compare what is in different products, is really important. If I have five different price points, that is sort of interesting, but if I understand why those prices are different because there are different levels of coverage, that is really meaningful.
Minimum product levels are also quite difficult. You referenced the hollowing out of products. We have seen some of that, and paring back of coverages, but I argue that having some cover for home insurance, for instance, or for adverse events such as a flood, a fire, a theft or subsidence, is better than no coverage. What you see with minimum products—essentials products—is that they do not offer the frills. You will get covered for a flood or a fire but you might not get cover if you stain your carpet. Different people’s life circumstances and situations mean that they have different needs. Focusing on the minimum product requirement and standardising that feels like the lowest common denominator rather than giving people, as you were saying, the ability to compare apples to apples across different products, brands and companies.
Alistair Hargreaves: We also use price comparison. We provide products for a wide range of customers. For our customer base, affordability is really important, which is why they go on price comparison sites to make sure that they are getting great value for money. If I look at our product suite, for most of our products, it is about strengthening, so we have increased limits for things like alternative accommodation, personal belongings, contents, keys and locks, and things like that. I do not recognise the hollowing out. It is really about a very competitive UK market, and then consumer duty holds us accountable for making sure that we live up to that promise. That combination works well.
On the essentials products, one thing I would call out is that we launched an essentials product for motor, but if you cast your mind back to 2023-24, one of the big concerns was about motor affordability. So it was about ways in which we can provide motor insurance more cost effectively, and essentials was a slimmed-down version of that. But, for example, essentials cover does not include windscreen, so we actively ask the customer, “Do you need windscreen?” and get a yes or no, so that they know what they are buying and it still fulfils their needs. Price comparison, combined with that consumer focus and the accountability on all insurers through consumer duty, can work well.
Baroness Bowles of Berkhamsted: Would you say that if things were standardised, it would interfere with competition?
Alistair Hargreaves: There is a risk that it oversimplifies. A standardised product could be too much cover for some customers and not enough for others, and it might reduce the obligation to say, “Is my product fit for this customer?”, which is what consumer duty asks us to do by looking at the outcomes and the data, and then to drive that continuous improvement. I agree with my colleagues; although standardisation sounds good, I am not sure it is a quick fix.
Jason Storah: It could also lead to reduced competition and innovation, as well as reduced personalisation. If you look forward and think about AI, for instance, it is giving people the ability to get really hyperpersonalised advice. To suddenly take a step back and say, “We should standardise product sets”, would feel as if it was going against the grain—going to Alistair’s points—of the value of having fit-for-purpose insurance for every individual but then also the value of having coverages that are tailored to people’s needs.
Baroness Bowles of Berkhamsted: So the drive is really in the opposite direction, for more tailoring.
Jason Storah: As well as the comparability—the ability to be able to compare apples to apples. Just looking at price alone is very misleading; you really need to understand the cover beneath that to understand what you are getting. There are also Defaqto reviews, and other information is available out there for consumers to try to compare different products and different companies.
Baroness Bowles of Berkhamsted: Do you want to add anything or has it all been said?
Nick Turner: No, I think it has all been said.
Q97 Lord Griffiths of Fforestfach: Before I ask the question, can I just ask you how you define standardisation? For example, if I go into a store to buy food, I look at how much fat and salt is in it—it is very clear. Instinctively, I feel that standardisation will add to cost and probably reduce competition and your ability to compete with each other, so I am not really sympathetic to it. On the other hand, when I go into M&S or Waitrose, I value it. I just wonder how you would see standardisation in relation to, say, food?
Jason Storah: With standard excess levels—standard levels of coverage for certain perils—you could provide that, and there will be quite a lot of overlap between what different companies offer in their insurance products. But when it comes to setting standards that are rigid and from which there is no flexibility, I agree with you entirely: it would reduce competition and innovation. To my colleague’s point, it would mean that you would create a vanilla average for everybody, and not everybody is vanilla—everybody is a different flavour.
Lord Griffiths of Fforestfach: What did you make of the Which? supercomplaint?
Alistair Hargreaves: We have always been customer focused and strive to improve things for customers. Our customer promise is about giving customers great value, ease of use of insurance and being trusted at point of claim. I think that Which? highlighted that we have a problem with trust with consumers, and the biggest aspect of that was about consumer understanding. The points that you have been focused on so far in the committee are that these products, home and travel in particular, are more complicated than something like motor and that as an industry, we need to face into that and things such as wear and tear.
I say it is consumer understanding because our travel insurance products are broadly in line with the market. It costs on average for us a bit less than £300 to cover your home and contents for unforeseen events. Some of those claims can be expensive if there is a storm et cetera. But in terms of maintenance of your home, checking your roof and things like that are the sorts of things that, if you look at the guidance, should be sort of at least annual. Maintenance of your home and things like that cannot be covered by a £300 policy. That is not its purpose. That said, we need to explain that well to consumers. I think some of the claims acceptance are queries, and we should not be putting customers off at point of claim to understand. It is understandable that they would reach out to us as experts and ask about it. But within that, I still think that, as an industry, we need to work harder because at the moment it is damaging that trust in us as an industry.
Nick Turner: We have a good relationship with Which? We won its insurance brand of the year award for the last four years, which puts us in a reasonably good place. Whether the learnings come from FCA good and bad practice, which I would encourage more of, or whether it comes from a supercomplaint, our first go-to place is to look at all the things that people are coming up as issues and do a gap analysis with our own operation. We found no significant changes that we needed to make on the back of the supercomplaint.
Jason Storah: I agree. Which? plays an important role. We had already done an awful lot of work and improvements, and the FCA had looked at those. The one thing I will say is that Which? plays an important role but when you look at the basis of the data it used for the supercomplaint, it surveyed 3,300 claims and spoke to 24 people. That is not a big sample size. We settle 3,500 claims every two days. I have personally spoken to and met probably over 100 customers in the last couple of years through customer insight and feedback sessions that we do. So, there is always validity to somebody finding some data and putting it out there but the validity and size of that data set is also worth considering sometimes.
The one other thing is that I worry a little about broad-brush statements and some of those headlines that get attention when actually there is a lot of texture beneath them. As to any headlines that perhaps call into question the value of the insurance industry and the value that insurance companies provide, we all pay thousands of claims every day. That tends to get forgotten. There are definitely always areas for improvement but the vast majority of people make a claim and get it paid. We all see examples of high claim satisfaction scores. Maintaining that balance sometimes is important.
Q98 Lord Lilley: Is there any change you would like to see in the regulatory regime you have to adhere to that would make life simpler and less costly to you without appreciable loss in quality to your customers?
Jason Storah: The regulation that the insurance industry has worked through over the last number of years has been significant. One of the issues and parts of dialogue over recent months in the last year or so is the significant volume of regulatory work on companies. If the Government and companies were aligned around wanting to reduce the level of regulation overall, that is challenging when you have principles-based regulation with outcomes-based regulation and things get layered on top of each other. But the FCA takes its responsibilities seriously. Personally, we would sometimes rather the regulator went after bad actors or individual instances of outliers rather than take something and then try to do an across-the-board piece of work or data request. It has certainly implemented consumer duty and all the other components of the regulation thoroughly but we would definitely like to see a bit of prioritisation in the ongoing regulatory burden going forward.
Nick Turner: I echo some of those comments because obviously everybody must comply with regulations, whether you are the target of the problem or not. So, if you are operating in a consumer-friendly way, you can have a cost of implementing a regulation that does nothing different to the way in which you are currently dealing with your customers. It just becomes a cost that you then have to pass on to members, and that is painful. Again, we have the consumer duty, which should give the regulator all the power it needs to go after the bad actors, as Jason said. It does so increasingly, by the way. So, rather than increase the number of regulations going forward, use the powers that you have to tackle problems that you have in the market. That was one of the drivers for consumer duty—to give the regulator the power to do that with a bit more energy and gusto.
Alistair Hargreaves: I agree with what has been said. Consumer duty gives the right framework. The FCA talks about predictability and proportionality. It is about using that going forward and being mindful of the cost of large data requests, some of that activity and prioritising it, as people have said.
Lord Lilley: If, on reflection after this meeting, there are concrete changes or you think of something has been inherited from before the consumer duty but is still there and is now unnecessary, and could be removed, perhaps you would let us know.
Q99 Lord Davies of Brixton: Did you watch our session last week when we spoke to the price comparison websites, or have a briefing?
Jason Storah: We saw few headlines, yes.
Lord Davies of Brixton: Did you pick up anything that was said that you would want to respond or add to? It is a bit of an unfair question. I did not quite understand on the NFU. As you say, you always do well in the Which? reviews but you say you do not participate in the price-comparison websites. I do not know about other members but I got the impression that there was fairly universal coverage but the price-comparison websites just do not report on your policies.
Nick Turner: Think of it as like a supermarket. You can choose to put your products on their shelves or not. We choose not to.
Lord Davies of Brixton: To the other two—you choose to go on. The price-comparison websites raised the issue of claims rates information. They say they want it but there is a reluctance in the industry to provide it. What is your take on that?
Jason Storah: The FCA collects claims data and such data is available. The FCA publishes it. The price-comparison websites could take that. Certainly, we are very open to providing claims data and it being published. We just want to make sure it is consistent across all companies. When there are differences, you can create uncertainty and distrust. It is just because the data is slightly different. So, to put that in the hands of price-comparison websites would not feel like necessarily the right idea. The best solution is that the FCA collects that data. It would validate that it is apples-to-apples data and that different interpretations of the data are not reported. It then publishes that, and price-comparison websites could use that as they as they saw fit.
Alistair Hargreaves: I echo that. We have talked a lot about claims acceptance. There has been an acknowledgement that that is not necessarily consistent across the piece, which demonstrates the challenge to that. The other point is that consumer understanding is about understanding what the consumer does not understand and making sure that they do. Insurance is quite a complicated product to a lot of people. We do a lot of testing of that consumer understanding.
As I mentioned, we are also looking at what is causing the confusion and targeting it throughout the life cycle of our customers to ensure that we are giving good customer outcomes. So, on top of the comparability, I worry slightly about giving out more data and whether that confuses customers. We work with the price comparison sites and we are actively engaging with them to find ways in which we can get that comparability but, for both the underlying data and how you communicate it to customers, it is not a quick fix. It takes time to do it well.
Alistair Hargreaves: I would echo that.
Lord Davies of Brixton: But are those of you who engage with the price comparison websites trying to agree a standard basis?
Alistair Hargreaves: Yes, absolutely.
Jason Storah: It bears some reinforcement that the texture of reported data is important. I will give you an example: time to settle a claim is a stat that can be really meaningful for certain kinds of claims and misleading for other kinds of claims. If somebody has a simple stain on a carpet, and the piece of carpet needs replacing and is covered, how quickly that gets done, as it is a pretty simple claim, can be a meaningful data point. If somebody has a catastrophic flood or a fire or has a subsidence issue with tree roots causing problems with the foundations, the time to settle that claim can be quite significant.
You just want to make sure that the way that reported data is used is apples to apples. That is why we think about going through the FCA and using that data. It is being collected for understanding and, ultimately, customer regulation and customer benefit; the primary use of it is not commercial benefit.
Lord Davies of Brixton: So there is a slight difference in emphasis between you and the price comparison websites about resolving this issue, but you are agreed that the market would be better if that information were available.
Jason Storah: Yes, but some of that difference might be because we operate in slightly different parts of the insurance ecosystem. We have different business models, so it is probably not a surprise that there are some differences.
Lord Davies of Brixton: Another issue that was discussed was that you can go on the website on a Thursday and get a price, but you may go back the following Monday when there is a different price. I have always felt that underwriting is perhaps a bit more of an art than a science. The explanation we had was that, at any given point in time, you are looking for different sorts of business. Would you like to expand on that point to explain this to us?
Another aspect of that is I that I may get a renewal premium and speak to the person on the helpline. They say, “I’ll check if I can make you a better offer”, then they come back and they make you a better offer. You made a comparison with a supermarket: if I go to Sainsbury’s, I know the price and the price is taken. Occasionally, there are special offers, but they are there as special offers, whereas insurance companies seem to be making it up as they go along.
Jason Storah: Pricing does change on a regular basis. In your example, though, somebody is making the effort to phone up. Sometimes people get a quote, then go back a few days later and go through it, but they might enter slightly different information or use a different quoting engine. If somebody has been given a quote that they have completed online, but they phone up and give additional information, that additional information can change the price quite frequently: it can sometimes increase the price and it can sometimes reduce it.
There are other factors that an agent can take over the phone that they cannot necessarily take through a rigid 10-question set, but they could understand other products that a customer has and other variables that would factor into the price that the company feels is appropriate for the risk. Again, it is a bit more nuanced than just saying that you can pick up the phone and the price gets changed. There is a lot more to it than that.
Alistair Hargreaves: We are trying to cover customers for the risk of them potentially needing to be looked after at point of claim. We are trying to do that as affordably as we possibly can for that customer. Price comparison is very transparent: you want to be as cheap as you possibly can for that customer, but you need to make sure that you are covered enough across the piece, so that you can look after them well at point of claim. We try to use data and update that data regularly to understand the risk of that customer as well as we can. That is the principle.
For customer renewals, the general insurance pricing practice means that the price that we offer our renewal customers is the same or lower than they would get as new business. Again, it is a very competitive price; it is the same as they would get if they came to us as new business. But I think customers are quite used to negotiating in their general life so, if a customer calls us, maybe gives us some more data or has found a cheaper price elsewhere, we will give a discount to that customer. That would be a better price than they would have got as new business; we will not give the same discount to new business. It is a commercial decision, because that customer might have been with us for a number of years so we would think, “We would like them to stay longer”. It is something that we evaluate—we think, “Is it the right thing to do?”—but, on balance, we think that being able to discount customers who have been with us for a few years is appropriate, in those circumstances.
Lord Davies of Brixton: What you say suggests to me that there is a struggle there. Rather than “consumer understanding”, perhaps we could rephrase it “provider explanation and the lowest possible price”. There is a struggle there that you have to work your way around.
Nick Turner: Also, the number of data points that go into a price these days, particularly in personal lines, can be in the hundreds. Any one of those could change or be updated, either from an internal understanding of the risk or from people putting in different data points. Actually explaining that in such a way that someone can absorb the impact of something is a constant battle for insurers generally.
AI does help with some of this stuff, because it can surface the rationale for any price changes. We have it at renewals: people come up to renewal and say, “Why has my price gone up by this amount?”, or even, “Why has it gone down by this amount?” For some of the more mechanised pricing that is generated, there is actually quite a large amount of work to truly understand what has moved in a scenario to explain the price. That will always be a battle between consumers’ understanding and our ability to explain that in an economic way.
Q100 Lord Vaux of Harrowden: To follow up on that point, one of the things that we were told very forcefully last week by the price comparison websites was that, if you go to the price comparison websites, you will typically save about 20% off the renewal from your existing provider. Why is that? What is happening there? Does it not somewhat undermine what you have just told us about the process you have explained?
Alistair Hargreaves: We have changing prices and, at renewals, customers tend to notice when their price goes up, but some customers see their price go down. If you see your price go down when you look at your auto-renewal—that is the nature of the beast, the changing risk—you will think, “I’ve got a good deal this year” and, to be honest, will probably forget about it and move on. If your price goes up, you think, “Why has my price gone up?”
A lot of our customers will shop because they have come to us from price comparison. That is where our rating structure has changed and maybe somebody else’s rating structure had changed in a different way, and they can save money. The price comparison sites are looking at people who have probably decided to shop because their price has gone up, rather than looking at the whole market.
Jason Storah: Again, they might be able to get a cheaper price, but are they getting the same level of cover? How aware are they of the cover comparison? They are probably not as aware. But price comparison websites play a really important role and a big part of that is the choice that they offer and the ability for people to get options and reduced premium levels, in some instances. But, again, price is not the only thing that we want customers to look at; we want them to assess cover versus cover and the prices available.
Lord Vaux of Harrowden: On my main question, you talked about acceptance rates and said that probably the main reason to decline a claim is down to customer understanding of what they are actually covered for. You have also talked about the work that you are doing to try to improve that customer understanding. You have given us a snapshot of acceptance rates at a point in time, but what are the trends? If you are making all these efforts, one would expect to see a decline in the rate of declined claims—if that is the right way of putting it. What are the trends?
Alistair Hargreaves: We have seen an improvement in the 2025 results versus the 2024 results. As I say, we have been very focused on this for a number of years, and we have been working on a number of changes. We found that it is quite hard to engage with customers via communication alone. I am probably one of the few people in the UK who gets excited by insurance. Generally, our customers will be engaged at point of sale but, afterwards, they will not be thinking about it very much unless they need it. With a product like home insurance, where that wear and tear exists, a customer might want to query aspects of the cover with an expert with whom they have a relationship but we will have to accept that it will never get to the same stage as motor insurance. But there are things that we can do.
As I have said, we have gone beyond on our communications. We communicate at point of sale, we follow up with areas to reiterate what the customer is covered for and what they are not covered for on points that could cause confusion. One example would be if there is an event, such as a storm: we will reach out to customers and explain things again to them, because they might be interested at that point. I mentioned some of the excess levels; we have made changes to the product. We see that coming through to the results. As I said, though, I am not however expecting it to get to same level as motor insurance because of the nature of what we are dealing with.
Jason Storah: I would echo that. I think we probably all get excited about insurance. We have definitely seen an improvement. To be clear, we want there to be an improvement. We want the decline rates to get narrower. There is a point at which there is only so much we can do. If you take a macroeconomic step back and look at various factors—such as the cost of living crisis, where people are spending their money, what they have in terms of disposable income and the choices they are making—perhaps fixing the roof or making decisions on insurance products is one of those things that they are not necessarily going to be able to address. We have seen improvements, however, and we continue to work to make them.
Nick Turner: Without being complacent, I think we probably feel like we are close to the frontier of what is possible. With the building rates, we designed the claim service around what people need. It is almost an insurance policy worked backwards. You work out what you need when your house burns down or floods, and that designs the policy. The price is then a function of that. That is how our mindset works on this stuff. On the declinature rates, there is little that can be done: the issues involve fraud, wear and tear, and things beyond the policy scope. Helpfully, we also have a 94.1% persistency rate of renewals. People renew with us at that sort of level.
We also have the agency network of 600 agents and nearly 3,000 staff out in the rural community and we actively encourage people to come and talk to us, whether it is for a motor policy or a home insurance policy. We train the team in Mutual Direct in Glasgow and Cardiff very well, and do not limit their ability to speak to people, saying, “Talk as much as you want”—which is a differentiator for our business. We see that as an educational thing. Every year people come back to us, and we have conversations and reinforce the scope and range of the policies through those conversations, which is difficult to do through other distribution channels. If you have been with us for 25 years and we have had a conversation every single year about what you are covered for, the chances are that you know what you are covered for by the time you come to claim. It gives us an advantage in that regard, so I think the model works for us.
Lord Vaux of Harrowden: I have a very specific question on storm damage. We have heard about storms being defined by wind speed. As someone who has a house, you do not care what the wind speed is, just that your roof has been blown off. Does that need changing?
Alistair Hargreaves: There is a standard definition of a wind speed that defines a storm. That is something that we would use. One of the things that we have looked at here is the introduction of more than one measure. We make sure that we use the highest measure of those two, and we then give a 10% tolerance. If the customer is within 10% of that definition of a storm, that would meet our definition, and we can go ahead with the claim. If a customer disputes it, we will look at it and we can send people out to take a look and understand exactly what has happened. But we need a way of distinguishing between a storm and the wear and tear, and that is why the industry standard exists. We are keen, though, to ensure that we are on the side of the customer, hence we give that flexibility and ensure that we are interpreting that definition well.
Jason Storah: I totally agree; we have the same. We look at the industry standard and then provide a number of parameters around that to understand the impact, whether it be wear and tear or the event of the storm that has caused the loss. So we have a very similar approach.
Nick Turner: It is the same for us. We would try to get eyes on the problem and understand exactly what has happened from the consumer’s perspective. We focus on being very pragmatic in the grey areas, because there often are grey areas. If we can pay, we will. If we can pay using some of the other covers that are within the policy, we will. The only statistic I can give you is that last year in total we paid out £1.4 billion in claims, and we received five out of five on our CSAT scores 83% of times.
The Chair: Did you target improvements in your acceptance rates?
Alistair Hargreaves: Yes. We drive continuous improvement in terms of acceptance rates. This is something that we have looked at a great deal and we have seen improvements in 2025 versus 2024.
Jason Storah: Absolutely. We drive improvements in the acceptance rates through understanding the issues, the areas of a customer’s lack of clarity and where there are decline rates. We do that as a consequence of focusing on the feedback that we get from customers.
The Chair: They do seem to be quite stubborn.
Lord Vaux of Harrowden: Among the wider industry, do you share the sort of common reasons for declining, particularly where there are common misunderstandings? Is that something that you share between yourselves?
Alistair Hargreaves: We work with the ABI on common issues and we have, as I mentioned, an industry-defined definition of wind speed and how that works. This is to ensure standardisation of interpretation. There are property working groups that work together in order to drive improvements.
Q101 Baroness Donaghy: I want to ask you about AI but, before I do, would you say there has been any change in your relationship with the FCA over the last three or four years? Have some things improved? Is there room for improvement in other areas?
Jason Storah: We have probably got to know it better over the last three or four years. We have spent a lot of time with its staff and gone through a lot of data and dialogue. There is much dialogue with the FCA teams across a number of levels of the organisation. In terms of areas for improvement, as I said earlier, the only thing I would address is the level of regulation, namely the burden that comes on to us. You mentioned that you would reference AI: the implications of future regulation were explored in the recent Mills review. If that regulation is just layered on top of the existing regulatory burden on companies, that is probably not ideal from our perspective. We have already spent a lot of money and time on that. It is about managing the balance between regulation evolving versus what will fall away and what will be prioritised.
Nick Turner: Just to build on that, our relationship with the regulator has not changed. It is a good, respectful relationship and we engage with it fully openly. On the AI point, I think that Jason is bang on. We have a consumer duty which says that you are still accountable whatever methodology you use within your organisation to deliver great outcomes for members, in our case, and customers. Using AI as one of your component parts within the organisation does not let you off the hook. You cannot abdicate anything, or say that “the bots did it here” or “the algorithm did it there”. You are still entirely on the hook for it, which is perhaps one of the reasons why AI has not got into the regulated world as ubiquitously as it may have done in other industries, because we have to be very careful that we do not damage any consumer outcomes through its deployment. I am sure all of us are doing this in a very controlled way, and in a very cautious but proactive way, because it also has positives for customers.
Baroness Donaghy: Which? said in its evidence on the supercomplaint that there was a frustration with the lack of data, despite it spending months, if not years, trying to get that information. There was an implication that the FCA was not doing its job, but that is not your view.
Alistair Hargreaves: No—as we have been saying, we have close relationships with the FCA. Some of the data that we have been talking about, such as the GI value measures data, has been introduced as part of the consumer duty looking at outcomes and having data that is visible and that we can discuss. When the FCA looks at specific topics, it asks for a lot of data to dig into them.
Baroness Donaghy: On AI, one of you—I am sorry; I cannot remember who—talked about hyperpersonalised advice. What progress are your organisations making on AI, bearing in mind what Nick just said about cautious analysis and progress?
Nick Turner: I can give you a couple of examples of how we are using it. I would put AI into three buckets. There is the use of things like Copilot, which are just generally helpful. Like a Google search, you can ask questions of the AI and it will give you a shortcut answer or you can use prompts that create simple routines. We have rolled that out to all our staff in a controlled way. It is used in a controlled environment and it helps people to draft letters or reports, so that it cuts away some of the work that we have that would take longer, so that people are more productive, in effect. That is one way.
The second way is that we have done things like tipped our 6,000-page underwriting manual into a large language model that allows our agents to speak to a machine and get an answer on whether we will accept something or not, from an underwriting perspective. That then leads you to which page in that manual it is, so it is like a really engaging index that stops our agents speaking directly to experienced professional underwriters. They are asking questions such that they could look it up themselves. It helps educate them that way and creates a triage. That is just one example on productivity; I could give you others.
We could get more value from a piece of work that we are doing on processes. We are remapping all our processes with customer experience as the North Star. Once we have mapped those processes, we can look at where we can automate or improve some of the outcomes within those processes towards that agenda of customer experience. That is being worked on right now.
So there are three buckets. The one that I think will be most impactive for us is the process re-engineering, but you can test that to destruction and make sure it is delivering the right outcomes for customers before you roll it out fully. The human is in the loop on everything, across all of these. As an organisation, we see that as a differentiator, as you have probably picked up. Our people at the front will remain, so a lot of AI is focused more on productivity improvements in the back office. That is how we are using it, in a nutshell.
Jason Storah: I will maybe use a couple of similar examples. I am thinking about claims summarisation tools, which offer benefits both internally for our people and externally for our customers. During the life of a claim, somebody will phone up and tell us the first notice of loss: they have had a claim; there is water in their kitchen or there has been a fire or a theft. The claims agents would take a lot of information, give them a claim number, there would be follow-up and somebody might phone back a few days or a week later.
The claim summarisation tool, using AI, means that the person who the customer speaks to does not have to spend time summarising the claim. It is all done instantly. Next time that customer phones, they will invariably speak to a different claims agent, but that agent will instantly know everything that is going on—the most pertinent parts of the claim and the conversation. So that is a more effective, more efficient usage of time for our people and a quicker response for the customer. That is one example.
Another one is that Aviva was the first insurance company, or one of the first, to offer the ability for customers to use ChatGPT to get a quote. We did that not thinking that there would be hordes of people who would start getting quotes and buying their policies using ChatGPT; we did it because the insight you get from the customer journey is really interesting. I will give you a couple of examples. Of the people who get a quote online, 60% complete the quote online. With ChatGPT, over 90% of people were getting through to a quote, so it is a much higher throughput rate.
However, this is not the real value. If you think about a normal quote journey, customers answer a prescribed set of questions, particularly with a price comparison website. It is the same question set, then it pumps out a number of different quotes from different companies. But with ChatGPT—it was me who mentioned hyperpersonalisation—customers can be asking questions and giving information that is very relevant to them. Any two people would have very different perspectives on what is important to them in their insurance and what is most pertinent, as they may have had a claim or be worried about X in their house. So you get a much more individualised customer journey and the data and insight that you get are really interesting.
Alistair Hargreaves: They have not left me much to say, have they? We are similarly embracing AI and doing so in a controlled way. I will use a couple of slightly different examples, just to build on what has been said. Another use that we found for AI is quality assurance. We have a lot of transactions with customers, and AI enables us to look at all those transactions and then spot the exceptions in terms of any dissatisfaction, so it can be a driver of quality and feedback in that sense. Also, in terms of fraud, it enables you to look at a lot of information and identify inconsistencies, or, increasingly, people outside of the business have AI and so they might have photographs that have been manipulated or things like that. It enables us to identify that as well.
We think AI is a very powerful tool and will be powerful for us going forward, but we have controls around it. We have a data ethics team that is independent of the teams that are testing AI and which is saying, “Look, is this in line with consumer duty?” Although AI is a great opportunity to improve things for customers, we want to make sure that we are still accountable for the outcomes for those customers, whichever tools we are using. So we have those appropriate controls and checks and balances as we try those new user cases.
Baroness Donaghy: You have all mentioned the importance of keeping in mind humans. I will not bore you with the argument I had with a bot once, but yesterday it took me half a morning to get to speak to a human being. There is another side of this: of people in society feeling disenfranchised and alienated from the system. I am concerned that this will all add to that particular feeling. What measures are you putting in place within the regulatory regime to ensure that does not happen?
Alistair Hargreaves: On a principled basis, we want to provide great service for the customer. Where AI can do that, that is great, because it should be that customers are going there because it is easier and better for them. That would be our objective, but we would still want to give choice of channel. In the same way that digital changed things a lot for customers but there are still customers who want to speak to somebody, we like to give customers that choice of channel. That would be our starting point as an organisation. If you look at it through the lens of consumer duty, consumer support and things like that, we would frown on anything that we thought was blocking that access to customers. Customers should have that choice.
Nick Turner: Our business model is built around human interaction. People who want a human service ring or talk to us, or walk into one of our shops, but I completely understand that not everyone does. We are in the process of having full multichannel, but it will still remain a people-based business.
Jason Storah: I totally agree. We will always have human oversight assessment of journeys and experiences, whichever channel they come through. From a claims perspective, we have a baseline claims service and people who respond to claims, no matter which channel people come in to us from, no matter what product they have bought, what coverage they have bought, or what Aviva brand they have used to come into the company. When we have people who are augmenting their skills and their productivity with AI, they are still very much involved in the process. But I am sure we have all had bot-type experiences like the one you mentioned.
Q102 Lord Vaux of Harrowden: Jason, you talked about consumers using AI to seek out policies and quotes, and you can imagine a world where ChatGPT, Google, et cetera, end up bypassing the price comparison websites. Price comparison websites are subject to regulation and consumer duty; Google, ChatGPT, et cetera, are not. What are your thoughts on that and how should that be regulated?
Jason Storah: The only way to buy insurance is through a regulated company at the moment, and I do not know that anybody would want to change that. We would not want that to change, and I do not think the regulator would want it. There are two components to it. The first is understanding the kind of AI and the impact on the customer journey that different regulated firms are using. We are all regulated and we are all embedding AI into different parts of our business—I do not think there is a concern there. However, consumers going outside of that regulated envelope is what you are flagging up and you are concerned about. We will certainly identify any of those; we do not want that sort of Wild West behaviour in the industry, so we would flag those, but I also believe that the FCA would be aware and other consumer groups would flag those kinds of experiences.
Q103 Lord Davies of Brixton: Just to be clear, if a customer comes to you direct, you do not interrogate why they have come to you.
Jason Storah: We do not interrogate.
Lord Davies of Brixton: They might have come to you because they have received some duff advice from ChatGPT, Claude or something else.
Jason Storah: Maybe, but then they will go through our quote journey and we will get the information directly from them.
Lord Davies of Brixton: But I presume you are not saying, “Well, you really ought to go to somebody else to get your insurance because we are not the people for you”?
Jason Storah: In the case of Aviva, our home premiums range from £80 a year to over £1 million a year, so we have a pretty broad appetite and we think we can serve everybody’s needs. But no, that is not an outcome that we foresee.
The Chair: Thank you very much, Mr Hargreaves, Mr Storah and Mr Turner—this has been a fascinating morning. Thank you very much for your time and for answering our questions in the way that you have, which has been very encouraging. That completes this evidence session.
