The claim after the storm

Executive Summary

In February 2025, we asked consumers with buildings insurance how they defined ‘flood’ and ‘storm’. Most of those surveyed thought that a flood was when water builds up in your home, no matter how fast this happens; most thought that rain, hail and snowstorms can happen without storm force winds. The Association of British Insurers (ABI), the government’s Flood Re scheme and guidance from the Financial Ombudsman Service (FOS) agree. As this report shows, many insurers do not, leaving consumers at risk of having claims unexpectedly rejected when their homes are damaged.

In this report, Which? presents a novel analysis of storm and flood definitions in home insurance policies, examining 133 recent policy documents from 67 providers. We found that one in three contain flood definitions and one in five contain storm definitions that contradict what we found to be most consumers’ common sense expectations and industry guidance. We believe each of these policies is potentially unfair to consumers. We also know from published FOS decisions that the application of similar definitions has caused avoidable distress and financial harm, and that where flood and storm are not defined, firms have applied definitions that go against most consumers’ expectations. We found that more than half and more than a third of documents reviewed did not define flood and storm respectively.

Under the Consumer Duty, the FCA expects firms to consider the reasonable expectations of customers in their target market in relation to the product or service. The FCA also requires manufacturers and distributors of insurance products to keep this under regular review to prevent foreseeable harm. We think people’s expectations for coverage in the case of floods and storms are reasonable given that they are broadly in line with the definitions of the UK’s largest trade body for insurers, the ABI, and the government’s Flood Re scheme. That many products adhere to these definitions proves that these are insurable risks. 

The findings outlined in this report clearly show that consumers’ reasonable expectations are not being met by a significant number of firms in the UK market, which includes some of the largest providers by market share. They also suggest that flood and storm definitions could partly explain the persistently high rates of rejected claims for buildings insurance, which firms do not appear to be sufficiently addressing by reviewing their products and claims-handling practices. 

We want the FCA to investigate all firms’ practices as part of its ongoing review of firms’ claims-handling arrangements and take enforcement action where necessary. The review’s broad approach of investigating whether ‘systems, controls, governance and oversight structures drive good consumer outcomes’ should enable the FCA to look at the root causes of high rates of rejected claims for home insurance, including whether firms have met requirements under the Consumer Duty regarding the design of their products. 

The FCA’s work should include considering the nature of the market and whether new guidance is needed for firms. Since it’s very difficult for consumers to compare definitions of terms between providers, we think competition is unlikely to be effective here in guarding against poor outcomes for consumers. Given widespread agreement among consumers of what reasonably constitutes a flood or a storm, and that this understanding is broadly in line with industry and government-backed guidance as well as the definitions used by many firms, the FCA should be clear that it expects firms to at least meet this definition. We are also concerned that where firms do not define flood or storm, consumers may receive worse outcomes than if a fair definition was included in the policy terms. The FCA should consider how this is impacting consumers and whether firms should define terms such as these in their policy terms. 

Recommendations

Based on the findings detailed in this report, we recommend the following measures to address the harm which can result from potentially unfair flood and storm definitions:

• Recommendation 1: The FCA should investigate whether all firms’ flood and storm definitions comply with the Consumer Duty as part of its ongoing claims-handling review and take enforcement action where necessary. This should include investigating whether firms have:
  • evidenced how they have met the reasonable expectations of their target market in any flood and storm definitions 
  • reviewed rejected claims for floods and storms to understand how to better align their approaches to their customers’ reasonable expectations
  • addressed any identified issues with product design.
• Recommendation 2: The FCA should make clear to firms, such as via issuing new guidance, that using the ABI and Flood Re definitions of flood and storm is one way for them to ensure they are currently meeting the Consumer Duty, with the FCA keeping these definitions under review. 
• Recommendation 3: The FCA should review consumer outcomes from claims involving policies that do not define flood or storm, and consider requiring firms to define these terms and any others where doing so may better prevent harm from arising. 

Introduction

Policy background

Something is wrong with the UK home insurance market. The FCA’s most recent value measures data showed that 37% of claims made on buildings insurance were denied, and 14% of claims led to a consumer making a complaint. That so many consumers are having their home insurance claims rejected suggests a growing chasm between consumers’ expectations of cover, and the policies they are being sold. 

This gap in expectations causes real harm to people dealing with damaged homes and shouldering unexpected costs, who find themselves without the safety net of insurance. We know that, where consumers have been significantly impacted by the events that led to their claim, insurers’ actions make a difference – when we surveyed 3,300 people in 2024 to understand their experiences of the insurance claims process, almost half (44%) of people surveyed who were severely impacted by their incident said their insurer’s actions negatively impacted their mental health.

Which?’s 2025 study of how consumers purchase insurance showed that one driver of rejected claims may be common consumer misunderstandings about insurance, exacerbated through the ways insurance is sold. Many people do not understand the limits of their insurance cover and often broadly assume that anything that isn’t ‘their fault’ will be covered. In some cases, this leads to claims being rejected where no insurer would have offered cover, for example, due to moral hazard around wear and tear. There is, however, another possible point of failure here, resting in the design of insurance products for risks that are insurable. Where policies have been worded in a way which doesn’t meet consumers' reasonable expectations, consumers are likely to be surprised to find their claim rejected.

As the UK's climate gets wetter, increasing the risk of storm and flood-related structural damage, home insurance will become more crucial to consumers than ever. Met Office projections suggest that increasing periods of heavy, prolonged rainfall will worsen surface water flooding for urban areas. By the mid-century, the number of properties in England at risk of natural flooding is predicted to rise from 6.3 million to 8 million – that's one in four properties. Whether water enters through overwhelmed gutters after heavy rain or through floors from rising groundwater, homeowners will need to know that they can rely on their insurer to be there for them.

In this report, we present an investigation into the terms contained in UK home insurance policies, using AI combined with expert human review to investigate terms in 133 policy documents that our new survey findings show are likely to contradict consumers’ reasonable expectations of coverage. This work focuses on how insurers define two terms: ‘flood’ and ‘storm’ – not because these are the only areas where we believe insurance terms may be unfair, but because they represent cases where clear, fair definitions, agreed by industry bodies and government, are readily available. They are also events which can cause lasting harm to people – harm which is likely to become more commonplace in future.

Methodology

Developing a test for fair flood and storm definitions

Tests to establish a fair definition of flood and storm were informed by four phases of work:

1. Internal exploratory sessions, drawing on Which?’s expertise in insurance to establish areas of potential unfairness to investigate further.
2. A review of evidence that policy terms in these areas had caused harm to consumers in the real world. This review led to a decision to prioritise floods and storms.
3. A review of expert, industry, regulatory and ombudsman guidance on defining flood and storm.
4. A nationally representative survey of 2,103 UK adults to establish consumer expectations around what constitutes flood and storm, and the coverage they’d expect their home insurers to provide in these situations.

Measuring the prevalence of potentially unfair definitions

Having established a clear method for testing fairness in flood and storm definitions, we applied this to home insurance policies across the UK market. To do this, we: 

1. Established a broad collection of online insurance related documents. 
2. Built a filtering system to remove everything from this dataset which wasn’t a policy wording document for standard UK homeowners’ buildings insurance.
3. Analysed this collection of policy documents for missing insurers, and obtained documents directly from insurers where gaps were found.

This left us with a final dataset of 133 home insurance policies from 67 different distributors, either dated since 2023, or sent to us in response to a recent query. 

This dataset is not exhaustive, and does not cover every insurance policy currently available in the UK. In the absence of a requirement for insurance firms to publish their policies online, and without legal powers to compel firms to share documents, assembling a complete dataset here was always going to be beyond Which?’s ability. After extensive manual checking, including ensuring that major providers were included, we believe that this dataset gives us a comprehensive view of the insurance market in the UK, and that our findings here are likely to give a faithful picture of the nation’s home insurance market.

Extracting and labelling flood and storm definitions

For each document in our dataset, we used a large language model to extract the policy’s definition of both flood and storm, where these were present (for more detail, see Annex A in this report). Each definition was then independently coded by four expert human coders to establish whether it passed our test for fairness, with disagreements discussed and reconciled by the team.

While AI was used to extract definitions, it was not used at any point to test whether a document was likely to be fair; this decision was always made through the agreement of multiple human analysts. These decisions form the basis of our analysis of unfair storm and flood definitions, below.

Full details of our survey methods, our collection and filtering process, and the approach used to extract definitions from documents are included in Annex A of the report.

This report

The rest of this report is presented in three sections:

Defining flood and storm fairly explores how we developed our tests for fairness in flood and storm definitions, including evidence that poor definitions can harm consumers, a summary of industry’s stance, and detailed findings from our survey to establish consumer expectations around coverage.

Measuring the prevalence of potentially unfair definitions in current UK insurance policies details what we found when we applied these tests in practice, discusses the proportion of reviewed UK home insurance policies that contain potentially unfair flood and storm policies, explores what makes them unfair, and highlights cases of poor practice by some providers.

Finally, we present recommendations to the FCA, designed to address the issues uncovered in this report.

Chapter 1: Defining flood and storm fairly

As discussed above, data published by the FCA on levels of complaints and claims accepted suggest a fundamental disconnect between what consumers expect their buildings insurance to cover them for, and what insurers cover in fact. Previous work by Which? identified consumer misunderstanding of insurance which isn't corrected during the sales process. This report builds on this by investigating whether the design of home insurance policies could further explain persistently high claims rejection rates. 

To do this, we first established where we should look in order to uncover areas which could cause harm, and then worked out how we should test for unfair practices in these areas. This happened in four phases. Exploratory analysis of potentially harmful terms, followed by a review of evidence of harm in the real world, led us to a focus on flood and storm definitions. We then looked at how industry bodies define flood and storm, and conducted a nationally representative survey to establish consumer expectations around these events. 

All of this work informed our final test for fairness in flood and storm definitions, presented at the end of this chapter.

Phase 1: Exploratory analysis of potentially harmful terms

Our first aim was to identify insurance terms where we felt consumer understanding was likely to be at odds with coverage provided by insurers. To this end, the research team held a series of exploratory sessions with Which? experts in insurance, financial policy, law and consumer research. 

In the first session, this group collectively analysed a sample of home insurance policy documents, a process involving much argument and annotation. Experts highlighted areas of concern, and flagged anything missing we should look for. In the second, Which?’s data science team presented the group with a few dozen policy definitions in each of these areas, and asked them to code each for clarity and fairness, before discussing the results. Both sessions were designed to tap into our experts' deep knowledge of the challenges faced by consumers in this market, while using sample policies to stay grounded in terms as they actually exist. 

The result was a shortlist of areas to consider which, alongside definitions of flood and storm, included questions around coverage for scorching and warping caused by extreme heat, and terms concerning information disclosed by a consumer when a policy is first taken out. To prioritise, we then looked for evidence in each of these areas that harm was in fact being caused to consumers.

Phase 2: Evidence of harm in the real world

We drew from sources including FCA data, published complaints, guides and case studies from the FOS, and work on insurance claims conducted by Which? to establish which of our potential areas were causing harm to consumers in practice. This process quickly confirmed flood and storm definitions as areas particularly likely to be impacting consumers. 

Previous survey work by Which? has shown that many consumers who need to make a home insurance claim do so to fix storm damage, and that rejection rates for these claims are high. In 2024, Which? surveyed 3,322 people who had made an insurance claim in the last 3 years, asking them about their experience of the process, and the type of claim consumers had made. We found that damage caused by ‘high winds, heavy rain or hail’ had by far the highest rejection rate, at 18%, and was the second most prevalent type of claim, after escape of water from inside a property.

The FOS have also recently reported an increase in flood and storm claims. A report in August 2024, discussing the drivers of a 10 year high in buildings insurance complaints, flagged an increase in ‘numbers of cases where flood and storm damage is complex and expensive’.

Published FOS decisions also provide some insight into the harm caused by floods and storms. We found more than a dozen complaints (nine published after 2023) in which consumers had claims rejected by insurers enforcing restrictive interpretations of flood or storm. This can’t give a complete picture of the issue – decisions are only published when a consumer chooses not to settle with an insurer or drop a complaint, and many complaints are resolved following an initial FOS assessment – but these complaints provide a human view of the disruption and distress caused by flood and storm damage. Many of these are outlined in this document, including a case regarding ‘Mr and Mrs D’, below, who were told that the groundwater which damaged their kitchen had built up ‘too slowly’ to be covered as flooding.

As the complaints cited in this report illustrate, having a claim for flood or storm damage rejected can cause significant harm to consumers, who often face heavy financial and emotional costs from unresolved or worsening damage to their homes, as well as time spent and distress caused in dealing with the claim and complaints processes.

Phase 3: How do industry bodies define flood and storm?

Having established flood and storm definitions as areas of focus, we next reviewed how industry defined these terms.

Flood

The Association of British Insurers (ABI) endorses a legal definition for flood outlined in the Flood Reinsurance Scheme (Flood Re). Flood Re was developed jointly by the insurance industry and government to provide flood coverage for high flood risk properties, and defines a flood as:

water, from any source external to a building, which enters a building

(a) at or below ground level, or above ground level, provided that part of the body of such water is at ground level; and 

(b) does so with a volume, weight or force which is substantial and abnormal.

 The following do not constitute a flood: 

(i) the gradual seepage or percolation of water into a building (such as rising damp); 

(ii) water escaping from a water main, drain, sewer, pipe or other thing inside a building, unless such escape was solely the consequence of a flood falling within paragraphs (a) and (b).

Guidance from the Financial Ombudsman Service (FOS) is helpful in clarifying how they assess events while making decisions. Their guidance to insurance on floods states that:

A flood doesn't have to be a sudden or violent event. It can occur when water enters (or builds up) in a property slowly and steadily, and this doesn't necessarily have to be caused by a natural event. The key factor is that water builds up, regardless of where the water came from.

In addition to stating that a flood needn’t be a ‘sudden or violent event’, the FOS lists several possible sources of flooding, including a rise in the water table.

 Both the Flood Re definition and the guidance from the FOS interpret a flood as an event where water enters the property from any external source. This water doesn’t need to come from a natural watersource, for example a river or the sea, and more importantly, the event doesn’t need to happen rapidly - there just needs to be a significant or abnormal ingress of water entering from outside the building. 

There is some nuance here. This reading assumes that, as their parentheses suggest, Flood Re’s exclusion of the gradual seepage or percolation of water into a building (such as rising damp) is indeed designed to exclude rising damp and similar damage which could be prevented through good building maintenance – and not flooding which happens slowly, but with ‘substantial and abnormal volume, weight or force’. This is reinforced by guidance issued by Flood Re which states that they will cover groundwater flooding where those conditions are met.

Storm

The term ‘storm’ does not have a precise meteorological meaning - it can be used in a variety of ways. As the Met Office puts it on their website, it ‘means different things to different people’: 

The term storm is usually applied to any violent atmospheric disturbance, whether it’s a thunderstorm, squall or snowstorm.

A storm means different things to different people and while it doesn’t have an official meteorological definition, like a Hurricane, it is commonly used to describe a deep and active area of low pressure with associated strong winds and precipitation.

In insurance, one commonly cited definition for storms draws on the Beaufort Scale, which provides an empirical measure of wind intensity alongside descriptive terms for various speeds of wind. Winds described as ‘storm’ force occur at Beaufort Level 10, defined as gusts of between 55 and 63mph. Guidance from the ABI, however, is clear that a storm doesn’t need to involve high winds: any extreme weather event of rain, snow, or hail can be a storm. 

The ABI define a storm as follows:

A period of violent weather defined as: 

• Wind speeds with gusts of at least 48 knots (55mph)* or; 
• Torrential rainfall at a rate of at least 25mm per hour or; 
• Snow to a depth of at least one foot (30 cm) in 24 hours or; 
• Hail of such intensity that it causes damage to hard surfaces or breaks glass

*Equivalent to Storm Force 10 on the Beaufort Scale.

The FOS offer a definition similar to the ABI’s:

We say that a storm generally involves violent winds, usually accompanied by rain, hail or snow. But in some cases we may find there’s a storm without there being high winds. There are occasions where rain, hail, or snowfall by itself can constitute storm. Any extreme form of bad weather has the potential to cause damage to a property.

Phase 4: What are consumers’ expectations around flood and storm?

The Consumer Duty is clear that the FCA expects firms to act in a way that is consistent with reasonable expectations of customers in their target market and design products and services to meet consumers' needs and objectives. To investigate what consumers expect of buildings insurance products, Which? surveyed 1,324 UK adults with buildings insurance in February 2025. Our survey aimed to establish how consumers defined flood and storm, and to test their intuitions around what buildings insurance should cover them for. Our findings are below.

Most of those surveyed with buildings insurance defined a flood as when  water enters and builds up in the home, and defined a storm as involving multiple types of weather

69% of those surveyed with buildings insurance defined a flood to be when water enters and builds up in the home, regardless of the speed or quantity of water.

Figure 1: A clear majority of UK adults surveyed with buildings insurance define a flood as a build up of water regardless of speed

Source: UK adults with buildings insurance (n=1,325). Respondents were able to select multiple options, so the total % is higher than 100.

Figure 1 suggests that most consumers are broadly in line with industry definitions of flood. There is, however, some nuance in each definition not covered by our question: 

Firstly, FOS guidance doesn’t state an opinion on how much water is needed to constitute a flood, so we don’t know whether they’d agree with the majority of respondents who defined a flood ‘regardless of quantity of water’.

Secondly, as it could technically be argued that rising damp is ‘water building up in your home very slowly’, Flood Re’s exclusion of ‘gradual seepage or percolation (such as rising damp)’ means that it doesn’t technically cover water build up ‘regardless of speed’. This exclusion is clearly designed to protect insurers from moral hazard associated with damp problems which could be prevented by standard building maintenance, and is fair as long as it isn’t used to deny coverage for gradual flooding. As it allows all but the very slowest forms of water ingress, this definition is broadly in line with most consumers defining flood as happening at any speed. 

Finally, the Flood Re definition also has a quantity requirement in that the water must enter with a ‘volume, weight or force which is substantial and abnormal’ – though this is consistent with small amounts of water entering a building slowly under high pressure.

Taking these technical differences into account, Figure 1 suggests that most consumers in fact understand flood in broader terms than industry do; there is nothing covered by Flood Re or the FOS definition which wouldn’t be included in the way most consumers think about floods.

For storms, most of those surveyed were broadly in agreement with the ABI definition, defining storms as involving any combination of heavy rain, hail, snow or wind. 60% of respondents viewed a storm as a combination of extreme weather conditions, any of which could occur alone - only 26% asserted that a storm was not a storm without high winds.

Figure 2: A clear majority of UK adults surveyed with buildings insurance define a storm as an event involving multiple types of weather, possibly occurring independently

Source: UK adults with buildings insurance (n=1,325). Respondents selected a single option.

Most respondents expect to be covered for gradual flooding and storms with extreme weather without high wind, and feel that to deny claims based on definitions which exclude these events would be unfair

In reviewing policies during the exploratory phase of this project, we identified some edge cases where we thought consumers were likely to be caught out: circumstances in which we thought consumers would reasonably expect to be covered, but which would be excluded by some of the definitions we’d seen in documents. 

To test our assumptions about consumer expectation in these cases, we presented respondents with a set of questions designed to test people’s expectations of insurance cover in various flood and storm scenarios. Scenarios were designed to reflect gaps found in flood and storm coverage, and edge cases which had, or could, cause damage to UK homes.

Figure 3: A majority of consumers surveyed with buildings insurance expect a typical home insurance policy to cover them for flood damage which happens gradually, or arising from a burst water main

Source: UK adults with buildings insurance (n=1,325)

Which? thought consumers would expect to be covered by floods which happened gradually, or which originate from a range of external sources; a reasonable expectation, given guidance from the FOS and Flood Re. Figure 3 shows that, in three scenarios illustrating these cases, we were correct: at least two thirds of those surveyed with buildings insurance expect their insurance to cover them for each of these types of floods. 

As well as asking respondents whether they’d expect to be covered in various scenarios, we asked whether they would be surprised if an insurer rejected a claim because it didn’t meet their definition of a flood, and whether they felt this rejection would be fair. In both cases, a clear majority told us they would be surprised to have these claims rejected, and that they would consider these rejections unfair.

Figure 4: Over two thirds of consumers surveyed with buildings insurance would be surprised if an insurer rejected a home insurance claim for flooding which was gradual or caused by a burst water main, and at least 80% think a rejection on these grounds would be unfair

Source: UK adults with buildings insurance (n=1,325)

These proportions are relatively consistent across our three scenarios - although one in three stated that they wouldn't be surprised if their insurer declined a claim after damage from a burst water main. Tellingly, for each scenario, the majority of people who indicated that they would not be surprised by a decision to reject also felt that decision would be unfair, suggesting a degree of cynicism towards insurers amongst those we surveyed.

Figure 5: A majority of consumers surveyed with buildings insurance expect a typical home insurance policy to cover them for rain, hail and snow damage occurring without high winds

Source: UK adults with buildings insurance (n=1,325)

While respondents tended to be less confident about being covered for scenarios involving storms than involving floods, Figure 5 shows that more than six in ten of consumers surveyed expect to be covered for damage caused by heavy rain (62%) and hail (69%) without high winds. 53% of respondents also expect to be covered for damage caused by heavy snow without winds. The smaller group here may reflect the fact that heavy snow is a less common occurrence for homeowners in the UK.

A substantial minority of respondents also told us they didn’t know if they’d be covered in each case. This is especially high for damage caused by heavy snowfall, with 28% of those surveyed with building insurance expressing uncertainty as to whether they’d be covered. Such uncertainty isn’t surprising in the context of recent Which? research showing that most consumers don’t understand the limits of the policies they buy. 

This dynamic – consumers unclear on their policies and insurers selling policies with non-intuitive exclusions – increases the risk that consumers experience financial, time, and psychological detriment when claims are rejected; both the cost and stress of unexpected expenses, and the frustration of hours spent engaging with their insurer to try prove their case, understand why the claim was rejected, or make a complaint. 

Figure 6: Most consumers surveyed with buildings insurance would be surprised if an insurer rejected a home insurance claim for rain, hail or snow damage without high winds, and at least 7 in 10 would feel this decision was unfair

Source: UK adults with buildings insurance (n=1,325)

In each of the three scenarios we tested, a majority of our respondents stated that they would be surprised if their insurer turned down a claim for damage caused by heavy hail (64%), rain (61%) or snowfall (58%) without high winds. While this effect is less prominent than the reaction to flood rejections shown in Figure 5 above, we still found that, across each of our scenarios,at least 70% of those surveyed with buildings insurance felt that a rejection would be unfair, and more than half would be surprised. 

As with floods, this question unearthed poor expectations about insurers’ behaviour. In all scenarios, at least 70% of those who said they would not be surprised by a rejection also said this would be unfair (76% for hail, 77% for rain, and 70% for snowfall).

Our tests for fairness in definitions of flood and storm

Drawing from Flood Re, ABI and other guidance, as well as consumers’ expectations as revealed in our survey, Which? developed the following test for fairness in definitions of flood and storm in home insurance policies: 

We believe flood definitions to be potentially unfair if they:

• Exclude gradual, slow and steady entry of water into a building, e.g. by specifying that water must have entered the building 'suddenly' or 'rapidly;' or
• Exclude gradual, slow and steady build up of water outside the building, e.g. by specifying that water must have built up 'suddenly' or 'rapidly'; or
• Exclude events involving water ingress from certain external sources – e.g. from a rise in the water table, or non-natural events (such as a burst water main outside of the property).

We believe storm definitions to be potentially unfair if they:

• Stipulate that a storm must involve any single type of weather - e.g. high winds; or
• Exclude weather consisting of any of rain, hail, high wind or snow alone either explicitly or through omission by specifying only some types of weather.

Chapter 2: Measuring the prevalence of potentially unfair definitions in current UK insurance policies

Having established a clear test for fairness in flood and storm definitions, we applied it to a collection of 133 home insurance policies from 67 different distributors, either dated since 2023, or sent to us in response to a recent query.

We found that:

Almost a third of UK home insurance policies reviewed contain a flood definition which we view as potentially unfair.

A fifth contain a storm definition which we view as potentially unfair.

A sizeable percentage of UK home insurance policies define flood and storm in a way which we think contradicts industry guidance, as well as most consumer’s expectations around what constitutes a flood or storm based on our survey findings.

Figure 7: A substantial number of home insurers policies contain a potentially unfair storm or flood definition, or leave these undefined

Source: UK home insurance policy documents dated since 2023 (n=133) Numbers show counts of documents.

Across all 133 home insurance documents in our dataset, 42 (32%) have a potentially unfair flood definition, and 27 (20%) have a potentially unfair storm definition.

This practice is not restricted to niche firms or specialised policies. Distributors selling policies which use these definitions include some of the largest underwriters in the UK home insurance market: we found policies with what we believe to be potentially unfair flood definitions being sold by three of the ten largest underwriters by market share (Aviva, AXA and Ageas) and policies with, in our view, potentially unfair storm definitions being sold by one of the ten largest (Lloyds Banking Group). 

Of 54 documents which define both flood and storm, we found four policies, offered by John Lewis, Nationwide and Tesco Bank, in which both definitions are, in our view, potentially unfair.

In the case of the Tesco Bank policy, this has changed over time. A Tesco Bank policy dated 2021 has a simple, fair definition of flood:

Flood: Water external to the home entering the home at, or below ground level.

However, in the most recent policy we have, dated 2024, this wording has been changed to require a sudden entry of water (emphasis here added by Which?):

Flood: A sudden and large volume of water external to the home entering the home at, or below ground level.

While policies will occasionally change over time, these changes should not lead to policy terms which do not meet consumers' reasonable expectations for coverage. Based on our consumer survey findings, we think this 2024 definition would not meet consumers’ expectations.

When we contacted insurers, a Tesco Insurance spokesperson said: 

“We regularly review our policies to ensure we offer the best possible outcomes for our customers." 

“Upon further review, we accept our wording in relation to floods could be improved”

“We are currently conducting an in depth review of our policy documents and will incorporate this feedback into this review.”

A Nationwide spokesperson said: 

“Flood or storm-related claims often touch on other areas of our policy which will provide cover for customers. That’s why our claims pay-out rate on flooding and storms outperforms the industry average. While we have defined sections in our policy on floods and storms, we will always take a holistic view of any claim. This comprehensive approach means claims are properly assessed in the round as opposed to being limited to a specific area of our policy.”

John Lewis declined to comment.

Definitions like these make a difference to consumers, and we know that similar ones have been used to reject claims in the past. Complaints upheld by the Financial Ombudsman, detailed below, include examples of insurers declining claims for flooding because water didn’t build up sufficiently quickly, or for damage caused by torrential rain because wind speeds weren’t fast enough. These decisions can leave consumers bearing significant and unexpected financial burdens, on top of the inconvenience of living in alternative accommodation, or the worry of remaining in a damaged home.

Almost all potentially unfair flood definitions define a flood as a ‘rapid’ or ‘sudden’ event

Where flood definitions are potentially unfair, insurers usually require water to enter the home or build up outside the home rapidly or suddenly. A small handful of policies also exclude damage caused by water from specific external sources.

Figure 8: Most potentially unfair flood definitions require rapid water build up or entry, a few restrict certain sources of flooding

Source: Which? analysis of UK home insurance policies (n=133 policy documents) Percentages don’t sum to 100 since documents might be coded as potentially unfair for more than one reason.

17 of 42 potentially unfair policies define flood in terms of the speed at which water enters the home. Many insurers use language like ‘rapidly’ and ‘suddenly’ while others make explicit exclusions to water which enters the home gradually. For example

‘A substantial volume of water suddenly entering your Buildings from an external source at ground floor level or below’ - Aviva, 2023

'By flood, we mean water that comes suddenly into your buildings from outside and which enters at the ground floor or below' - Ageas, 2023

24 of 42 potentially unfair policies define flood in terms of the speed at which water builds up outside the home. As with speed of entry, insurers often use the words ‘rapid’ and ‘sudden’. For example:

‘Invasion of the property by a large volume of water, caused by a rapid build-up or sudden release from outside the buildings’ - AXA, 2025

We chose the three examples presented above as they were included in policies from insurers in the UK’s ten largest underwriters by market share, meaning they are likely to affect a high number of consumers.

When we contacted insurers, a spokesperson for Aviva said:

“We strongly refute any suggestion that our policy wording is unfair.”

They went on to add that they believe their policy wording is consistent with Flood Re, and that they accept the overwhelming majority of flood claims they receive. 

A spokesperson from AXA said:

“The definition of flood and storm perils in AXA home insurance policy wording is in line with the advice and guidance set by independent bodies.”

They went on to say they believe their definition is fair. 

Full responses from insurers are included in Annex B of the PDF report. Ageas declined to comment. 

Finally, seven policies of 42 were coded as potentially unfair for excluding flooding from some external sources. Some insurers exclude flooding caused by a rise in the water table, while others exclude flooding from artificial sources external to the home like tanks and pipes.

Altogether, restricting the speed at which water either builds up or enters a home occurred in all but two of 42 potentially unfair definitions. Both Flood Re and FOS guidance aligns with consumers’ natural expectations (shown by our survey) that a flood ‘doesn't have to be a sudden or violent event’ and yet some insurers seem to insist that it does. 

We know that insurers applying similar definitions have caused harm to consumers in the past. FOS decisions show that insurers have previously relied on what we would now view as potentially unfair definitions to reject consumers’ flood claims. And it’s not just policies with potentially unfair definitions - we know that even where insurers don’t define ‘flood’ (accounting for 56% of our dataset), they sometimes still reject claims for flooding which does not occur as a sudden event. Ultimately it is consumers – with flooded basements, dropped floors, and rainwater in their kitchen – who pay the financial, emotional, and time costs of these definitions.

‘What your policy doesn’t cover:’ rise in the water table

The water table is the level below which the ground is saturated with water. When the water table rises, slow and steady flooding can creep under floorboards, or into basements, cellars, and garages. As climate change affects rainfall and temperatures, groundwater flooding is expected to become increasingly common in the UK.

Flood Re are clear that they cover groundwater flooding, and distinguish this in their definition from rising damp. Of the consumers we surveyed, the majority (66%) believed they would be covered if a rise in the water table caused groundwater to enter and damage their property. Many insurers, however, disagree: we found that 38% of policies we analyzed explicitly exclude flooding caused by a rise in the water table or groundwater. Considering water table exclusions alongside the issues described above around the speed of water build up, water entry and source, 59% of documents in our dataset have some kind of potentially unfair flood clause.

We understand that there may be cases in which insurers feel that they cannot provide cover for rising groundwater – where the water table has risen over decades as a result of changes to the climate, for example, making risks unmanageable – but the blanket exclusions for all flooding caused by groundwater which we’ve seen in our research go against industry guidance and consumer expectation.

Almost all potentially unfair storm definitions exclude damage caused by rain alone, and more than half only cover for storm damage with high winds

By definition, each of our 27 potentially unfair storm definitions exclude damage by at least one of rain, snow or hail alone from their storm coverage. Of these, rain is the most commonly excluded event, followed by snow, hail, and then documents which exclude all three. In this last category, 13 documents (almost half of potentially unfair policies) define a storm as an event involving high winds, implicitly excluding each of severe rain, snow, or hail alone. 

 Figure 9: Almost all potentially unfair storm definitions exclude rain without high winds

Source: Which? analysis of UK home insurance policies (n=27 policy documents containing a potentially unfair storm definition)

Figure 9 shows that coverage for rain, snow, and hail in the absence of high winds are not evenly distributed. Out of 27 potentially unfair definitions, 25 exclude heavy rain on its own compared to 18 for hail and 14 for snow. This not only seems to go against consumers’ expectations of what they’re covered for, but we know from FOS complaints that storm damage won’t always involve wind – any extreme form of weather could potentially leave consumers facing expensive home repairs.

The high numbers of rain exclusions are explained in part by six policies from constituent parts of the Lloyds Banking Group that share a common storm definition:

‘When we say ‘storm’ we mean strong winds over 55mph, and/or hail or snow that’s extreme enough to damage hard surfaces or break glass. Rain alone is not a storm. We won’t pay for any damage if your home is not in good condition. We will, however, pay claims for internal damage if you were unaware that your home was not in a good condition and the damage was caused by torrential rain of 25mm or more in a day’.

This definition is a marked improvement on other potentially unfair definitions we have seen, in that it allows for internal damage caused by rain, and is therefore likely to cover most of the damage consumers would want to claim for. As a result, in our view it is likely to mostly meet consumers’ reasonable expectations for coverage after a storm, and therefore, we think, is more likely to be in line with the Consumer Duty than other examples in this report.

However, the insistence that ‘rain alone is not a storm’ seems at odds with industry guidance. The ABI – of which Lloyds Banking Group is a member – is clear that ‘torrential rainfall at a rate of at least 25mm per hour’ is considered a storm. We also think there is the potential for confusion in the wording of the definition, with the words ‘rain alone is not a storm’ appearing to exclude damage from rain alone, followed directly by wording which states that coverage is provided in some circumstances. This complexity may confuse some consumers, and could discourage them from making valid claims. When contacted, a spokesperson from Lloyds Banking Group provided background information, but declined to provide a comment for publication. 

Furthermore, we think there are plausible cases of external damage that might be excluded here - as the FOS complaint cited below shows, rain alone can cause external damage to homes. 

Given the findings of our survey, we also believe that a refusal of coverage under this definition would likely surprise consumers: most of those surveyed with buildings insurance believed that heavy rain without wind would be covered by their insurer. In fact, 77% of the consumers we surveyed felt that rejecting a claim for extreme rain just because winds were low at the time would be unfair. 

Over half of home insurance policies reviewed have no definition for flood, and a third have no definition for storm

56% of policies reviewed have no definition for flood, and 32% have no definition for storm. In these cases, we think it’s difficult for either Which? or a consumer to predict how an insurer will respond to claims for damage caused by torrential rain, for example, or flooding caused by a rise in the water table.

In cases where a policy does not define a term, consumers should be able to expect that an everyday understanding of that term will be applied. This was made clear by the FOS in a 2022 decision:

‘An insurer, whether as part of an on-line application or as part of the policy, can’t reasonably give a definition for every word used. Rather an insurer will provide a definition for any words with specific meanings for the insurance cover, and any other words used in the application or policy will have their ordinary everyday meaning’.

Furthermore, if a court was asked to rule on what a definition should be, it’s likely that they would apply a definition that favours the consumer, following a legal doctrine known as contra proferentum. This doctrine essentially states that where there is doubt about the meaning of a contract, it should be interpreted against the interest of the party who wrote the terms.

While clear definitions can hold insurers accountable to consistent, fair practice, a lack of definition in a policy can allow insurers leeway to apply definitions that may be unfair, causing consumers harm. Indeed, we know from FOS complaints included here that, given this leeway, providers have applied what we would now view as a potentially unfair definition where none was given in the policy.  

Chapter 3: Recommendations

Our research suggests that there is a significant mismatch between the public’s expectations of flood and storm coverage and the definition used by many firms, including some of the largest home insurance providers in the UK. According to our survey, consumers’ expectations are broadly in line with how the ABI and Flood Re define flood and storm, and many firms follow these definitions. This suggests that consumers’ expectations are reasonable, leaving many firms potentially failing to meet the FCA’s expectations that firms consider the reasonable expectations of their target market when designing products and services.

The risk to consumers of financial, emotional and even physical (e.g. breathing difficulties due to unresolved damp) harm from rejected and partially rejected claims is significant. While some consumers may pursue a complaint if they feel their claim for flood or storm damage has been handled unfairly, the FOS typically adheres to any definitions set out in a firm’s policy terms so these can be difficult to challenge. Moreover, relying on consumer-led complaints to ensure the system provides fair outcomes is likely to discriminate against people who are less able to pursue complaints. With people who have experienced a flood or a storm by definition likely to be in vulnerable circumstances and needing to resolve the situation as quickly as possible, this makes it even harder for consumers to complain to ensure they are treated fairly. 

It’s crucially important that firms are held to requirements under the FCA’s Consumer Duty that aim to prevent foreseeable harm from arising. Based on the findings detailed in this report, we recommend the following measures to address the harm which can result from potentially unfair flood and storm definitions:

Which? have not been able to analyse every insurance policy currently available in the UK, nor have we been able to gather data from firms on the outcomes specifically from flood and storm claims. It is now for the FCA to investigate the scale of harm caused by firms’ approaches to product design, including the extent to which claims involving floods and storms are driving the almost four in ten buildings insurance claims that are rejected. The FCA’s ongoing claims-handling review covers how firms’ ‘systems, controls, governance and oversight structures drive good consumer outcomes’. This broad approach should enable the FCA to look at the root causes of high rates of rejected claims for home insurance, including whether firms have met requirements under the Consumer Duty regarding their approach to flood and storm cover. 

This research has highlighted two key areas where we believe the FCA should investigate further to ensure firms are meeting the Consumer Duty:

1. The FCA’s Consumer Principle and considering the target market’s reasonable expectations

In acting to deliver good outcomes for their customers, the FCA expects firms to act consistently with the reasonable expectations of customers and to consider the reasonable expectations of customers in their target market in relation to the product or service. While the FCA is clear that the Consumer Duty does not remove consumers’ responsibility, it also makes clear that consumers can only take responsibility for their actions when they can trust that the products and services they can choose from are designed to meet their needs, and where they are supported by firms to make decisions.

Our survey findings clearly suggest that most people’s expectations are not being met by many firms’ products when it comes to coverage of floods and storms. We think these expectations are reasonable given that they are broadly in line with the definitions of the UK’s largest trade body for insurers, the ABI, and the Government’s Flood Re scheme. That many firms adhere to these definitions also proves that these are insurable risks. The FCA should investigate what evidence firms have used to determine their flood and storm definitions, including their reasoning for taking a different approach to the established definitions used by the ABI and Flood Re.  

2. The FCA’s Consumer Duty governance of products and services outcome

Manufacturers and distributors of insurance products are subject to specific requirements under the FCA’s Consumer Duty regarding meeting the needs, characteristics and objectives of their target market, including the need to regularly review that this is occurring. High rates of claim rejections or complaints involving floods and storms should be a key metric that firms are monitoring to trigger a review. Where the distributor is a different firm, they should be requesting these data from the manufacturer. Where they identify issues, they should be seeking changes from the manufacturer and, if this is not possible, ultimately considering whether to remove the product from their offering. We believe the FCA should be investigating manufacturers’ and distributors’ review processes and the actions they have taken to avoid foreseeable harm. 

 We want firms to prevent consumers from experiencing avoidable distress, expense and lost time by ensuring that clear, common-sense definitions for floods and storms are used in their home insurance policies. To meet the reasonable expectations of their target market, this could most easily involve adopting the Flood Re definition of a flood and the ABI definition of a storm. Our evidence does not support definitions that provide less coverage than the Flood Re and ABI definitions as being sufficient.

Some firms may choose to innovate and offer broader coverage for floods and storms, which may appeal to some consumers and may become increasingly salient due to the impact of climate change. However, given widespread agreement among consumers we surveyed of what reasonably constitutes a flood or a storm and that this is broadly in line with industry and government-backed guidance as well as the definitions used by many firms, we generally do not believe this is an area where competition over terms is likely to be effective. Our previous research on the insurance sales process has highlighted that consumers commonly lack understanding of the limitations of insurance and quite often think that insurance products are ‘much of a muchness’. 

The inherent complexity of policy terms makes it difficult for consumers to compare policies on the definitions used for specific areas of cover such as flood and storm. As the examples we’ve cited in this report show, this can lead to harm when claims are unexpectedly rejected. While some consumers will feel able and willing to challenge decisions, this puts many consumers – who may be more likely to be in vulnerable circumstances – at greater risk and may prevent them from obtaining a good outcome. 

While a lack of definitions in policy terms may not necessarily lead to harm, as firms’ claims-handling practices could still be in line with their customers’ reasonable expectations, on balance we think that this is likely to increase the likelihood of unfair outcomes. In these cases, we think insurers should at least apply the definitions endorsed by the ABI, Flood Re and the FOS. As mentioned above, the contra proferentem rule provides that any ambiguities in a contract should be interpreted against the drafter. In practice, however, this principle may only be invoked where consumers, unhappy with their claim being rejected, are able to complain and possibly seek redress from the courts or the FOS. Where this is the case, this is likely to lead to systematically poorer outcomes for some consumers, including those in vulnerable circumstances who are less able to advocate for themselves. 

While our analysis here has focused on floods and storms, where we saw the most pressing evidence of current harm and prioritised our limited resources, we expect that similar patterns of definitions which do not align with consumer expectations, either within policies or used to decide claims outcomes, may be causing wider consumer harm. In its claims handling review, the FCA should look more widely across the causes of rejected claims to identify any other underlying issues, and require firms to take similar steps to address them to prevent foreseeable harm to consumers.

Annex

The annex of the report can be found in the PDF copy of this report.