This piece is by Greg Hanson, VP and Head of EMEA North at Informatica from Salesforce
Insurers are moving fast beyond one-size-fits-all products. AI now shapes decisions across underwriting, pricing, claims and customer engagement, and the market reflects that change. From a value of $8.63 billion in 2025, the AI insurance market is projected to reach $59.5 billion by 2033. Allianz’s partnership with Anthropic and Aon’s collaboration with DataRobot signal a much broader race to personalise. That race matters most where the stakes are highest.
Commercial and high-net-worth risks are more complex, change is more frequent and this demands a far richer understanding of customers and assets. A static, one-size-fits-all policy cannot reflect a portfolio of properties, a growing business, or a fast-changing risk profile. The real opportunity for insurers lies in moving to a broader “party model” — understanding how a customer connects to a wider household, business network and ecosystem of risk.
Hyperpersonalisation, tailoring cover, pricing and service to the extended network rather than an isolated profile, is fast becoming the key differentiator. However, it cannot be successfully achieved without trusted data. The more personalised AI becomes, the more it depends on data AI doesn’t really ‘understand’ customers or risk; it only interprets patterns in the data. It cannot tell the difference between a good decision and a bad one, only between the data it has and the data it does not. That’s why hyperpersonalisation is less of an AI challenge than a data one. A 2026 AI Impact Survey from Grant Thornton found that although 52% of
insurance companies report a revenue increase from AI, fewer than a quarter are confident they could demonstrate adequate governance and controls over how it’s used. The gap between AI ambition and data readiness is where hyperpersonalisation efforts fall short.
Three foundations for trusted hyperpersonalisation
A unified customer and network view: Insurers cannot fully personalise decisions if the information behind them is scattered. Customer, policy, claims and third-party data is often fragmented across underwriting, claims and service systems, duplicated or stored in incompatible formats. Master Data Management (MDM), and the connected systems that support it, offers insurers a single, reliable view of each customer, asset, and relationship network.
An individual may need cover for their commercial business alongside a domestic personal policy, multiple vehicles, pets and dependents. Without a robust model, an underwriter sees isolated, disloyal single-policyholders who are easy to lose to competitors. With MDM, insurers can identify cross-policy opportunities using trusted data — locking in loyalty through multi-policy household offers covering everything from family cars to business ventures, with the option of bundling to offer discounts.
For policyholders, bundling makes life simpler, it’s perceived as offering better value and making renewal much easier; for insurers, it means deeper, harder-to-replicate relationships and a natural source of stickiness that discourages shopping around. It’s no surprise that bundling drives loyalty — J.D. Power research found that home insurance policyholders who also bundle their auto coverage retain at a 91% rate, compared to just 67% for customers with standalone policies.
Trusted data: A unified view only helps if what sits behind it can be relied on. Personalised pricing, underwriting and claims decisions are only as good as the data feeding them, so accuracy, consistency and completeness are vital. Missing or inconsistent data can otherwise lead to genuine claims being flagged as suspicious or risk being mispriced, eroding trust and margin. Ongoing monitoring, standardisation and clear data ownership can prevent such problems escalating.
Contextual data: Context is everything. Even accurate, connected data can lead to the wrong conclusion without it. A commercial property policy can be linked to the business owner’s personal lines and broader risk history. Insurers need to understand not just what data they hold, but where it came from, how it has changed and how AI can help them reach a decision. Context lets insurers explain a decision rather than defend it, which is increasingly what regulators and customers ask for.
Explainability, the natural output of good data, should be easy to understand at every level of the business.
What happens when hyperpersonalisation gets the data wrong
Get any of these three foundations wrong, and hyperpersonalisation can work against the insurer. Risk gets mispriced because underwriters are working from an incomplete or outdated picture. Customers get inconsistent experiences depending on which system, channel or agent they touch. Cross-selling attempts may be irrelevant or tone-deaf if relationships within a household or business network are misread.
Genuine claims could be delayed or wrongly queried if the data trail doesn’t hold up. A poorly targeted bundling offer is a good example: pushed to the wrong household or mismatched to a customer’s actual assets, it damages trust rather than building insurance stickiness. At scale, a single data gap doesn’t stay isolated — it compounds into a pattern of poor outcomes, often visible only once the damage is widespread.
Hyperpersonalisation will separate tomorrow’s market leaders
Hyperpersonalisation will define competitive advantage in insurance over the next decade — and nowhere more so than for insurers mapping a complete party model across personal, commercial and household risk. Those who pull together MDM and broader data capabilities to offer seamless, unified, multi-policy protection and bundling unlock greater retention and reduce churn. That’s the stable revenue insurers need to grow, stay competitive, and serve customers better.
Ultimately, hyperpersonalisation is about more than competitive advantage — it’s showing up for policyholders when it matters most. Over the next decade, the insurers who thrive will be the ones who truly understand their customers, whether through complex commercial policies or the everyday protection of a family home.
Hyperpersonalisation isn’t an end in itself; it’s proof that an insurer’s data is connected, trusted and thoughtful enough to see the whole picture, not just a policy number. When insurers treat high-quality data as core infrastructure, everybody wins: fair pricing, faster and more empathetic claims processing, and lasting relationships with loyal policyholders. Getting data right is the clearest sign of an insurer that truly knows its customers.
Be the first to comment