Your P&C Personalisation Problem Isn’t a Data Problem

Open a comparison site and look at motor quotes. In most markets across Europe and Asia Pacific, the top twenty results sit within 2–5% of each other. The same convergence shows up in home, SME, and even commercial lines, where brokers anchor the renewal price and insurers work to match it.

That’s not a coincidence, and it’s not a pricing failure. It’s the natural result of an industry where nearly everyone works from the same inputs. Same data. Same pricing engines. Same claims history feeding the same models. Swiss Re’s sigma research points to the structural cause: market concentration has been declining across most major markets, with the top five insurers now holding smaller shares than in 2004 in nine of eleven large markets. More players, more competition, and broadly similar tools create the conditions for price convergence. It’s structural, not temporary.

So when almost every insurer is investing in “personalisation,” it’s worth asking what that word is really doing.

Presentation is not personalisation

The industry has largely delivered on one version of personalisation: demographic segmentation, transaction history, third-party data feeds, and increasingly sophisticated targeting. It lifts conversion rates and makes communications feel more relevant. That’s useful. But it isn’t differentiation, because every competitor has access to the same tools and the same data sources.

If everyone works from the same inputs, better messaging isn’t personalisation. It’s presentation.

Real personalisation is when data changes decisions. That goes well beyond what you say to a customer. It shapes what you actually offer them, by knowing which covers are genuinely relevant rather than serving a generic menu. It shapes how you price. And it shapes when you engage: an experience that adapts in real time as circumstances change, not a slightly better script wrapped around a standard product. This is what hyper-relevance really means: the right solution for the right customer, in the moments that matter.

The demand is there, even if customers can’t always name it. A 2025 Deloitte survey on P&C insurance found that almost 60% of customers want personalised products, yet over 70% still believe standard products meet their needs. That gap between latent demand and a market that hasn’t shown people what’s possible is a real opportunity. But it will only be captured by insurers whose operating models can support it.

The operating model is the congestion point

Here’s the part most personalisation strategies skip over. In EMEA and APAC, most P&C insurers are running personalisation initiatives on top of operating models that were never built for them. Pricing, underwriting logic, product configuration, and claims feedback sit in separate systems, updated on separate cycles, governed by separate teams. The signals exist. They just can’t flow into decisions quickly enough to matter.

We call this the Decision Gap. Pricing, underwriting, and claims signals become disconnected. Risk selection grows inconsistent. Customers end up with products that no longer reflect their actual needs or risk profile. Personalisation becomes a front-end promise the back-end can’t keep.

This is why the difference between leaders and everyone else rarely comes down to customer-facing technology. Capgemini’s World Insurance report found that only 5% of insurers globally deliver truly outstanding customer experience, and that best-in-class insurers achieve 38% higher Net Promoter Scores and 11% lower expense ratios than their mainstream peers. That difference is a function of operating model maturity, not a slicker front end. McKinsey’s research raises the stakes further: over the past five years, insurance sector leaders in AI and technology generated 6.1 times the total shareholder return of laggards, a spread wider than in most other sectors. The gap isn’t narrowing. It’s compounding.

AI accelerates whatever you already have

There’s real and justified enthusiasm for AI-driven personalisation. There’s also a risk many insurers are underestimating.

Embed AI into a fragmented operating model where decisioning logic is siloed and data doesn’t flow cleanly between functions, and you don’t fix the underlying problem. You accelerate it. You speed up flawed underwriting logic. You scale inconsistent pricing. You produce bad decisions faster.

The numbers show how wide the gap between ambition and readiness has become. A McKinsey survey of more than 50 leaders from the largest European insurer groups found that over half expect AI to deliver productivity gains of 10 to 20%. Yet only a third have initial use cases in production, and 60% describe their traditional data as merely “evolving.” The ambition is clear. The foundation that would make it real is, for many, still a work in progress.

The insurers who will make AI work for personalisation are embedding agentic AI into the core of their policy administration and decisioning building systems where every interaction generates a signal that signal informs the next decision, so product and pricing logic adapts continuously. The human stays in the loop; the underwriter’s expertise trains the system, and the system handles the scale no team can. At that point, personalisation stops being a feature you bolt on. It becomes the way the business runs.

Fix the operating system, close the gap

The personalisation gap in P&C is real. But at its core it isn’t a data problem or a technology problem. It’s an operating system problem.

The insurers who close it will stop treating personalisation as a customer experience initiative and start treating it as an operating model outcome. When data changes decisions at every point in the value chain, and those decisions adapt continuously as customer and market signals shift, personalisation stops being something you do to customers. It becomes how you run.

That takes capabilities that work together rather than in isolation: cover selection based on a customer’s real profile, adaptive product logic that evolves without cancellation and reissuance, real-time pricing that reflects current exposure, and product versioning that updates continuously without disrupting the existing book. The technical components already exist. What most insurers are missing is the operating system that connects them.

This is only the starting point. For the full picture, including where personalisation breaks and what structural readiness looks like, read the whitepaper, Pinpointing Your P&C Personalisation Problem.