The real test of digital transformation: claims

For years digitalization in insurance was discussed around pilot projects, proofs of concept and flashy technologies. The question is no longer whether we’ll digitalize. The real question is: where and how will we create real impact? The answer, I think, sits in a very clear place: claims.

Because claims is the only moment the customer truly experiences insurance. What happens at purchase is a promise; whether that promise holds is revealed only at the moment of claim. Pure digital players permanently raised customer expectations by making claims fast, simple and transparent. Now everyone expects the same ease.

Why claims?

McKinsey’s work shows an end-to-end digitalized claims process can raise customer satisfaction by 25 to 30 percent, cut claims costs by around 20 percent, and significantly reduce classic problems like over- or under-payment. So improving claims both pleases the customer and eases the balance sheet. Few areas achieve both at once.

A customer may renew their policy for years without a single claim; the only real contact they have with insurance happens when a bad day comes. Managed well, that moment creates lifelong loyalty; managed badly, a single experience erases years of trust. So claims is where the promise is kept, not where marketing happens. And it’s where we most need to talk about digitalization.

Not a single technology, but a holistic redesign

But this transformation doesn’t come from one shiny technology. In successful cases five elements are handled together:

• A new digital value proposition for claims.
• A customer journey designed end to end, from first notice of loss (FNOL) to payment.
• AI-supported decisioning and segmentation.
• An ecosystem integrated with services, adjusters and repair networks.
• A new, agile and data-driven operating model.

Adding a mobile app or a chatbot alone goes no further than polishing the shop window without changing the core. The real work is rebuilding the whole process through the customer’s eyes.

The heart of operations

At the center of this redesign is operations. Three dynamics stand out: moving core systems to cloud-based, API-enabled structures; smart automation (using RPA and AI together) speeding up policy and claims processes; and the simplification, through unified platforms, of the middle layer where underwriting, claims and finance intersect. The result is fewer manual steps, more speed and transparency.

But there’s a dimension as important as technology: culture. Real transformation begins not in the systems but inside the organization. Without agile teams, data literacy and leaders who champion this change internally, even the most expensive system melts away inside old habits.

Agentic AI on stage: Allianz Project Nemo

A concrete example shows where this transformation is heading. Allianz’s Project Nemo, with an agentic-AI architecture, settles low-complexity claims like food spoilage in hours instead of days. In the background, different specialist AI agents work together on planning, scope checking, weather-event verification, fraud analysis and payment calculation; claims-processing time is cut by up to 80 percent.

The most critical point: the final payment decision is always made by a human expert. This human-in-the-loop approach combines AI’s speed with human judgment. Instead of drowning in many small claims during crisis periods, operations teams can focus on the truly complex cases. In a world where disasters are more frequent, high-volume claims periods are no longer the exception but the new normal; and that makes such scalability a necessity.

Nemo’s modular structure isn’t specific to one country or line either; it’s designed to be adaptable to areas like travel delays, simple auto claims and low-complexity property claims. So this isn’t a single success story, it’s a scalable model.

The global picture differs, the direction is the same

Operational maturity isn’t at the same point in every market. While new markets like Singapore and the UAE leap straight to cloud- and AI-based systems, established Western insurers struggle to clear decades of technical debt. The pandemic made this gap visible: the flexible adapted fast, the heavy fell behind. For Turkey the opportunity is exactly here; it’s possible to turn lateness into a leap by going straight to modern infrastructure.

But full automation is a myth

It would be wrong to conclude from this that everything will be automated. According to data from the British Columbia Insurance Council, only 1 percent of companies run the quoting process fully automatically. The bulk, at 64 percent, is at a low automation level: the human gathers the information, the system works as a calculation tool, and the human still makes the decision. Even at high automation levels, reviewing the quote and managing risky situations stays with the human.

The only reason isn’t technological limits. The human is kept in the system because trust, explainability and consumer protection require it. So the future of digitalization in insurance will be shaped not by the claim of full automation, but by models that can strike the balance of automation in the right place and humans in the right place.

So the real question isn’t how much of the process we can automate; it’s where and how we’ll protect trust, transparency and consumer protection as automation deepens. As speed gets cheaper, explainability and fairness become more valuable. When a customer asks why their claim was paid a certain amount, behind it there should be not just an algorithm’s output, but a defensible logic and, when needed, a human. A well-designed claims operation is one that can hold the machine’s speed and the human’s accountability under the same roof.

What does an insurer that can’t digitalize claims lose?

The cost of missing this transformation is concrete. As platforms and embedded models capture contact with the customer, the strongest card left in the insurer’s hand is the claims experience. Even if you lose the moment of purchase to a platform, you win the relationship back by being the party that stands beside the customer on a bad day. The reverse is also true: an insurer slow, opaque and bureaucratic in claims can’t keep that relationship, however well it markets its product.

In a geography like Turkey, with high earthquake and flood risk, this is even more critical. At moments when high-volume claims hit the table at once, an operation that can act fast, fairly and transparently isn’t just an efficiency matter; it’s where social trust is tested too. The insurer that puts claims at its core makes the difference at exactly these moments.

The insurer’s new role

Behind all this is a role change. The insurer must now be not just the one who pays claims, but the actor who manages the claims experience and the ecosystem around it. Because in the age of platforms and embedded models, what keeps the customer relationship alive is not the moment of purchase, but the experience when a bad day comes.

An insurer that can’t digitalize claims will struggle to protect the customer relationship in this age. But in doing so, AI should be placed beside the human, not in their place. Because speed comes from the machine, and trust still comes from the human. The best insurers will be those who build these two not as rivals but as two halves of the same process. On a bad day the customer wants both a fast solution and an interlocutor who understands them; whoever can offer both at once wins. Digital insurance isn’t a shop-window job; it’s a transformation that touches the core. And the most sensitive point of the core is still the same: claims.