The AFM wants insurers to judge outcomes, not mistake smooth service for customer interest.
A customer adds insurance while buying a telephone online. The offer appears at the right moment, acceptance takes one click and payment is effortless. The journey receives a good satisfaction score. Months later, the customer discovers that the cover overlaps with an existing policy, or that an important exclusion received little attention.
Nothing in the purchase felt difficult. That is precisely the tension.
On 27 October 2026, the Dutch Authority for the Financial Markets published findings on how insurers embed customer interest in their organisation and daily decisions. Insurers completed self-assessments, and the AFM validated the information through office visits. The review covered the customer journey, digitalisation and AI, and organisational culture.
Its message is direct. Insurers should distinguish customer interest from customer satisfaction, both in policy and in daily work.
Convenience deserves respect. Customers value fast answers, clear screens and accessible staff. Yet a pleasant journey only describes how a customer experienced a process. Customer interest asks a harder question: did the product, price, explanation and eventual outcome serve that customer properly?
The metric can hide the decision
Customer satisfaction remains useful. It can show whether a service team answers calls promptly or whether customers find an online process clear. The trouble starts when management treats those measures as the whole story.
The AFM's complaint figures show why one dashboard is not enough. Around 121,000 complaints were reported by 170 insurance providers in 2025, down 3.2% from the previous year. Average handling time fell from 19.4 to 15.5 days.
Those are worthwhile improvements. At the same time, non-life insurance complaints rose by 5.1%, while legal-expenses complaints rose by 24.8%. A company can close a complaint quickly and still leave the underlying weakness untouched. It can also lose sight of customers who decide that a claim, exclusion or dispute takes too much time to pursue.
The same tension appears at the digital checkout. In July 2026, the AFM reviewed 22 online environments offering embedded insurance. It found prominent offers, repeated prompts, less visible refusal options, social proof and time pressure that could steer customers towards taking out cover.
Information on costs, duration, limitations and overlap with existing insurance was often harder to reach than the benefits. The customer buying telephone insurance may still say the purchase was easy. The insurer has to decide whether the choice was also properly informed.
Good friction can protect the customer
Many businesses train themselves to remove friction. Product teams track conversion. Service teams track waiting times. Claims teams track speed and closure. Digital teams seek fewer calls and less manual work.
These targets can improve a business. They can also make the wrong result look efficient.
Some friction earns its place. An extra question during acceptance may reveal unsuitable cover. A warning can prevent duplicate insurance. A visible refusal route may lower conversion while improving choice. Human review may slow an automated decision, yet catch a result that staff cannot explain or defend.
That is where governance enters the commercial room. Customer interest cannot sit only with compliance or complaint handling. Product owners, claims managers, technology teams and directors each shape the eventual customer outcome.
When sales, retention or cost targets pull against that outcome, someone needs the authority to challenge the process and require a change. The decision should leave a usable record: what concern arose, what the firm found, who decided and what changed.
The AFM also places culture inside this question. It points to leadership, room for employees to raise dilemmas and errors, and attention within the HR cycle. The first person to spot a recurring weakness is rarely the board member reading a monthly dashboard.
It may be the claims handler explaining the same exclusion every afternoon. It may be the service employee who hears customers misunderstand a renewal. A business learns when those observations can travel upwards and affect the process.
Monitoring is not the same as knowing
The AFM draws an important line between monitoring a policy and evaluating whether it works. Monitoring may show that employees reviewed automated outcomes when the procedure required it. Evaluation asks whether those reviews reduced customer risk and produced careful results without needless delay.
A completed control is not automatically a useful control.
AI makes this distinction more urgent. The AFM identifies AI use across product development, information provision, acceptance and claims handling. Flawed rules, biased data or incorrect data can affect customers. Refusals and premium increases still need an explanation that a customer and employee can follow.
For management, the practical test is simple. Can the firm reconstruct one customer journey? Which information entered the process? What rule or model shaped the result? Could an employee intervene? Were exceptions recorded? Did a recurring pattern reach the right management level? Did anyone correct it?
This is not paperwork for its own sake. It prevents fragmented systems from producing expensive surprises. When product, claims, service and technology teams each hold one part of the story, a customer dispute or supervisory question creates manual work at the worst possible moment.
Late correction can involve system changes, customer communication, staff training, file reviews and remediation. Earlier testing is often cheaper because it catches pressure before it becomes a wider operating problem.
Customer interest needs an owner
The October AFM publication is a supervisory signal about the maturity expected from insurers. Written intentions are only the beginning. Firms need to monitor execution, evaluate effects and improve the process when customer outcomes require it.
A useful management review can start with one journey where speed, sales or automation carries real weight. Place the satisfaction score beside signs of misunderstanding, cancellation, repeated contact, override, dispute or duplicate cover. Then ask staff where customer interest conflicts with their targets.
Suppliers belong in that conversation when their software, data, call-centre work or claims processes shape the journey. Insurers need reporting that lets them understand actual customer outcomes, identify recurring weaknesses and act on them.
The customer buying insurance with a new telephone should not need to understand the insurer's governance structure. That burden belongs inside the business.
Dutch supervision is pointing towards a simple discipline. Do not confuse a smooth moment with a sound result.
A happy customer is worth listening to. A properly served customer requires more: a decision the company can understand, challenge and stand behind.
If your customer journeys reward convenience without testing outcomes, we can help you bring customer interest into governance and decision-making.
The data, sourcing, and analysis behind this article were conducted by Paolo Maria Pavan. AI was not used to identify sources, build the factual basis, or produce the analytical judgment contained here. AI was used only as a drafting aid. The final English text was personally reviewed, edited, and approved by Paolo Maria Pavan before publication.
References
- Maak expliciet onderscheid tussen klantbelang en klanttevredenheid
- Autoriteit Financiële Markten - Governance beyond commercial incentives
- Autoriteit Financiële Markten - Complaints data versus actual customer outcomes
- Autoriteit Financiële Markten - Digital choice environments and embedded insurance
- Autoriteit Financiële Markten - AI, personalisation and explainable customer outcomes
- Ondernemersplein - AI transparency obligations
- Autoriteit Financiële Markten - Customer attention and financial capability
