Leadership lessons from Zurich’s chief data officer

Data transformation, like any transformation, is fundamentally about people. While technology is important, the true challenge is creating a culture and an organisation ready to embrace change.

Transformation isn’t a one-off event; it’s a continuous journey towards ensuring that you can meet the needs of your organisation now and in the future. It’s crucial to create scalable modular foundations from an operating model, architecture and technology perspective.

Don’t believe anyone that tells you it will be easy. Transformation is hard, and patience and transparency are key. You will suffer setbacks and a two-speed approach is key to success. The ability to demonstrate and communicate value in the form of ‘no regrets’ activity and incremental progress will enable you to keep buy-in until you complete each phase.

Support is fundamental – from the ground up and the top down. Without this, you risk creating a technically beautiful solution that doesn’t deliver value.

CRM is crucial in any customer-centric sector like insurance. A strong CRM foundation enables a unified view of the customer, and AI can take CRM to the next level – but only if your data is integrated, clean and trusted.

The biggest challenge is that many CRM systems are siloed or underutilised. Embedding CRM into your wider data architecture is what unlocks real business value.

When aligned with AI, CRM has the potential to transform customer engagement, driving positive customer and commercial outcomes.

It’s essential. Diverse perspectives help challenge assumptions, uncover blind spots and ensure that the insights and solutions we generate are fair, representative and relevant to our customers and stakeholders.

Diversity drives innovation. In complex environments like insurance, having a team that reflects that complexity leads to richer problem-solving and more relevant solutions.

This is even more critical with Al. Bias in data or models often stems from a lack of diversity, whether in the data itself or those building and validating the models. A team with varied experiences, backgrounds and ways of thinking is far more likely to question the status quo or spot ethical or practical issues early.

AI enables hyper-personalisation, from tailored policy recommendations to proactive engagement during life events. With AI-driven segmentation and behavioural modelling, we can move beyond one-size-fits-all products to truly individualised offerings.

There’s a huge opportunity to enhance risk selection and pricing. When paired with geospatial data, IoT inputs or external datasets, AI models allow for more granular, dynamic underwriting. This can create a competitive advantage, especially in specialty lines or emerging risks.

In claims, AI is driving transformation through intelligent triage, document classification, fraud detection and straight-through processing. Generative AI opened up a new frontier, from automating customer communications to summarising complex legal or policy documents.

All of this relies on strong data governance, ethical frameworks and a deep understanding of where humans should stay in the loop. As we scale AI, we also must ensure it’s used responsibly with fairness, transparency and sustainability in mind. If we get that right, AI won’t just make insurance more efficient, it’ll make it more relevant, responsive and resilient.

One recurring issue is the expectation gap in data leadership. Many organisations are still unclear about what they truly need from a data leader. They often hire for deep technical expertise but then expect strategic business transformation, which sets both the leader and the business up for frustration. The industry needs a clearer, more mature understanding of what effective data leadership looks like.

Another critical challenge is operating model design in increasingly decentralised environments. In most large organisations, data is generated, owned and acted on within business units. The traditional centralised model is giving way to a hybrid world where value is materialising at the edges. The question is, how do you enable synchronised autonomy? That is, empowering domains to move fast while maintaining alignment through shared platforms, governance and principles.

The pace of change, particularly around AI, adds complexity. Technology is moving faster than most organisations, regulators and even boards can respond to. Data leaders are being pulled into the spotlight to help make sense of these changes and ensure organisations act responsibly and with foresight.

Leadership in data isn’t about technical brilliance – it’s about influence, vision and impact. Many aspiring data leaders come from deep technical roles, but to truly lead you need to pivot towards understanding your business, its goals, its challenges and how data can drive outcomes.

That means developing strong soft skills. You need to be able to translate complex insights into clear business value, tell compelling stories and build trust with both technical and non-technical audiences. A successful data leader creates a culture of innovation and collaboration. They encourage experimentation and build psychologically safe teams where diverse perspectives are valued.

Stay informed about emerging technologies, but focus on what truly serves your organisation and customers.

My guiding principle is simple: build the best and brightest team you can and then create the conditions for them to thrive. I never want to be the smartest person in the room, I want to surround myself with people who challenge ideas, bring fresh perspectives and push the boundaries of what’s possible.

As a leader, my job is to give people space to be brilliant, to remove obstacles, and to amplify their achievements so they’re recognised and celebrated. But equally, when things go wrong and they inevitably do, I take responsibility. Success belongs to the team; failure sits with me.

That kind of trust-based leadership builds loyalty, fosters innovation and ultimately delivers the outcomes that no individual could achieve alone.