Share article
Martin Švamberg isn’t the type to lecture on customer experience from a distance – he’s spent twenty years pulling it apart process by process, number by number. From banking and energy, he made his way into insurance: an industry where the true quality of CX only reveals itself when something goes wrong. In this interview, he talks about how to separate signal from noise, why speed isn’t always the priority – and what his early dead ends with AI taught him.
Insurance and great customer experience – that sounds almost like a contradiction in terms. How did you end up deciding to work in CX at an insurance company?
My path to CX was actually shaped earlier by banking, where I learned to work with data and relational NPS categorisation. From there, a stint in the energy sector showed me just how fragile customer trust can be during times of market upheaval – like the collapse of Bohemia Energy.
I see insurance as a natural next step. It’s an industry that combines the stability of banking with the sensitivity of energy. But what really drew me in was the technological shift that’s happening right now – the opportunity to use AI to turn “just insurance” into something genuinely simple, intuitive, and efficient for the customer. That was a challenge I simply couldn’t walk away from.
When did you first realise that customer experience isn’t just about “being nice to customers” – that there’s serious analytical work behind it?
There were several moments, but probably the most powerful came with mortgages, about fifteen years ago. We realised that even the most empathetic branch adviser couldn’t ease the frustration of a customer waiting a month – sometimes longer – for mortgage approval. When you’re buying your first flat or building a house, you need support that’s fast and reliable.
That’s when CX became pure analytics for me. We had to break down the entire customer journey into individual processes, identify the critical bottlenecks, and compress what took weeks into days. That’s when it clicked: real CX isn’t about being nice to customers – it’s about process analytics that removes the barriers standing between a customer and their goal.
What does a typical day look like when you’re “hunting” a problem in the customer experience? Do you have a ritual or a process?
I start the morning by checking new items in the system we use to analyse customer feedback – a quick scan to see whether anything unusual is happening and how much of it there is. If anything in the queue needs more attention, I connect with the relevant team and we work out the next steps together. I also keep an eye on our CX metrics in Power BI. During the day I’ll dip into the CX community on LinkedIn to see if there’s anything worth reading 👍.
On the longer-term side, I’m always working on three or four development areas where we’re trying to move the needle significantly on customer experience.
How do you know when customer feedback is a genuine signal and when it’s just noise?
When you’re looking at customer feedback every day, you develop a feel for what’s routine and what’s new – it becomes almost instinctive. When it comes to a specific customer, you need to dig into their story: what products they have, how long they’ve had them, how they use them, and what their recent interactions with us looked like. Sometimes you start investigating a piece of feedback and realise the customer actually had that experience with a different insurer – or they’re reacting to something a friend or family member told them 😉.
What was your biggest “aha moment” in data analysis – a moment when the numbers told you something completely counterintuitive?
Everything around us is speeding up, so you’d expect speed to be the primary driver of customer satisfaction today. And in many cases, it is. But the human element still plays a significant role. When an adviser properly sets expectations upfront for a case that might take longer to resolve, customers end up just as satisfied as they would have been if the issue had been resolved quickly.
Direct pojišťovna has a reputation for taking CX seriously. What do you think is behind that – culture, processes, or specific people?
Essentially a mix of all three, and then some 😉.
It’s a culture where the company has built its success on several pillars, one of which is the client experience. In the insurance market, we stand out through an innovative approach – simplicity, flexibility, friendliness, and technological maturity.
Insurance is unique in that customers only really *need* you when something has gone wrong. How do you design a great experience around a moment nobody wants to have?
In that “moment of truth,” empathy is essential – alongside speed, process simplicity, and clear communication. The customer wants to get back to where they were before the unexpected event, as quickly as possible.
And when – despite every effort – you hit a snag, you owe the customer a clear explanation and, where necessary, an honest look at what went wrong: a risk not covered in the policy, a windscreen undervalued, breakdown assistance not included in the cover…
I know you well, Martin, and I know you’re constantly coming up with new ways to do CX better. Can you share an example of something you tried that turned out to be a dead end – and what it taught you?
Nothing comes to mind more readily than our first steps with AI 😉. The initial attempts at categorising CX comments using ChatGPT about two years ago involved… quite a few dead ends 😀:
- The early assistants in ChatGPT didn’t work the way we expected. We had to learn how to write prompts and think carefully about how to build an effective assistant.
- The new AI capabilities set our expectations very high – we assumed we’d be creating dozens of new categories and constantly finding new connections between them.
The reality at the time was that we were learning on the go. We discovered that rather than crafting individual prompts for everything, it was far simpler to use a universal approach and train the system to categorise based on validated, tested taxonomies – grouped into two to four categories that can vary slightly depending on the touchpoint.
We don’t use ChatGPT for categorisation anymore, but as an introduction to what AI can do, it was a brilliant school 👍.
How do you convince people in the organisation that a CX change is worth the investment – especially when results aren’t immediately measurable?
When it comes to process changes, concrete examples work best – showing where things are grinding to a halt. For instance, if customers are struggling with photographing their cars correctly during onboarding, your conversion rate drops, you’re losing business, and you’re creating dissatisfied customers all at the same time.
The ideal scenario is when people inside the organisation are also customers of the company – so they can live the CX story themselves 😉.
What’s the thing that frustrates you most as a CX analyst – something you see over and over and think, “this really isn’t that complicated”?
I have very little patience for recurring open feedback items – an unresolved query, a policy left incomplete, an email that never reached the customer. If we don’t get it right first time, it costs us far more effort later – effort that could have gone into something genuinely meaningful.
What do you think everyone who wants to do CX well should be able to do – regardless of industry?
- Know your customer and be able to identify the key personas you’re serving.
- Be a customer of your own company – experience the CX first-hand and live the customer journey (the brilliant parts and the imperfect ones alike). Visit a competitor occasionally 👍. Draw inspiration from everywhere 😉.
- Understand your company’s processes – what works brilliantly and makes you stand out, and where the gaps are.
- Propose changes, innovations, and simplifications.
- Use common sense – and don’t be afraid to 👍.
If you had to explain what you do to your grandmother at a family gathering – how would that go?
I make sure our customers have a great experience today – and an even better one tomorrow 💚.
How do you see the expectations of today’s customers, shaped by other industries?
Customer expectations are rising in line with the ambitions of the most progressive players out there. Take Alza: “Order by evening, pick it up from the box in the morning.” That sets the bar extraordinarily high – and it influences every other sector of the economy, not just consumer electronics and delivery services.
Insurance can be progressive too. Getting insured based on a single photo of your car’s number plate is no longer just a distant dream 😉.
And finally – do you have a favourite CX story?
I’m sure I could find one – how about a dating story 😉? Every now and then, a comment appears in call centre feedback from a customer inviting an operator out for coffee – or straight out on a date 😀. The key ingredient behind those comments is usually the operator’s empathy, or simply a very pleasant voice. We make sure the feedback gets passed on – whether anything ever came of it, we honestly have no idea 😀.
