From 8 policies to 100,000 claims: what scaling embedded insurance really looks like
%20(1).png)
In a nutshell
- Scaling embedded insurance isn't linear. Some of our early product failures set the foundation for the focus and resilience of the platform today.
- Operational crises like 'code yellow' and 'code red' drove Qover to adopt AI and automation tools that have multiplied team productivity while keeping customer satisfaction high.
- Compliance by design – treating regulatory requirements as the foundation of product development rather than a final check – changes the entire trajectory of a launch.
Alex is in charge of content marketing at Qover, meaning she's obsessed with all things storytelling and tone of voice.
With more than a decade of experience in content strategy, copywriting, editing and SEO, she's worked for media companies, marketing agencies and startups across the globe.
When she's not writing or strategising, you'll likely find her in a language class (Dutch, currently), on a spin bike or diving into her latest Netflix binge.
Our CEO reflected on a decade of insurtech entrepreneurship in his recent column, and now three other members of the management team set out to tell their sides of the tale in our 10th anniversary webinar.
Chief Operating Officer Ed Ackerman, Chief Product Officer Alix Moggia and General Counsel Caroline Hanotiau gave their takes on what it really means to build and scale embedded insurance from the ground up.
They covered ground that many companies don't share publicly: products that flopped, operations that buckled under pressure and the lessons that reshaped how Qover builds today.
Watch the full webinar replay →
From rockets to results: the early product pivots
We try to be just as open about our misses as our hits, which is why Ed kicked off the webinar by sharing a few product flops that helped us pivot in the early days of Qover.
In 2019, we launched Nomad Assistance, a travel insurance policy that quickly reached €1 million in gross written premium across Belgium and France. Then came COVID. The risk carrier pulled out of travel insurance entirely, and we spent six months winding down the product.
'We launched lots of rockets,' Ed said, 'but not all of them reached space.'
The GAP motor insurance experiment – our first official product on the market with Toyota – was more sobering still: 17,000 quotes, eight policies sold.
These early failures weren't detours, but lessons learned. They shaped our conviction that building something truly scalable means identifying the white spaces: problems that aren't yet solved, with solutions that aren't yet on the market.
Sometimes those solutions don't exist because they haven’t been invented yet. Sometimes it's because they don't work.
The key is finding out quickly and pivoting accordingly.
Claims crises as catalysts: code yellow and code red
Fast-forward to 2023. Qover's digital claims management operation had grown significantly, but growth had outpaced the system.
'In 2023, we hit critical mass with our claims operations. Volumes rose over 500% between January and December that year. We're no longer talking about one claim, but hundreds and thousands of claims,' Ed explained.
To ramp up operations to meet demand, we doubled the team and temporarily worked weekends. Ed sent daily Slack updates to track progress against their targets. They called it 'code yellow'– a three-week push to clear a growing backlog and hit SLAs we pride ourselves on meeting.
It worked. But it was a huge reality check.
Anticipating volume growth and scaling proactively, rather than reactively, had to become the operating model. After all, claims are the moment of truth in insurance. If you lose trust there, you lose customers and business clients. Simple as that.
The second alarm, 'code red', came when Trustpilot scores started dropping. The pattern was clear: slow claim settlement times were the single biggest driver of negative reviews. The higher the number of interactions needed to resolve an issue, the worse customers felt about the service.
The response was equally focused: target 24-hour claim settlements, train teams to spot signs of frustration early and reduce friction at every step. Trustpilot scores recovered, and we’ve maintained strong public review ratings since.
AI as the operations multiplier
Both claims crises pointed to the same answer: speed is critical, and automation can help us get there. Not to replace the team, but to dramatically expand what the team could do.
We doubled down on AI in insurance claims and customer care and built AI chatbots to help customers submit the right documents at first contact. Then we deployed claims straight-through processing and voice AI.
Productivity per claims officer and customer care agent rose many times over, without a proportional increase in headcount.
'The beauty of AI and automation is it also helps us service customers more quickly and more efficiently,' Ed said.
By mid-last year, we handled 100,000 claims. By the end of the year, we were on track to double that – and to handle half a million customer service tickets in the same period.
These numbers don't necessarily reflect the past 10 years of experience, but the last three to five – and that's what hyperscale looks like.
Read more: what we’ve learned from using AI in claims processing →
Building embedded insurance at scale: why repeatability beats go-live
Alix picked up the story from the product and technology side, introducing the philosophy behind how the platform works.
With a background in ad tech, she noticed when she entered the insurtech space that most technology was built bespoke, product by product, partner by partner. Like tailoring a unique dress for every customer: it works, but you have to start again every time.
We decided to take a different path. By putting product thinking at the core, we built for repeatability and speed, not just the next launch.
It didn’t always work that way in reality – every partner has different needs – but it’s something we’ve followed fiercely since Alix has been on board.
Today, our insurance orchestration platform is modular and end-to-end – covering policy engine, payment flows, claims management, reporting and regulatory compliance, all built on a configurable core that partners can deploy in weeks rather than months.
'We don't want to be a feature shop,' Alix said. 'We’re building a platform and we’re very clear on what the platform is, what the features and capabilities are, how we evolve and what goes on the roadmap.'
A dual-team model makes this work: professional services teams handle configuration and speed-to-launch, while product managers and designers focus on platform strategy.
Every quarter, we co-builds our roadmap with top partners, keeping development collaborative and genuinely useful rather than generic.
Next on the horizon is moving from AI-supported to AI-native – aka embedding AI directly into the platform's core modules, not as an overlay but as architecture.
'Brands no longer want to own an insurance entity,' Alix added. 'They're searching for agility and orchestration.'
Compliance by design: from afterthought to architecture
Caroline rounded out the session by talking about our journey from compliance as a final check to compliance as a starting point.
In the early days of Qover, the approach was compliance by detection. Teams would design and build a program, then review at the end whether it met all regulatory requirements. If it didn't, delays and costly rework followed.
'Compliance by design means seeking compliance not as a last act, but as the starting point – the foundation of everything we design,' Caroline explained.
That shift has real consequences for how we build programs. The compliance team is now involved from the discovery phase, translating regulatory requirements into technical specifications that engineers and solution designers can act on from day one.
Operating across multiple European markets adds real complexity: directives must be interpreted at a local level, every country has unique insurance regulation and product-specific rules vary significantly. There's no one-size-fits-all.
To stay on top of constant regulatory change, we’ve built a layered monitoring approach – external regulatory watch tools that flag relevant changes, and more recently, AI agents connected to the legal, compliance and privacy teams that alert the company whenever a regulatory shift in a market of operation could affect a product or program.
'Thanks to these tools, we manage to catch a lot of the regulatory changes and anticipate them,' Caroline said.
What distribution partners get wrong about scaling embedded insurance
The session closed with a Q&A that pulled together all three perspectives on one question: what's the biggest misconception distribution partners make early in their embedded insurance journey?
Caroline's answer was direct: 'The biggest mistake is underestimating the compliance element of the project. If you come around to that too late, there can be additional delays and additional cost.'
Alix added that beyond compliance, partners often underestimate the full ecosystem integration required – the CRM connections, point-of-sale experience, what end users actually expect, etc.' The impact of all those elements sometimes isn’t seen upfront. That's why orchestration is key – calling all those elements at the right time.'
Ed rounded it out from the operations side. It's easy to tick every technical and compliance box while losing sight of the person at the end of the journey. Keeping the customer at the centre, from day one, is the hardest discipline of all.
Ten years in, what comes next
A decade of building embedded insurance at scale has produced a modular platform that can go live in weeks, a claims operation handling hundreds of thousands of cases a year and a compliance framework that works proactively rather than reactively.
More than that, it's produced a set of hard-won lessons that no amount of planning would have delivered on its own.
As Ed so eloquently put it, 'Incidents tend to unlock an element of human creativity that you wouldn't necessarily get outside of that sort of emergency period.'
It’s not somewhere you want to live, but the war stories are part of the process. And so is what came after.
Curious about building embedded insurance at scale? Get in touch with our team.
.jpg)
