Why MIC Global’s Founder Says ‘Don’t Build InsurTech’
When asked what advice he’d give to founders building insurance technology, Harry Croydon doesn’t hedge or equivocate. His answer is two words: “Don’t do it.”
This isn’t coming from someone who failed at insurtech and turned bitter. It’s coming from someone who’s been building insurance technology since 1999 and knows exactly why it’s such a brutal business.
In a recent episode of Category Visionaries, Harry Croydon, Co-Founder of MIC Global, an embedded microinsurance company that’s raised over $13 million, shared the hard-earned wisdom from twenty-five years of trying to sell technology to one of the world’s most resistant industries. His core insight cuts through all the insurtech hype: the problem isn’t the technology, it’s the business model.
The Fundamental Problem with Selling to Insurance
Harry’s warning is specific and experience-based: “Insurance companies, you know, they spend a lot of money on things, but you know, they take a long time to buy technology.”
This isn’t about insurance companies being cheap. They have massive budgets. It’s about how they buy and what they actually want.
The sales cycles are punishing regardless of your target customer. “This is why we aren’t an insurtech, we are an insurance company enabled by technology. You know, that’s what we are. I’ve been selling technology insurance for a long time and the sales cycle, whether it’s to brokers, if it’s to insurance companies, or is very long.”
But length alone isn’t the killer. The real problem is what happens during those long cycles: “They all want semi bespoke solutions and it’s very difficult to service them, I find.”
The Policy Management System Trap
Harry has particularly strong feelings about one category that attracts countless insurtech founders: policy management systems.
“I built my first policy management system in 1999 and I haven’t really seen anything that’s much different. Despite everything that people say, even today, they’re all the same, they’re all just there. So don’t build one of those.”
Think about that for a moment. Twenty-five years of innovation. Cloud computing. APIs. Microservices. AI. Machine learning. And policy management systems are essentially unchanged.
This isn’t because no one has tried to innovate. It’s because insurance companies don’t actually want innovation in their core systems. They want systems that work exactly like their existing processes, just slightly faster or slightly cheaper.
Every insurtech founder believes their policy management system will be different. They’ll use modern technology. They’ll have better UX. They’ll be API-first. None of it matters if the fundamental business model is selling to companies that resist change.
What to Build Instead
Harry’s advice isn’t purely negative. He has specific recommendations for founders determined to build in insurance.
“Build something that is smart and clever for people really, you know, and understand the need. You need to find a need first. Don’t just think, you know how the insurance industry works and you can solve their problems because of. Yeah, I don’t think you can.”
His concrete suggestions focus on tools for specific, acute problems: “If you can keep the system, you know, focused one particular problem that helps or a complete solution that is easier for them to use, like a modeling tool for hurricanes or wildfire or something, these are the things that they want tools.”
Notice the pattern. Hurricane modeling tools solve a concrete problem that insurance companies can’t easily solve themselves. They’re specialized enough that customization demands are limited. They have clear ROI. And critically, they’re tools that enhance existing processes rather than replacing them.
Why the Industry Resists Technology
The insurance industry’s resistance to technology isn’t irrational or backwards. It’s structural.
Insurance companies operate on decades-old infrastructure that works. Their processes have been refined over generations. Their regulatory compliance depends on specific workflows. Their employees are trained on particular systems.
When an insurtech startup pitches a “better” policy management system, what the insurance company hears is: “Replace your entire operational infrastructure, retrain thousands of employees, navigate complex regulatory approval, migrate decades of policy data, and hope nothing breaks.”
The semi-bespoke customization demands Harry mentions aren’t pickiness. They’re insurance companies trying to make new technology fit into existing regulatory frameworks, compliance requirements, and operational realities.
Harry understood this intimately: “You know, it’s difficult to just understand. The industry is very multifaceted and there’s lots of people in it and it’s kind of challenging from that point of view.”
The Alternative Path: Become the Insurance Company
This is why MIC Global chose a completely different model. Instead of selling technology to insurance companies, they became an insurance company that uses great technology.
“We aren’t an insurtech, we are an insurance company enabled by technology,” Harry emphasizes, drawing a distinction that explains their entire strategy.
As an insurance company, MIC Global controls their own technology stack. They don’t need to convince committees to adopt new systems. They don’t need to customize their software to fit legacy processes. They build what they need, when they need it.
The capital requirements are higher. The regulatory burden is significant. But the business model actually works.
The Broader Pattern in B2B
MIC Global’s experience reveals something important about B2B technology companies in general: sometimes the best customers for your technology are the ones you never have to sell it to—because you’re using it yourself.
This pattern appears across industries. The best restaurant technology often comes from restaurant operators. The best construction software comes from construction companies. The best insurance technology comes from insurance companies.
The reason is simple: when you’re the customer, you truly understand the problem. You know which features matter and which are nice-to-have. You know the regulatory constraints. You know the operational realities. You don’t need six-month sales cycles to understand requirements.
When InsurTech Makes Sense
Harry’s advice isn’t absolute. There are successful insurtech companies. But they typically fit specific patterns.
They solve narrow, acute problems with clear ROI. Hurricane modeling. Fraud detection. Specific underwriting tools. These are defensible because they require specialized expertise insurance companies don’t have in-house.
They focus on distribution, not infrastructure. Embedded insurance products that insurance companies can white-label avoid the core system replacement problem.
They target new markets rather than replacing existing systems. Building for gig economy workers or microinsurance means creating new categories rather than disrupting established ones.
Notice what’s missing from this list: core policy management systems, claims processing platforms, and other infrastructure that insurance companies already have.
The Real Lesson
Harry’s blunt advice—”don’t do it”—isn’t cynicism. It’s pattern recognition from twenty-five years of watching insurtech companies burn capital on sales cycles that never close or implementations that never scale.
The insurance industry will adopt new technology. Harry sees it happening now around AI, claims management, and customer service. “I think in three to five years there’s going to be a massive change in the adoption of technology into insurance.”
But the companies that will win aren’t the ones selling technology to insurance companies. They’re either solving highly specific problems with specialized tools, or they’re insurance companies that happen to use great technology.
For founders eyeing the insurtech space, Harry’s question is fundamental: are you building technology for insurance companies, or are you building an insurance company with great technology?
The first path leads to brutal sales cycles, endless customization, and policy management systems that look exactly like the ones from 1999. The second path requires more capital and deeper industry expertise, but it actually works.
Twenty-five years of experience distills down to simple wisdom: if you’re going to build in insurance, don’t build insurtech. Build insurance, enabled by technology. The difference is everything.