Simon Glairy brings a nuanced perspective to the insurance sector, where he has spent years dissecting why some firms thrive under pressure while others buckle under the weight of their own legacy. As an expert in Insurtech and risk assessment, he has witnessed firsthand the friction caused by outdated systems and the “performance gap” that prevents insurers from excelling across operational, financial, and compliance-driven domains. In this conversation, we explore why the true solution to modern volatility lies not in a “big bang” technology replacement, but in the careful management of institutional knowledge and the courage to overhaul ancient workflows. We discuss the shifting priorities of global regulators, the myth of technology as a standalone cure, and the necessity of building a dynamic operating model that treats data as the ultimate competitive fuel.
Claims inflation is currently outpacing premium growth while regulators shift their focus toward operational resilience. How are these dual pressures creating what you describe as a “performance gap” for modern insurers?
The performance gap is essentially a structural inability to move at the speed the market now demands, and it manifests in five critical areas: operational efficiency, financial stability, compliance, decision-making, and time to market. When you look at the P&C world today, it feels like the ground is constantly shifting beneath our feet due to climate-driven catastrophes, social inflation, and interest rate dislocation. Because many insurers have failed to keep up with necessary systemic changes over the years, they find themselves vulnerable to this extraordinary volatility where claims inflation is still outpacing earned premium growth. This puts combined ratios under mounting pressure, making it nearly impossible to maintain a steady course without a fundamental change in how the business functions. It is no longer just about having a strong balance sheet; it is about having the agility to react to whatever comes next, whether that is a reinsurance cycle compression or a new geopolitical tension.
Many organizations treat cloud platforms and AI as a “magic wand” for instant agility, but why is this perspective often considered a myth in the context of operational transformation?
There is a persistent myth perpetuated by techno-optimists that simply installing a new IT system will create instant adaptivity, but the reality is much more grueling. Technology is often treated as a magical solution, yet platforms like the cloud or emerging AI trends cannot, on their own, afford an organization new thinking or a faster speed of movement. To truly transform, an insurer must be willing to banish ancient processes and address the underlying restrictions that have hindered their flexibility for decades. If you layer expensive new software over a broken, rigid operating model, you are simply automating inefficiency rather than solving it. Real change requires a deep dive into the operating model to ensure that it can absorb change and align decisioning across the entire enterprise.
How is the shifting gaze of regulatory bodies and rating agencies toward operational resilience changing the way insurers must justify their business models?
We are seeing a significant shift where supervisory bodies and rating agencies are looking far beyond just the financial statements to understand an insurer’s underlying operating model. They now view operational resilience and technology readiness as core signals of a company’s long-term preparedness and survival. When a company relies on disparate legacy systems, it naturally increases running costs and creates information silos that lead to fragmented audit trails. These silos are a major red flag for regulators because they lead to delayed filings and increased regulatory exposure, which can be just as damaging as a financial shortfall. A modern insurer needs to demonstrate a single, cohesive view of all information to prove that their underwriting logic is sound and that they can update products quickly enough to avoid dangerous coverage gaps.
When an insurer needs to adjust underwriting rules due to a sudden shift in catastrophe loss patterns, what are the primary obstacles that prevent this from happening in days rather than weeks?
The primary obstacle is almost always that institutional knowledge is dispersed, unmanaged, and often quite literally frozen in time within legacy code. When a sudden shift in catastrophe loss patterns occurs, the challenge is rarely as simple as changing a single field in a system; instead, the relevant knowledge is scattered across actuarial assumptions, underwriting guidelines, and regulatory requirements. Many of these older systems have rules hard-coded into them, meaning that processes and workflows cannot be easily configured without bespoke programming that risks breaking other dependencies. Unless that knowledge is structured and accessible, even a seemingly minor pricing change can take weeks to navigate through the various silos. To fix this, insurers need to organize their institutional knowledge first, unlocking those inaccessible data points so they can be used to drive rapid, informed decisions.
For established insurers with decades of data, how can they implement continuous, iterative change without disrupting their day-to-day operations?
Most commercial insurers do not have the luxury of starting from scratch like a startup, so they must adopt a bimodal approach where change takes place alongside daily operations. This shouldn’t be a “one and done” project, but rather a process of continuous, iterative improvement that focuses on the biggest pain points or “quick wins” first. There is a significant behavioral adjustment required here, as staff must learn to trust AI-assisted processes rather than defaulting to the slower, more familiar habits of the past. Leaders have a massive role to play in this by communicating why change is necessary and demonstrating how these new systems actually make complex tasks easier to achieve. When employees see that their productivity is increasing and that they can let go of manual, repetitive steps, they become much more amenable to the transition.
With the current excitement surrounding artificial intelligence, what foundational data requirements must be met to ensure that these tools don’t actually amplify dangerous errors?
The old rule of “garbage in, garbage out” has never been more relevant than it is today with the rise of AI, because errors in these processes are amplified and can become incredibly misleading. AI can only truly excel if the foundational data it ingests is high-quality, clean, and appropriately structured from the start. We use technology like machine learning to ingest data rapidly and identify anomalies faster than any human could, but these tools still require a framework of best practices and preconfigured guardrails to stay on track. SaaS applications can help by providing a stream of new features and security upgrades as background processes, ensuring the data lineage is clear by default. The ultimate goal is to achieve a state of “All Systems Go,” where data and processes are so perfectly aligned that external volatility actually becomes a competitive advantage rather than a threat.
Do you have any advice for our readers?
My advice is to stop looking for a single technological silver bullet and start focusing on the “organizational brain”—that unique combination of your people’s expertise, your data excellence, and your internal processes. The companies that will outperform their peers in the coming years are those that can integrate signals faster from exposure data, claims, and emerging loss patterns to reprice products in real-time. You must build a culture where controls are designed directly into the workflows and where every piece of data has a clear, observable lineage. If you can master the art of iterative change and learn to treat your operating model as a dynamic asset, you will find that your combined ratios remain steady even when the rest of the market is in turmoil. Focus on the plumbing of your data today, and the sophisticated AI applications of tomorrow will take care of themselves.
