IE caught up with Tim Hurst, Chief Operations Officer at TXP, to learn more about updating legacy tech;
1. Why is legacy tech a major issue in the insurance market right now?
Many insurers have been operating for decades, and in some cases centuries, so it is natural that they carry a significant amount of technology heritage. Those systems have grown in layers. New platforms have been added on top of old ones, while critical business processes have evolved around large volumes of historical data. Over time, policy, claims and finance platforms have become so closely embedded with one another that overhauling any one of them is complex and potentially risky.
As a result, innovation can be slow at a time when insurers want to move fast. For example, more than a decade since cloud started to become widespread in insurance, many insurers are still struggling to migrate legacy systems. With AI becoming a business imperative over the last few years, many insurers are unable to drive full value from technology because they are held back by legacy.
2. How has the growth of the insurance industry exacerbated this?
The UK insurance market today is unrecognisable from just 40 years ago. Since the 1980s, it has grown almost 25-fold, with premiums rising from around £15bn in 1980 to around £372bn today. Many of the systems still in use were designed for a smaller, more uniform market. Insurers are now being asked to process far greater volumes for a far more diverse customer base.
Much of that growth was absorbed by bolting new products, channels and acquisitions onto existing platforms rather than upgrading them. Every addition made the underlying systems more complex and intertwined, so the cost and risk of change have grown in step with the business itself. The bigger an insurer has become, the harder it has been to modernise.
3. How do you balance modernisation with the relationship-led service specialist insurers are known for?
Modernisation should protect, not dilute, the relationship-led advantage of specialist insurers. By automating high-volume, low-judgement work such as submission intake, document retrieval, rekeying, status updates and routine renewals, underwriters and claims teams can spend more time on complex risks, difficult claims and broker conversations.
The principle should be human-led, technology-enabled: use data and AI to prepare insight and recommend actions, but retain human ownership where judgement, empathy or material customer outcomes are involved. Technology should remove friction – not the conversation.
4. What will happen if insurers don’t digitalise?
Insurers that fail to digitalise risk becoming progressively less competitive. For UK specialist and commercial carriers, manual submission, quotation, endorsement and claims processes increase the cost of doing business, constrain underwriting capacity and make business harder for brokers to place. Fragmented data also slows portfolio decisions, product development and regulatory evidence, while limiting the value available from analytics and AI. This matters as digital trading becomes more established across the insurance market and the FCA expects insurers to map, test and address vulnerabilities in important services.
The answer is not to automate away expertise. Human judgement, negotiation and trusted relationships remain differentiators. Digitalisation should remove avoidable friction so specialists can focus on complex risks, difficult claims and valuable broker conversations.
5. Insurers’ systems are vast. Where should they begin with digitalisation projects?
Insurers should start by understanding where technical debt is constraining the business. Years of integrations, workarounds and incremental change can leave technology estates increasingly complex and difficult to change, so simply adding new digital capabilities can compound rather than solve the problem. The answer is not necessarily wholesale replacement. Insurers need to understand their architecture, data, and dependencies, then determine what should be simplified, modernised, integrated or retired. Modern cloud platforms, APIs and better data architecture can progressively reduce complexity and create stronger foundations for digital and AI.
Insurers should phase digital transformation, but not fragment it. Large-scale change creates significant execution risk when too many systems, processes and dependencies are changed simultaneously. The London Market’s Blueprint Two programme, for example, demonstrates the complexity of coordinating technology, data standards, processes and multiple market participants at scale.
However, phasing does not mean pursuing disconnected projects. Incremental initiatives can also fail when they create overlapping point solutions, duplicate data or pilots that never become embedded in day-to-day operations.
The most successful programmes tend to balance a clear long-term vision with practical, phased delivery. Insurers should establish the target architecture, data standards, governance and desired business outcomes upfront, then deliver transformation through manageable, end-to-end phases. Each phase should solve a meaningful business problem, demonstrate measurable value and provide learning for the next.
6. What did the Shipowners’ Club project involve?
The Shipowners’ Club is a major mutual insurer for specialist vessels, with 8,296 members and 35,378 vessels insured. Its brokers needed simple, fast access to the information they rely on to service their clients.
TXP built a digital portal that brings key information into one place, including policy documents, claims loss ratios and broker statements, available whenever brokers need them. This process included extensive discovery, architecture, user-centred design and knowledge transfer, connecting historical and live data to the new system. Since launching in 2022, the portal has been taken up by around 1,300 brokers worldwide and now also lets them bind renewals online.
Brokers can now find what they need themselves, which leaves the Club’s teams free to focus on the conversations where their expertise makes a difference.
7. AI is becoming a business imperative for insurers. What challenges do insurers face? Our recent research with UK mid-market leaders found that ambition is high. Almost all (98%) already have an AI strategy in place. Organisations have invested an average of more than £1m in pilots and proofs of concept over the past three years, and nearly half (47%) of those pilots have scaled into production.
Getting the rest over the line is the challenge. Almost half (49%) of leaders say their AI initiatives have so far failed to deliver value or proved underwhelming, and 61% have paused or wound down initiatives that fell short. The barriers are mostly practical: the cost of moving from pilot to production (49%), governance and risk concerns (39%) and poor data quality or accessibility (38%). Integration with the wider technology estate is another sticking point, and it is especially difficult for insurers with deep legacy estates.
Knowledge is a less obvious but crucial problem. The research found that critical knowledge often sits with individuals rather than being documented and shared, so a pilot can stall simply because the person who understood a system has moved on.
8. How can insurers address these challenges and make sure their AI initiatives stand out?
Most insurers have access to the same AI models. As the technology becomes a commodity, differentiation comes from what each insurer brings to it, such as proprietary data and underwriting expertise.
Our research points to three steps. Start by capturing the knowledge that lives with individuals and making it available across teams. Then redesign workflows around AI where it adds value, instead of bolting it onto processes that were never designed for it. Finally, focus on orchestration: map technical dependencies and connect AI to the systems and data that support it.
Insurers that combine their own expertise with the right technology will get more projects into production and build an advantage competitors will find difficult to replicate.
Be the first to comment