Transformation initiatives have to strike a balance: the ambition to modernize quickly while bringing the entire organization along for the journey. Jeff Duncan (pictured left), EVP and head of commercial lines, and Erich Bublitz (pictured right), head of excess and surplus at AmTrust Financial, recently shared how the firm pushes to improve broker experience and operational efficiency.
Bublitz described transformation in broad terms, but his emphasis pointed to a structural challenge. “Transformation isn’t just about any one part of the business,” he said. “It’s transforming the entirety of the experience for everybody that does business with us, be it internal or external.” That scope introduces complexity, particularly in a business spanning multiple product lines and legacy processes that have evolved over decades.
He acknowledged that while underwriting performance and scale remain relevant, the focus has shifted outward. “A lot of our focus isn’t about improving underwriter profitability… it’s really about how do we innovate in a way that makes our broker partners and our agents have a better experience,” he said. That shift, however, requires reworking systems and habits that were not originally designed around ease of use.
“One of the biggest single challenges to any innovation… is the human element,” Duncan said.
“When you innovate without bringing people along… innovation can become something you’re doing to them instead of for them,” he said. The implication is that even well-funded technology programs can stall if they fail to secure internal buy-in. In practice, this creates a pacing issue: transformation must move slowly enough to carry stakeholders, even when competitive pressure demands speed.
The company has attempted to mitigate that risk by ensuring more cross-functional teams are working together earlier in development. Duncan pointed to a shift away from isolated product design. “Too often, innovation happens with IT teams or product teams working in isolation,” he said. “By bringing them in at the idea stage… you get to better outcomes.” The approach reflects a broader industry recognition that adoption, not deployment, determines success.
Bublitz reinforced the same point from a distribution perspective. “Too often, someone gifts you something innovative without asking whether it’s a gift you actually wanted,” he said. Feedback mechanisms such as producer conferences are intended to surface practical constraints before systems are built, though they also introduce additional coordination and iteration.
Much of the visible change has centered on broker interaction points, particularly quoting. Duncan said the company identified quoting as a source of both value and frustration. “It can be either very useful to them or very painful for them,” he said, referring to the business owners policy platform.
Efforts to improve that experience have required targeted investment rather than wholesale system replacement. “We spent quite a bit of time, energy, and money revamping our quoting platform,” he said. The language reflects incremental modernization rather than transformation through a single platform overhaul, a pattern common across mid-market carriers managing legacy infrastructure.
Even these targeted improvements carry organizational implications. Simplifying workflows for agents often requires standardizing or reconfiguring underwriting inputs, which can challenge established practices. The result is a balancing act between usability and underwriting control, with neither side fully dominant.
At a broader level, both executives pointed to cultural expectations as a driver of continued change. Bublitz said transformation is “part of our day to day,” while Duncan noted that leadership pressure reinforces that stance. Yet their comments suggest that culture alone does not eliminate friction; it creates permission to address it.
The firm’s approach to data highlights another dimension of transformation complexity. Bublitz drew a clear distinction between technology and data sourcing. “We have a preference for building in-house,” he said, citing the need to align systems with internal processes rather than adapting workflows to external platforms.
However, Duncan acknowledged reliance on external data to improve pricing and reduce manual input. “We spend a significant amount on purchasing third-party data,” he said. That reliance introduces a strategic constraint, as widely available data offers limited differentiation.
“If you can buy it, anybody can buy it,” he said, underscoring the need for proprietary insight. The company has attempted to address this by reinterpreting its own historical data, though that process can challenge long-standing underwriting assumptions.
Duncan cited an example in small commercial lines. “When we actually looked at our own data, it was very persuasive that protection class and building age should be pricing variables — not risk acceptability variables,” he said. The finding runs counter to conventional underwriting heuristics, illustrating how data-led analysis can conflict with institutional judgment.
“The underwriter bias is against certain things — you can’t do that,” he said. “But when the data is persuasive, the opportunity is there to do something differently. I think that speaks to the AmTrust cultural mindset around challenging assumptions and never being satisfied by the status quo.”
