Insurance competitiveness now turns on whether firms can turn messy data and human judgment into a measurable edge, not on who buys the latest platform, according to Anju Arora (pictured), VP of data, BI & innovation at Novatae Risk Group. She argues that technology only matters when it sharpens underwriting reality.
That edge, she believes, begins with accepting that risk decisions are made on uncertainty rather than clean inputs, which is why she says “technology is finally catching up with the talent of insurance. We're not making widgets; we're making something we don't ultimately know the cost of, but it underlies so many economies, because we can take risk and sleep at night,” she said.
Arora’s stance is shaped by a career that started in actuarial science and moved across carriers, an MGU and then the brokerage side, leaving her impatient with transformation slogans that ignored actual loss experience and capacity economics. “I don't know how many folks you run into that have an actuarial background, but that is my background,” she said, and she treats every technology investment as a capital decision that has to prove its impact on combined ratios and margins rather than its ability to generate attractive dashboards or tick modernization boxes.
That mindset underpins Novatae’s push to rebuild its core platform after a rapid acquisition spree that has left the organization running 17 businesses on six legacy systems. Arora views the migration to a single policy administration system as far more than an IT consolidation project; she sees it as a chance to impose one way of working and a consistent view of risk across the firm. “Novatae’s history is a young history. It's grown by acquisition, so we're, I think, at 17 acquisitions and six legacy systems. Now we're probably three quarters of the way through adopting and coming on to a single platform. What's really nice is we get to create our way,” she said, arguing that standardization is the only route to reliable analytics and credible conversations with capital providers.
Her remit is to turn that platform into a data operating model that supports revenue growth, margin expansion and underwriting discipline, starting with a hard look at where the firm actually has usable history. Acquisition-led growth creates holes as well as opportunity, and she wants leaders to confront that rather than pretend every segment can be modeled with precision. “Some of the issues we're going to run into are just a lack of history. When you grow by acquisition, everyone does something a little different. What's really nice is this is our opportunity to assess what we have today, what we want for tomorrow, and figure out how we get at those gaps. Sometimes that might just be purchasing some data along the way,” she said.
At the same time, Arora is blunt that even a well-designed core cannot be the single answer. She rejects the fantasy that one system will meet every product, process and workflow requirement in specialty lines, where loss runs, TPA feeds and deal structures vary widely and change fast. “I should say we intend on having a tech stack. It is not going to be a one-system-meets-all. I've been part of implementations in my prior lives, and that never goes well. Worse yet, by the time you finish building, new shiny toys have become available, new products and so on,” she said, pushing instead for a plug-and-play stack that can swap tools around a clean data spine as markets evolve.
Her skepticism comes from seeing what happens when firms chase grand platforms and forget about the people who actually write the business. “I've seen so many implementations go horrible and millions of dollars ruined,” she said, recalling a project she joined midway through that had already blown through budget and timetable before leaders downgraded it to a basic replacement. “I was part of one where I came in halfway through an implementation that had gone so over budget and so over timeline that they ended up de-scoping and calling it ‘like for like’. When it actually went into production, the underwriters were so unhappy that it survived two months and they reverted back to the system that they had once built 15 years ago,” she said.
For Arora, failure usually starts at the requirements table, where organizations lean on the same traders who carry the P&L to design and test new workflows on top of their day jobs and then wonder why adoption stalls. She argues that “where does this stuff go wrong? I have my opinion. I think it's always about the requirements. You have to have the right people in the right room and they have to understand the value. Often we ask the same people that are trading to build this out. That's not how it works,” she said, and she pushes for cross-functional teams that understand both the economics of the book and the configuration details of the tools.
Her view of insurtech follows the same hard line. She sees real upside in bringing in partners to automate heavy, repetitive tasks like submission and email triage but believes too many vendors mistake slick demos for impact and underestimate how little they know about the products they want to optimize. “It is interesting because the other aspect I think a lot of these insurtechs are missing is knowledge. They've got the tech part. They're eager. They know there are processes and efficiencies that they can help us with, especially with automation, and I absolutely agree,” she said, stressing that the firms she wants to work with will sit alongside underwriters and brokers, map pain points step by step and admit where their own understanding is thin. “I'm not looking for a disruptor; I'm looking for a partner,” she said.
