Kimberly Holmes: Humanizing analytics at AXA XL

A watershed moment forever altered Kimberly Holmes’s professional trajectory by teaching her analytics success began with empathy and not with a model.

In the early 2000s, in a previous role, Holmes worked collaboratively with a team of underwriters to build a new pricing model. “When I presented the model to underwriting head she responded ‘thank you, but I’ve been pricing for 20 years, so a model is unnecessary,’” recalls Holmes, who is now SVP of strategic analytics for AXA XL.

Instead of considering ways to improve the model or her presentation, Holmes decided to figure out what was important to the underwriter. “It’s the moment my job became about learning to understand and analyze the decisions underwriters needed to make, rather than about building models,” she says.

Soon Holmes went from sitting alone in her office to entertaining a constant stream of underwriters seeking her assistance. “By focusing on the critical decisions underwriters were facing, I started building impactful tools unlike any on the market at the time,” she says. “Not only were the tools embraced by management and reinsurers, the process of discovery, creativity and innovation became what I love about data science.”


Prior to the merger that created AXA XL’s predecessor company, XL Caitlin, CEO Mike McGavick asked Holmes what excited her about her job. After telling that story, McGavick suggested Holmes lead a yet-unborn advanced analytics unit. “It was a blank page,” she says. “There wasn’t even a name for the team.”

With internal and external assistance, Holmes built her 20-person team and dubbed it “Strategic Analytics” to emphasize that analytics should drive business strategy. She tackled internal skepticism and resistance by creating value quickly while simultaneously building out her team’s foundation and direction.

“No amount of talking convinces people,” she says. “Evidence is the only way to transform skeptics into believers.”

Over the next five years, Holmes’ team completed so many successful projects that measurable business improvements exceeded the cost of her department. Much as with her previous employer, Holmes and her team achieved this by focusing on creating business value, instead of developing analytical tools.

“At the point we became a profit center, we could branch out from developing tactical solutions to delivering more strategic value,” she says.

One recent example is developing a machine learning tool which automates a manual process, reducing it from weeks to minutes. Still early in its implementation, the tool is delivering an estimated ROI of more than a half-million annually. It’s also so unique that AXA XL is applying for a patent.

“It was my team’s idea and execution,” says Holmes. “I was aware they were working on the problem, but my leadership style is hiring the right people, giving them sufficient guidance to focus their efforts and empowering them to get things done.” This included getting input on a set of 10 tenets that her team lives by. “After drafting eight, I asked my team to supply the other two,” she says. “They provided so much interesting feedback that I incorporated and condensed it into the final set. My favorite contribution is ‘trust is like money; treat it as such.”

Holmes finds ample time for formally and informally mentoring others, contrasting that to her early years when women lacked mentors. She says that her husband has helped be a sounding board.

“He pointed out that highly technical people, for various legitimate reasons, are

invested in being right,” she says. “But, others don’t want to interact closely with someone because they are right; they want to do so because the person is likable.”

Now, she tries to apply those lessons in her interactions with others. “For instance, when asked for hiring advice, I tell people the number one criterion is likability,” she says. “We can teach analytics, we can’t teach likability.”

For reprint and licensing requests for this article, click here.