A few years ago, the industry’s biggest worry was how to recruit enough talent to fill the gaps and prepare for the future of insurance. We talked endlessly about how to make insurance look interesting to people coming out of school, dedicated panels at every conference and leadership meeting to it and dedicated marketing plans around talent recruitment.
Then, almost overnight, the conversation flipped and the new worry is how AI will wipe out the entry-level that has long been considered a rite of passage in insurance: the processing, rekeying and repetitive work that junior people grind through before they get to the more interesting parts of the job.
The assumption is that taking away that grind also takes away how people learn the business. I don’t think it does.
I did the grind
Early in my career, I spent about two years processing data eight hours a day in what could fairly be described as a closet. It was a rite of passage, and I don’t regret it.
But if I’m honest about where I actually learned insurance, it wasn’t from the processing. It was from anyone I could corner who would explain what I was looking at. Is this a data problem or is human behavior? Is it the history of the risk, or something macroeconomic or geopolitical? The processing put the information in front of me. The people, questions and context taught me what it meant.
Other professions have been through this. Accountants and bankers watched spreadsheets and trading models absorb a lot of the work juniors used to do by hand. The junior roles changed. That’s what I expect to happen here. In the right situation, new underwriters and brokers will get to the real risk questions sooner than I did.
A location in a river
And getting there sooner matters, because years of processing don’t guarantee anyone is asking those questions.
In a previous position, we had won a piece of business with a large utility whose property data had been handled by a junior person with the same firm for 15 years. In all that time, nobody caught that one of the company’s major locations had been mapped in the middle of a river. As a result, those cat model inputs became a large driver in elevating insurance costs for the client.
That’s not a knock on the person doing the work. It’s what happens when the job is mostly moving data from one place to another. Attention goes to the moving, not the meaning.
Give someone brand new that same schedule, some good data techniques, and AI on top of it, and I’d imagine they go through the same work a lot faster and flag problems like that for the client long before year 15.
That’s still an apprenticeship. It’s just different.
The better pitch to the next generation
There’s an upside here we’re not talking about enough. The case for joining this industry gets a lot stronger when the pitch is “you’ll be working on real risk questions in your first year” instead of “you’ll be rekeying data for two.”
We spent years asking how to make insurance attractive to new talent. We may have just been handed the answer.
The next generation can still learn the business the way generations before them did: by looking at risk, asking questions and learning from people who know more than they do.
They just may not need to suffer through two years of a brutal data entry grind that I did.
Topics
InsurTech
Data Driven
Artificial Intelligence
Interested in Ai?
Get automatic alerts for this topic.
