My career as a footballer was short, but it taught me some invaluable lessons still I’ve applied throughout my life. During my third and final season playing, a new coach was brought in. Management hoped he would revive my team’s fortunes following our tragically poor performance so far during the year. Our record of 0-1-6 was the envy of no one.

Monsieur Dedieu pulled no punches. “Vous êtes tous des frimeurs”, he said. We were show-offs, he said. Too focussed on our fancy dribbling and fast breaks, paying too little attention to the object of the game: working together to put the ball into the net.

We spent the next hour passing balls back and forth. I was 12.

By taking us back to basics, Dedieu actually managed to perform something of a turn-around. We won more than half our remain matches that year. We quietly mocked opposing teams that didn’t seem to understand how to pass, shoot, or cooperate.

I’ve never forgotten what he taught us. When goals seem more distant than usual, the best course is very often to go back to basics. A flashy short-cut rarely changes the course of direction of travel.

Apparently US commercial motor insurers have learned this, too. A friend kindly sent me a late-October report from the desk of US ratings agency A.M. Best. Entitled “Commercial Auto May Finally Be Seeing Light at The End of The Tunnel”, it reports that underwriters covering commercial vehicles and fleets have partially offset rising claims through greater efficiency, which “has driven underwriting expense costs downward moderately and fairly consistently during the past ten years”.

However, AI isn’t the secret to this success. Instead, American motor fleet underwriters collectively went back to basics. Bests cites:

• Underwriting expense control
• Loss ratio improvement from
  • more effe­ctive and proactive fleet management
  • data-driven behaviour tracking and associated training/coaching, plus
  • advanced driver assistance technology

Fair enough, the telematics may deploy some AI, but note how Best’s analyst notes that it was traditional driver training required as a result that drove down claims, not some kind of magic dynamic premium adjustment routine (although that, I admit, strikes me as a very cool possibility).

As an MGA CUO, I reminded my team often that the secret to profitable underwriting was fourfold:

1. Spend no more cash or time than you need to
2. Charge the right price for the risk
3. Practice proper risk selection
4. Force risk management onto insureds where possible.

It’s a back-to basics agenda that doesn’t include anything flashy.

Even then I had no doubt then that spending a little more cash on technology would allow underwriters to spend a lot less time on mundane tasks (although the absence of a platform that would allow it is what drove me to launch Quotech). However, I didn’t believe then that silicon could provide any sort of bottom-line panacea, nor do I now.

Telematics technology (or the equivalent) that is used to drive risk management behaviour can help with claims. However, it won’t improve your bottom line if you undercut everyone else to push up the top, or if you blow all your cash on an AI project that you expect to deliver a panacea (it won’t).

AI can do magnificent things, but it’s no substitute for underwriting discipline. It can help underwriters to understand the risk, and therefore support their arrival at the correct price, but it cannot force underwriters to charge sufficient levels of premium.  

Getting those parts right is just basic.