First principles: why insurance is a relationship business
Insurance is a relationship business. It will be for a long time. And that’s the best thing for everybody, especially insureds. Some of you don’t need any convincing about this… but some of you do. We have encountered plenty of executives and insurance technology leaders who are dismissive of this aspect of insurance, and they hold that attitude to their own detriment.
Let’s try to boil this subject down to first principles, so we can get comfortable with why continued investment in relationships is warranted.
Fragmented value chain
The first and most unforgiving principle is this: unless you can compete and win across all the elements of the value chain, across all your product lines, in all your markets, then you’d better have partners who can.
For argument’s sake, let’s define the insurance value chain as follows: originate → advise → place → price → pool → service.
Your firm may be a leader in advising and placing Product X. But if a customer walks in the door asking for Product Y, you need a wholesaler! If you are a wholesaler, you may specialize in placing all sorts of risks… but do you maintain the actuarial acumen to price them yourself? No? Then you need carriers who do! If you are a carrier who specializes in pricing Product Y, can you pool risks with the capital efficiency to compete with, say, AIG? No? Then you need a reinsurer! If you are a reinsurer with a globally diversified balance sheet, then can you originate and advise for Products X, Y, Z (and hundreds of others), across all markets, at scale? No? Then you need to source business from thousands of agents/brokers/carriers!
Some large carriers have responded to this situation by vertically integrating. Makes sense on paper, but in practice, how often do you see a mega-company outcompeting a small specialist? Pretty hard to do. There are always partners willing to work with a good specialist. So the vertically consolidated carrier finds itself having to win on hundreds of battlefronts, while the specialist only has to win on one.
Both sides have skin in the game
When a book of business runs profits or losses, every party touching it has to have skin in the game for a good outcome. Sliding scale commissions to agents, quota share commissions to carriers, etc are the mechanisms which enforce this. The upshot: you have to look out for your partners’ interests as well as your own.
Your market may be softening now. Should a broker use that to their advantage to keep chopping rates, thereby jeopardizing the actuarial soundness of his partners’ books? A good broker pauses here and considers what might happen when the shoe is on the other foot!
Results and credibility take time to materialize
Insurance policies have long tails. It can take years — even decades — before the final performance of a book of business becomes known. This slow filtering over time drives out the pretenders, and only the credible survive. You need to form relationships with these people. You just can’t manufacture their expertise out of thin air.
A huge amount of knowledge is subjective
The veterans in any market know the good risks from the bad. A huge amount of this knowledge is subjective: there’s no clean structured data (at least not typically) to tell a machine why Company A’s general liability policy is good, while Company B’s isn’t. But the broker who has met with the management teams from Company A and B for the last 15 years has enough data points in his mind to triangulate on the right answer.
Everyone has opportunities to misbehave
There are plenty of gray areas in insurance. Claims that end up in litigation, when they could have been paid; agents who advise on risks they don’t fully understand; brokers who abuse their position in the market with rent-seeking behavior.
Bad actors get smoked out pretty quickly in insurance, and they suffer consequences at the hands of their partners. Relationships create a self-policing function in the industry which is more powerful than any regulatory influence.
The cost-benefit equation of automation gets ugly fast
Around this point in the discussion, we usually hear tech people pipe up and say “well but really all these functions in the value chain can be automated, so it’s just a matter of time before they do.”
Here’s the counterargument: beyond a certain degree of product complexity, the cost-benefit for automation gets so out of whack that humans are the better solution. It makes sense to automate highly-standardized, high-volume products (think personal lines). But how many carriers have you seen try to automate BOPs? Very few! What about CPPs? (Do I even need to ask about speciality lines?)
This is because automation gets difficult fast with more complex products. As in, really difficult. Seriously. I should write a separate blog post about it. The difficult part is not logging a policy’s information into a policy admin system. Rather, it is making good decisions along the way in order to tailor the policy to the exposure and still compete on price. Complex insurance products give you a decision tree with thousands of potential paths, some of which may be tragically incorrect. Until a true AI comes along, humans will perform this sort of work better.
Don’t believe the hyperbole
Take some time to internalize what we’ve said here. Then, when a senior underwriter at a prestigious New York carrier tells you “these brokers are all going to get disintermediated, it’s just a matter of time,” don’t believe her. And when a big-name London broker tells you “these carriers are all going to be reduced to dumb money price-takers, it’s just a matter of time,” don’t believe him either.
The future of insurance still belongs to companies who elevate their relationships and take care of their partners.
