by Angus MacCaull

In 2014 when we launched this blog, the buzzword in business was digital marketing. Companies were learning that their brands could live online as much as—or more than—they do offline.

Now the buzzword in business is data science. Companies are learning that the behavioural information they collect while delivering products and services can be worth as much as—or more than—the products and services themselves.

I’ll say that again. And I’ll bring it home to our context in the independent broker channel.

The behavioural information we collect while selling insurance may eventually be worth more than our commissions.



Here’s how it works. Everything we do with a policy is tagged as a particular kind of transaction in our business management system. With a big enough record of transactions of the same kind, data scientists (increasingly powered by AI) can cross-correlate kinds of transactions with other kinds of transactions—both inside and outside of the data set.

Though my academic background includes doing exactly this kind of work in applied linguistics, I have to give a shout out to Seth Zaremba at Zinc Insurance and Sydney Roe at b atomic for bringing it into focus for me in the context of what’s happening right now in our industry.

The general principle at play isn’t any different than insurance in the past. Look for a pattern that indicates a second pattern. Something about behaviour that signals or does not signal a degree of risk. It might be speeding tickets. An old oil tank. Or metrics gathered through smart devices connected to a car or a home.

The new behavioural patterns that insurers are looking for include things like the number of times a client calls or the number of days after a particular kind of transaction that a client cancels. And they don’t even need to ask brokers for it. They can ask business management system vendors.

Can this information really be worth more than brokers’ commissions? The data from telematics may provide a relevant example. When I signed up for telematics, the carrier collected a few months of my driving behaviour. The data included things like hard braking events and hours of nighttime driving. It was ultimately worth 21% of the premium.

Some of the biggest potential is in emergent patterns that no-one knows yet. That’s why the data is so valuable. We don’t even know what we don’t know about what signals risk in the behaviours we’re already tracking but are only beginning to amass at scale.



As someone who works in the broker channel, I have another way of looking at the broader social context of all this data. The information brokers collect while selling insurance forms the basis of a human relationship. In conversations about protecting cars or homes or businesses, brokers also learn about clients’ families and communities. About their kids and causes. Their hopes and fears and dreams.

These interactions may already be worth more than brokers’ commissions. It’s just a different kind of value. It’s about having a purpose and a role in life. I believe that brokers who know their clients are more likely to be good advocates. They care in a client’s time of need. And what’s more valuable than waking up each day and caring about the people around you?

Behavioural change often starts with an implicit or explicit promise. We’ll hear more and more in the coming years about risk management. In addition to feeling cared for, I also believe that clients who know their brokers are more likely to be good risks. They’re more likely to take care of their assets and less likely to defraud the system.

But we’re just starting to measure all these things as an industry. Time and the data will tell.