I started with a list of health plans with more than 100 participants (the only qualified size under ERISA), and eliminated any city that had fewer than 50. With that list, I was ready to break it down into a "top 20." The question is, how do you define "top"?

I had three options available to me using the commissions and fees collected in each city: the total amount, the average amount, and the median amount.

If I were to use the total amount, then the "top" 20 would fall almost perfectly down the city's actual size: New York, Houston, Los Angeles, Chicago, etc. Not only is this a boring list, the larger problem is that the $60 million in the New York health market that's up for grabs is spread out across 1,270 policies. Compare this to the $2.9 million in Chattanooga that's held by only 65 policies. Sure there are more accounts you could go after, but they're just not going to be as worth as much. Especially when you take the cost of living into account.

So that leaves average and median. What I wanted to get at was, if I'm a health broker and I move to a particular city, what is the amount I could expect to get from any one health plan? Because of my goal, I went with the median. Here's why:

Let's look at the personal holdings of some hypothetical residents of Bellevue, Washington:

John Doe – $50,000

John Smith – $60,000

Jane Doe – $70,000

Jane Smith – $30,000

Bill Gates – $61,000,000,000

The average of that list is more than $12 billion. However, if I were to get a job in Bellevue, it's far more likely that I'd be making "John Doe" money than Bill Gates money. And indeed, the median of that list is $60,000 – a reasonable amount.

The top three cities based on average compensation are Bellevue, Creve Coeur, and Wilmington. However, as a broker, I'd have a hard time landing accounts with Microsoft, Monsanto, or Bank of America (the "Bill Gateses" of those cities).

Using the median gives me the cities where a health broker would have the greatest chance of success, even without having to worry about landing a single massive account: Chattanooga, Kansas City, and Boca Raton.

I hope this sheds some light on why the data came out like it did, and that I've given you a better understanding of the important difference between average and median.