Photo Courtesy: iStockPhoto.com

It’s no secret that March Madness started this week. As I filled out a couple brackets, I wondered, what is someone else’s method for putting together their picks? How would their picks help me understand their potential skills as a web analyst?

First, a disclaimer:   I am by no means an expert at picking a bracket. There have been years when I won it within my group and years when I finished dead last and had completely lost hope before the tournament got down to the Sweet 16 level.

With that, I thought I would briefly break down the different approaches I’ve seen people do over the past few years:

1) Simplify the Data

Maybe your bracket has both Michigan (East) and Wisconsin (Midwest) making it to the final 4 (both #4 seeds) with the Wolverines moving onto the championship game based on the fact that Michigan won the head to head match 59-41 during the Big Ten Season. You don’t look any further, you just know that Michigan won so it makes sense they would again. You don’t check any further to see if a key player was injured for the Badgers, where the game was played, intangibles, etc.

If you take this approach as a web analyst, it can lead you down a dangerous path. You’re likely to make decisions on data that hasn’t been segmented, questioned, or fully understood. For example, you might look at conversion data as a whole (orders, leads, etc.) instead of breaking it down by specific traffic or promotional source.

2) Overanalyze the Data

This mistake is common to horse racing as well. If you overanalyze the data, you know that not only did Michigan beat Wisconsin 59-41 during the regular season, you’re aware that the Badgers couldn’t get any shots to drop (of course the fact they had only 41 points should give you a clue there).

3) Just Go With the Favorites

If this is your strategy, it’ll work about once every ten years (a few years ago I won my bracket by taking a lazy “all top seed”) method. An example of applying this methodology as a web analyst would be conducting a head to head test on a promotion, but making the (often) erroneous assumption that the results can be applied like a blanket to all your marketing activities.

4) Pick all Underdogs

There’s nothing more American than picking an underdog, right?  Sometimes I like this strategy. With apologies to the perennial top seeds, it is kind of fun to see the whole bracket get shook up when an underdog takes an early game.

So if you apply this approach to your web analytics, you likely have an unconscious prejudice that leads you to look for data that supports an unconventional approach. You love the idea of standing up in front of an executive team and saying “Hey, no one thought this would work, but look at the data.”

5) Just Guess

While this can work about once every 10-15 years for an NCAA bracket, you shouldn’t plan on having a career as a web analyst.

6) Mimic the “Experts”

You know what’s really fun? I love seeing my bracket do better than some “expert” on ESPN. If you do constantly mimic the experts, you employ a “bobblehead” mentality.  As a web Analyst, this approach can work up to a point. Follow the practical advice of industry thought leaders like Avinash Kaushik & Eric T. Peterson, but do more than just copy what they do. You know your brand better than anyone so you should have your own insights into how to best analyze and interpret the data.

7) Undefined Mixed Methodology

One of our team members told me that a friend of hers based her entire bracket just on which mascot would likely beat the other one in a fight.

Another one of my team members bases it on favorite colors.  You wouldn’t think it, but this approach actually translates OK in the world of a web analyst and here’s why: You need to use both sides of your brain.  Data only tells you so much. You also need to be able to think of some really creative solutions to be able to take action on what you’re interpreting from the data.

Combo of All (or some of) the Above:

Here’s the deal: All of the approaches in 1-7 can be right and they can also be wrong at different times.  If your bracket approach contains components of at least some of these methods, then you likely have that proper left brain/right brain balance needed to successfully analyze web analytics data and take action on it.

So where do you think you fit?  Maybe you have an entirely different approach. Either way, I’d love to hear from you.Thanks and good luck with your bracket!