Monday, March 28, 2011

Behavioral Target Marketing or A/B Testing – A Head Scratcher

I just read an article that pitted these two tactics against each other and asked which was better.
I scratched my head on this one. Why forego either?
If I’ve learned anything in my 20+ years of marketing, operations and management (which I view very connected, btw) it is that ‘getting it right’ is like peeling an onion. And if you can, you peel two or three layers at once and see where your highest gain is.
Example: If I had both options  behavioral marketing and A/B testing available to me, I’d use both!  Many times with clients I will recommend a tactic and they automatically assume (and most state) that I am ruling out other tactics. The truth is that couldn’t be further from the truth!
When I’ve done campaigns with behavioral data available to me, I still do an A/B test to a small, but statistically relevant group before I launch the campaign. Why use one tactic when you can test, use both tactics and identify the winner?
To get the very best ROI for my clients or company, I want to make sure that when I “pull the trigger” with the big gun (the campaign budget) that I am confident that I have identified the very best strategy. I want to yield the highest return that I can and this can be an area to navigate with some clients or management folks.
Unless you’ve used all forms of marketing (online and offline), selecting one tactic for the desired outcome is risky. Plus, depending upon your product or service, you may truly need an integrated approach to completely ‘touch’ all of your prospects. Many times, one size does not fit all.
I love having behavioral data at my fingertips. However, you will not find the golden key that unlocks the door to all of your customers. You will find identifiers which can help you with imaging, placement, messaging and more, but you should try at least an A/B test of differing approaches that fit within the characteristics of the behavioral data and see which one wins.  If you get a true front runner, that’s your winner and launch your campaign with the winner. Sometimes though you will find the results similar and this tells you that you’ve hit good messaging and look back to the responders and see what differs between the A responders to the B responders and then peel your data back one more layer and separate the “A type” from the “B type”.
For example, let’s say you find that the A responders were in a different geo target than the B responders. It would be silly to hit both geo areas with the same media. By testing behavioral data, you can enhance the data and thereby increase intelligence to the data for further campaigns.  

No comments:

Post a Comment