Don't listen to what they say...see what they do.

Gord Hotchkiss wrote an interesting entry today on the decision making process and the short comings of market research. It reminded me of my time in CPG advertising where we delved into market research and tried to reconcile the information with what we saw in the SRI scanner data. Essentially, people were telling us that they were exercising more and eating healthy foods while cutting back on things like soft drinks and potato chips. Yet, according to the scanner data, these unhealthy foods were being consume in growing numbers. In other words, people were saying they were doing the smart thing, but actually doing something else. Gord's piece gets into the why of this and how it applies to search.

But, I think we all have it in our power to address this tendency of consumers to think they do one thing, while actually doing another. With search, we have the ability to track and test ad nauseam. According to the piece, people said they read the listings, but actually only scan the listings. So, what is better, a long sentence (okay, 70 chars isn't long, but can be well formed), or short phrases or individual words that are very targeted? We can test until the cow come home to find out WHAT people do, rather than what they say they do, or we think they do, or our bosses guess they do, or the client supposes they do... ad nauseam. 

People say they carefully read all or most of the listings. But, most of the activity (attention and clicks) go to the top. So, does this mean the 6 and lower listings are not valuable? No. Measured on a cost per, you may want a lot of 6 and over position ads rather than one or two 1-3rd position ads. If the bulk of the volume is up top, so is the competition. So, see if the 6+ performs. It may be better. Test.

The point is, while we should continue to read and absorb as much research and information as we can, our ultimate decisions must be based on real world actions, not academic inquiry. Fortunately, for us, search allows us to do this with our own experience, control and data. Take advantage of this. Let articles like Gord's provide the catalyse for new questions, direction and exploration for improvement rather than as quick answers.