• by Mark Green , Op-Ed Contributor, January 31, 2008

I keep feeling like I am standing on top of an iceberg, unable to see what is below the waterline. At any time worldviews can flip. Then, what has been underwater sees the light of the day.

While the rest of the world was getting serious about virtual worlds in 2007, I was digging through a pile of single source multimedia consumption and product purchase data in the bricks-and-mortar world. It has been called many things: A moon shot. Too little sample for too much money. A fancy model. A once in a lifetime opportunity. And I was there when the data and findings starting pouring out in mid-to-late 2007.

Is this a big one-off experiment or a new way of managing business? It started three years ago as a call to action. Advertisers wanted ROI and were ready to test the concept. Now it's decision time.

Are the data and analyses compelling enough to warrant the investment? By the time you read this column, a decision will likely have been made to continue this venture or close it down. What's been interesting so far?

>> Targeting matters. We found that the advertising response for a brand was marginal and that the brand's targeting strategy was no better. Stuck with nothing to tell, we let our imaginations wander and investigated response rates for a couple of other target groups for the brand. Inside the total advertising response that aligned with an independent mix analysis, we found an unsuspected target group with a double-digit response rate. Extending to the next step, we mapped out the profit and loss of the brand for different target groups. Knowing the cost and response of media and other marketing techniques by target group uncovered the profitability of specific consumers.

>> Price elasticity is tied to advertising. We put the data into the hands of some smart people and asked them to investigate the relationship between advertising and product pricing. They found that price elasticity declined with increased advertising for specific target groups. Yes, targeting still matters and advertising works, but now we know that we can manage price with advertising levers.

>> Recency is not necessarily better than frequency. The last time single source data was available in the United States - a generation ago with just household level television program viewing - a handful of consultants, academics and specialists spent two years mining the data and determined that consumers responded to the last advertisements seen. They called it Recency. It got debated and refined. And the bottom-line: Media planning changed for a generation. We couldn't resist taking a fresh look with the new lens of commercial ratings from television. And we found reality was not black and white. We brought back the smart people. They examined a large brand and found that heavy buyers responded to frequency while light buyers responded better to Recency. More revelations: Advertising works and targeting matters, especially if you want to manage your media investments to optimize the product's profits.

>> Not all commercials are equal. Another team of smart people investigated competing brands to discern if certain media choices were good places to be, and what their diminishing returns looked like. We wanted to know if these media were good investments for brands in this category. Again we looked for one thing and opened a can of worms. The heavy investments that one brand was making were not particularly responsive. The competitive brand need not follow, as the other media were more responsive. Then, we took another look and found that the first brand's advertising was not competitively responsive. Clearly not all commercials are equal. In independent analysis, we saw that not all commercials had the same response curves across media. Of course, commercial pre-testing always suggested this, but now we had hard evidence. And more, we saw the consequent differential in media investment levels to achieve the same results. Clearly the thousands spent on getting good recall pays off when you don't have to spend millions more to get the same results. Instituting commercial recall and response criteria can save millions.

>> Loyalty is connected to advertising. Just last week, I saw the beginning of some new findings. It involved some fancy mathematics and suggested that brand loyalty has ties to frequency. It was fast work and there are still many unanswered questions, but these are the type of tidbits that lead to insights.

Mark Green is senior vice president for strategic measurement initiatives at The Nielsen Company. (mark.green@nielsen.com)