• by Ted Smith , August 20, 2010

"We'd be best off if we just showed one ad, the perfect ad."

This quote is attributed to Google co-founder Larry Page. If we did serve that one perfect ad, we might not even call it advertising, because advertising has become synonymous with interruption and surprise.

Health advertising is a three-dimensional chess game, and picking one plane of play says a lot about what you think about the others.

Level 1: Advertisements in search of a target

Level 2: Advertisements as experiential concepts

Level 3: Advertisements as information aligned with user context

Level 1 arguably has been the Holy Grail for interactive media because the numbers are just waiting to be tabulated and optimized. I'd like to suggest that the odds of getting to "one perfect ad" using an algorithm that optimizes click-throughs requires waste and, perhaps more importantly, liquidity of ad creative supply. If the goal is an ad welcomed by the recipient, we miss more than we hit with this approach. The robots have done a remarkable job with cost effectiveness but in many ways have given us permission to ignore the millions of individual "misses" that tune the algorithm.

The second level of the chess board is represented by old-school ideas like advertorial or immersive brand experiences. This approach carries the risk of overestimating the relationship a consumer really wants to have with a brand. This is particularly true online. In health, these initiatives walk a fine line between being supportive and completely misunderstanding the role disease plays in consumers' lives.

The third level brings deeper knowledge of the recipient into the proposition. Consumers who signal or even explicitly request communications from specific brands are directly increasing the odds that a marketer can find a welcome, efficient approach. The extreme examples of this make the industry news every few years. Most recently, Bynamite has entered the market to allow the consumer to specify the circumstances under which personal data and preferences can be shared with advertisers. This comes at a time when the Federal Trade Commission is increasing scrutiny of passive consumer profiling.

A related factor in these high-affinity matching scenarios is the absolute scale in which these frameworks operate. As we all know, many of the products in our category need to be introduced to consumers who may not be familiar with them but who could welcome that information. Perhaps the answer, then, is not getting more user disclosure but increasing contextual or psychographic matching. This approach to the third level of play -- bringing new approaches to user interests into the equation -- likely presents the most upside.

In health, where concerns about privacy and customized messaging are probably the highest in the industry, the odds of placing one ad informed by user intent are increased by both scrutiny of the ad's content and the need to place relevant ads in informational context. The irony here is that the best next step may be to evaluate the informational value of advertising in the largely transactional world of health information gathering.

Certainly, Google pioneered much of this algorithmically with AdSense, but I would argue it goes much further. Short of tackling user permission at very high scale, the informational content of advertising combined with informational contextual matching is a frontier worth exploring if we want to take the one perfect ad challenge seriously.

The words of Google's other founder, Sergey Brin, affirm for me that the path to the one perfect ad is likely to be one part targeting and two parts rational value to consumers.

"We try to reduce our coverage at the same time as improving the monetization. But clearly that's not the ideal strategy indefinitely, because we don't want to end up with no ads. And in fact from a quality point of view, we now find our ads are a significant addition quality-wise to our page. They are just a very important source of information. (...) We're all the time running experiments. We run some people without any ads at all, and we know that our ads add value so we know that we're happy about having them." (July 17, 2008 Google quarterly earnings call)

A big thank you to Alex Chitu and his thoughtful analysis of this topic on the Google Operating System blog.