speaking in clicks

[August 22, 2011]

Years ago, the pioneering English marketer, William Lever, is supposed to have said, “I know half my advertising isn’t working, I just don’t know which half.” More than a century later, the issue is still irresolvable. That’s because advertising is a process, not a one-shot quick-fix for slumping sales. It’s only one of many contributors to the ambient emotional, cultural and intellectual atmosphere surrounding every product. 

By corollary, buying itself isn’t a decision, but a reflex—a response to a body of stimuli, including previous experience, word of mouth, cost and brand identity. Accordingly, you can’t tell “which half” of your advertising drives sales because that’s not what advertising is meant to do. At the same time, without well-crafted advertising, your sales would be hobbled by the absence of a key-influencer. Because advertising does work, just not like a magic wand. 

Despite the claims of placebo-toting consultants, the impact of advertising is ambiguous—and, as I see it, that ambiguity is its greatest asset. In light of that, there’s something both charming and sad about the quest by Yahoo Research to analyze the impact of advertising on sales. 

This ongoing project is flawed, in part, because it fails to define a coherent quality standard for the advertising it plans to test. To yield meaningful results, Yahoo can’t study the impact of Advertising as an abstraction. It can only evaluate the relative impact of one particular ad at a time. Without a mechanism to ensure the ads meet industry standards, the resulting data will be completely meaningless. After all, we can’t expect a smarmy, pretentious ad for the iPhone to sell…wait, bad example. 

Methodology aside, a deeper problem lies in the misuse of words like “data” and “analysis.” 

Objective questions, subjective answers. 
For starters, everyday experience teaches us that the worst way to find out what motivates people is to ask them. The immense social pressure applied to every aspect of American life means an honest answer is unlikely or even impossible. A mind shaped by the lifetime of denial required to meet our strict social norms can’t be expected to deliver objective observations. 

That means any data acquired through a question/answer process—whether “cross-checked” or not—can’t claim the mantle of scientific rigor. Yet as AMC’s Mad Men reminds us, we’ve been fooling ourselves on this score for decades. 

That’s not to say there’s no value in gathering information and trying to interpret it. What else to we have to work with, now that American society is fractured by an incredibly obtuse political discourse about its identity? But the key word here is “interpret.” Qualitative or quantitative, marketing data delivers only a fuzzy snapshot of reality that can’t be analyzed with scientific precision. 

So given that the course of American culture is now about as predictable as the shape of molten lava, I have a hard time understanding why we continue to clamor for such specious statistical rigor. Maybe it’s a measure of the anxiety this state of affairs produces—a “rage for order” that’s sweeping the nation. Or maybe it’s a reflection of the decades-long trend toward injecting pseudo-science into all aspects of American life. Either way, the belief that human behavior follows a predictable curve is inherently toxic. 

The side-effects of a doubtful prescription. 
Between those who advocate interrupting a brand narrative with the literal repetition of search terms, and those who hear the voice of God in user-testing, we’re drowning in rigidity. Clearly, in the trickle-down process between valid scientific research and its assimilation by advertising culture, an essential bit of perspective has been lost. 

It boils down to this: No matter how careful you are, the act of observing human behavior transforms it. Besides, in user-testing facilities we can only observe how people behave in an artificial setting, not in their “natural environment.” Between the social pressure to perform—we are, after all, asking people to test something—and the social pressure on analysts to justify their paychecks, the resulting data is hardly pure and objective. 

Results like this are not rigorous enough to justify multiple, panicky revision cycles until everything fresh, engaging, intriguing, mysterious, thought-provoking—human—has been erased from our work. It’s time to acknowledge that the unsatisfying gruel we serve up to consumers is the direct result of this infantile longing for absolute, lifeless safety and comfort.  

In spite of that, we continue to produce reams of work that mirrors the mentality of our test subjects—people in the prime of their lives with nothing better to do than lounge around in a conference room and complain. Sad to say, with our reliance on pseudo-science, we’ve not only failed to evolve past the era of snake-oil advertising—we’ve turned its harmful influences back on ourselves.