Immeasurable Impressions

One cost of traditional advertising is never knowing for certain what worked  —
 but it’s still worth it.

There’s been a resurgence of an old debate on some first principles of advertising — can it ever be quantified, measured, and justified with any certainty?

One side champions the unmatched effectiveness of digital, with its purportedly sophisticated analytics and efficiency owing to highly specific targeting and tracking. Digital advertising still seduces with a siren song of something once thought impossible — a quantifiable system of metrics that can scientifically determine every ad’s performance.

Well. Digital advertising’s reputation isn’t in the best place right now. Thanks to staggering amounts of ad fraud, those vaunted scientific numbers aren’t held in much esteem.

Which is why the opposite view is gaining steam. That other side favors an older-school estimation of advertising. Namely, that no brand has ever been built with click ads and paid social media posts, which can’t hold a candle to the power of above-the-line traditional advertising like broadcast, radio, and print.

But here’s the thing. This faction also believes that predicting and gauging the success of the most powerful forms of advertising is more about probability than arithmetic. In other words, they’re resigned to the reality that advertising’s true effectiveness can’t be measured with certainty.

How Did We Get Here?

I just spent some time with a beautiful old oversize book called Advertising in America: The First 200 Years. This book is brimming with colorful images of ads, tracking the trends and evolution of the industry right up to its 1990 publication.

The authors, Helen Dalrymple and Charles Goodrum, begin by helpfully defining the purpose of all ads:

“They want 1) to get our attention; 2) to convince us that what they’re selling is desirable; and 3) to get us to go out — rise from our chair or couch — and do something. Half the fun is to ask, Did it work?”

The bad news, they argue, is that there’s really no way of knowing for sure if an ad worked at all.

Sure Ads Work — Just Don’t Ask for Proof, OK?

First, it really is true that traditional advertising is far more effective than digital. Ironically it even has the numbers (real ones!) to back it up. If you’re going to spend money on advertising, big-media advertising has a higher probability of having positive effects.

One big reason is “signaling.” Traditional advertising’s vastly higher budgets signal to consumers that any brand utilizing it is committed, confident, and competent enough to have sufficient money to spend on advertising. So that expensive Super Bowl ad, for example, implicitly communicates that the brand is successful and trustworthy.

A different but related idea is proposed in the essay “Ads Don’t Work That Way” by writer and computer engineer Kevin Simler. He identifies “cultural imprinting” and the logical principle of common knowledge as the reasons for traditional advertising’s superiority:

“For an ad to work by cultural imprinting, it's not enough for it to be seen by a single person, or even by many people individually. It has to be broadcast publicly, in front of a large audience. I have to see the ad, but I also have to know (or suspect) that most of my friends have seen the ad too. Thus we will expect to find imprinting ads on billboards, bus stops, subways, stadiums, and any other public location, and also in popular magazines and TV shows — in other words, in broadcast media.… Internet search ads and banner ads are inimical to cultural imprinting because the internet is so fragmented. Everyone lives in his or her own little online bubble.” [Emphasis in original.]

As Simler states, this explains why click ads can’t establish a strong brand or budge brand perception. If anything, a banner ad catches a consumer way further down the line and offers a convenient opportunity to act, but the internal justification and emotional response happen earlier, on a far broader scale.

The upshot of this is a conundrum. The most effective advertising is much more expensive than digital advertising, while the very cheapness of digital serves as the rationalization leading so many to rely on it to do things it can’t do. Meanwhile, the success of the most powerful ads (i.e., traditional media) can’t really be accurately measured with any degree of certainty. Again, it’s more of a probability game.

Actually it’s arguably worse (less reliable) than that. Because even campaigns perceived as being wildly successful may be harmful to brands.

You Never Can Tell

In their introduction of Advertising in America, Dalrymple and Goodrum point to a run of commercials that aired for Alka-Seltzer in the 1970s:

These award-winning Alka-Selter ads were wildly popular, yet, as the authors explain, “The more those ads ran, the more Pepto-Bismol sold.” After getting crushed repeatedly in sales, the enormously “successful” campaign for Alka-Seltzer was pulled, to the great surprise of the ad industry.

How could that be? The authors have a theory: “We all laughed at the miserable victim, but when we became miserable ourselves we thought, ‘Alka-Seltzer thinks it’s all a big joke and doesn’t take my symptoms seriously. I feel awful. I want something serious to fix me up. Better by Pepto-Bismol.’”

Everybody with a TV knew those Alka-Seltzer ads. Meanwhile, Pepto-Bismol didn’t have an entertaining ad campaign with relatable characters and funny lines — just a no-nonsense message that it does the job when you need it.

Advertising in America offers other examples:

Cigarettes: “The first full year after tobacco advertising was taken off TV, the sales of cigarettes went up 3 percent and the manufacturers were able to save $70 million that they’d spent on commercials just two years before.”

Cars: “The lying Isuzu car salesman [Yeah I hadn’t heard of it either — it’s this campaign] was the most remembered of all television commercials two years running with the highest ‘sympathetic’ response from the viewers researched. During the same two years, the number of Isuzu cars actually sold sagged steadily downward.”

“Prestigious” Ad Awards: The authors write that David Ogilvy “was bemused by Harry McMahan’s analysis of the winners of the prestigious Clio Awards…. Ogilvy reports: ‘Agencies that won four of the Clios had lost the accounts. Another Clio winner was out of business. Another Clio winner had taken its budget out of TV. Another Clio winner had given half his account to another agency. Another refused to put his winning entry on the air. Of 81 television classics picked by the Clio festival in previous years, 36 of the agencies involved had either lost the account or gone out of business.’”

Getting Philosophical

Advertising in America then poses an interesting question and offers a provocative answer: “Can’t research and polling tell which ads ‘worked’ and which didn’t? No, unfortunately, they can’t. There is still no way to determine whether it was the current campaign that got the product sold, or residue from previous campaigns” (or no campaign at all, I’d add).

So if data shows that traditional is the most effective form of advertising, how is it that the success of individual traditional ads can’t be measured with certainty? Again, the authors have a theory: “Because no one knows what a really good advertisement looks like…. With an advertisement, there is an endless, almost metaphysical number of ways it can be created but no way of knowing which one is best — or even which one works at all.”

Even if a campaign becomes an iconic part of culture, it can backfire. It’s another confounding reason good advertising isn’t as easy to make (or to recognize) as we might think — apparently, some publicity is bad publicity.

In the lovely short film Sunshine, commercial director and filmmaker Doug Nichols posits that advertising is filled with frustrated artists who once dreamed of purer creative endeavors. Maybe they’d find it comforting to know that advertising, as tawdry as it can be, is still more of an art than a science. And that’s not likely to change.

Posted on February 21, 2019