June 04, 2019

The Wrong Math


Daniel Kahneman, Nobel Prize winner, says people don't believe facts. They believe experts.

In some fields experts have credibility. Mostly it is in fields of hard science like medicine, physics, and chemistry where expert opinions can be tested.

In soft science, like economics and sociology, where enormous variables exist and controls are hard to establish, experts have far less credibility. There is also far less agreement within these disciplines. A quote attributed to George Bernard Shaw goes like this, "If all the economists were laid end to end, they would not reach a conclusion." Not because they are any less serious, but because their theories are difficult to prove or disprove.

Sadly in the field of advertising and marketing, experts are not usually hatched based on their record of producing reliable results, but on their ability to attract attention. Consequently we should be highly dubious of their "expertise." But we're not. Because as Kahneman also says, "a reliable way to make people believe in falsehoods is frequent repetition."

One of the most frequently repeated and, in my opinion, highly dubious tropes in our industry these days is the idea that the paragon of media strategy is "mass one-to-one" communication. In non-jargonista terms, this means reaching large numbers with individualized messages.

You would expect that this assertion would be met with skepticism. For one thing, there is no record of "mass one-to-one" communication achieving anything. You might argue that no one has yet been able to engineer "mass one-to-one" and that is why there is no record. Which is exactly my point. Shouldn't we exercise a little skepticism about a theory for which there are no examples?

All of our huge brands -- Apple, McDonald's, Coca-Cola, Toyota, Budweiser, Tide, Crest, Nike -- (I could go on here all day but you have work to do) have been created by the supposedly wasteful and sub-optimal mass media.

The power of the marketing feedback loop seems to have caused our industry to lose its ability to be sober or skeptical. Or as Kahneman might put it, facts don't matter. Experts do.

The reason we accept the fairy tale of "mass one-to-one" with absolutely no evidence is that a) experts are talking about it, and b) our math experts (in media and data) say it's true. 

I don't believe the experts, but I do believe in math. I believe math can offer us insights into how advertising works and how consumers can be influenced. The only problem is, I think we're using the wrong math. If you'll pardon my cliché, we have the wrong algorithm.

I don't know what concept of math the data experts use to persuade marketers that "one-to-one" is the media model of choice, but I believe the math model we should be using to understand media effectiveness is probability. In other words, what media strategy is most likely to produce the desired result? For large consumer-facing brands, there is ample evidence that (the prudent use of) broad based media has the highest likelihood of achieving the desired result of building substantial brands, and almost no evidence of anything else doing so.

The mathematics-based rationale for the primacy of mass one-to-one advertising and its alter ego precision targeting seem to go something like this: a) you are not wasting money on people not interested in your product, and b) customized ads are more relevant and persuasive.

This may be true for certain types of B2B marketers and highly-specific brand categories, but I think both these rationales are wrong for mainstream brands. I think probability would tell us three reasons why they're wrong.

First, I believe brands are far more likely to achieve big success if they are well-known. Public media (broad based media) make you well-known. Private media (one-to-one) don't. Perhaps the best argument for this can be found outside the advertising industry. As many have noted, in their early stages Google, Facebook, and Amazon were brands that became successful without advertising. How did they become successful? One component was that news media fell in love with them and gave them zillions in free coverage. These companies became well-known without advertising, and being well-known helped them grow. The rules of probability don't just apply to advertising, they apply across the board.

Second, I believe people are more likely to accept the legitimacy of brands that advertise in public than brands that advertise in private

Third, except for sociopaths, we all (secretly) want to fit in. Understanding what products fit with our peer culture is part of fitting in. This is why goths wear black and golfers wear plaid. Consequently, we are more likely to buy a brand about which everyone in our group knows what the brand stands for. Public media provide the framework to believe that your group has the same understanding of what the brand is about as you do. Private media do not. When advertising is customized for individuals, we have no idea if others know what we know.

Byron Sharp tells us the key to growing a brand is acquiring new customers. I believe probability tells us that the more people we communicate with loudly and in public the more customers we are likely to acquire. 

Another way to look at this...

The great Rory Sutherland says that "A flower is just a weed with an advertising budget." His point is that flowers expend a lot of resources to look and/or smell pretty. And about 125 million years of evolution have shown that the expenditure pays off. 

If there was a superior way for a rose to attract bees by individually or precision targeting certain types of bees with certain types of attractiveness, one would assume it might have evolved by now. Instead, roses produce a lovely, fragrant flower and let probability do its work.

Only time will tell if "mass one-to-one" is the formula for building big brands. I'm betting the under.

No comments: