• by November 19, 2004

REST IN PEACE - Well, it's official. The demo is dead. And it actually died this week. How do we know this and how can we pinpoint the precise moment of demographic demise?

Well, because Crain Communications, arguably the leading publisher of trade magazines focused on the business of marketing, announced that it had acquired American Demographics magazine from Primedia (the leading killer of trade publications focused on the business of marketing), and promptly announced it would discontinue publishing the monthly magazine effective with its current issue.

Sure, Crain says it plans to utilize the AD brand in other ways, such as editorial content in other trade pubs like flagship Advertising Age, but from where we see it, they just killed the magazine that has been the leading chronicler of demographic trends, and their implications on both the world of marketing and society at large, for nearly a quarter of a century. And in a sad way, they may be right. The demo was destined for demolition.

The truth is, demographics - the segmentation of people into neat little clusters based on their age, sex, income, and some other criteria - never really made sense. It didn't make sense from a social analysis point-of-view. It didn't make sense from a marketing point-of-view. And it certainly didn't make sense from a media planning point-of-view. It just kind of happened.

And it kind of made sense at the time that it happened, because we didn't have a whole lot else to hang our hats on at the time, so we all clung to the demo as some sort of divining rod of consumer patterns.

The whole thing may have been Ollie Treyz's fault. Treyz was a smart, and extremely pragmatic president of the ABC Television Network during one of its golden days back in the 1960s, and he is considered the man who - while he may not actually have invented the demo - certainly gave it commercial substance.

At the time, TV had emerged as the dominant medium for advertisers who were shelling out quite a few shekels for little snippets of commercial time on the basis of how many households Nielsen estimated they reached.

Those were the days of "boxcar" buys. It wasn't about who you were reaching with your media buys, just how many. And that was okay, because we were still riding off the great wave of mass marketing that crested in the period following World War II and through the 1950s and 1960s. The problem was the business got more competitive and considerably more cluttered.

TV, which had been a relatively clubby medium of a few fat cat advertisers "sponsoring" a few fat cat networks, suddenly exploded into a multitude of brands competing for a finite pool of "spots." At the time, ABC's spots didn't reach nearly the boxcar loads of viewers that CBS and NBC did, but it came to Treyz's attention that ABC's shows reached a higher concentration of certain kinds of viewers.

These viewers just so happened to be the kind of young, urban, and family-minded viewers that many big marketers actually wanted to reach. So Treyz did something that changed the world of both marketing and media. He asked his research team to work out a deal with Nielsen to begin reporting the demographic composition of network TV shows. And we've never been the same since.

The number and variety and nuances of demos have evolved considerably since the early broad demos pitched by ABC. They were even challenged, but ultimately ended up incorporating a potential rival during the 1980s and 1990s -- so-called psychographics that portended to go beyond age, sex, income, education, and occupation, to include the lifestyle patterns of consumers.

And in some ways, virtually every form of advertising media ultimately incorporated the demographic sell into its pitch. Marketers accepted it. And agencies planned, bought, and posted on the basis of it.

The problem was that demographics were never an especially good surrogate for actual consumer targets. And sometimes they were really bad ones. We kind of knew it and could occasionally prove it, but the whole thing worked so neatly that we stuck with it until it became embarrassingly obvious.

The first signs occurred during the first golden age of media planning that started in the 1960s at such shops as BBDO. (If you care to read more about this, the whole thing is documented in the soon-to-be- published December issue of MEDIA magazine.)

At about this point, people began to notice that some slippage occurred between the time that brand management departments developed their sophisticated profiles of the consumers they were looking to reach. These were then turned over to media planners who figured out which media were best at reaching them, which were then turned over to the buying departments to negotiate the best and most cost-effective buys to reach them.

When the whole thing was posted, some realized that those reached by their media buys often did not have much to do with the people those brands intended to reach. This phenomenon was described as the "Great Media Disconnect," and it only grew worse as media got more niche and more highly targeted themselves. And let's not even get started on the super-enhanced targeting of the Internet.

While it became eminently clear the demo was a crappy surrogate for consumer targeting, nobody did anything about it; they just talked and danced around. They came up with ways of factoring, or modifying demographic analysis that would bring it more in keeping with the way they analyzed consumers.

They appended or ascribed different databases to them - things that correlated actual brand preferences and buying patterns with demographic composition. They even came up with incredibly sophisticated "geo-demographic" clustering systems that could retrofit demos into ZIP code-level data that could then be ascribed to all sorts of neat marketing info based on U.S. Census data and other cool stuff.

And during this period, American Demographics reigned supreme. Never mind that its title connoted a geographic bias itself. The title became the most well-read and influential trade magazine on and off this continent for anyone who wanted to understand the patterns and science of demographics.

Many people no longer want to understand, or even believe in those patterns of science because new, better, and far more empirical data has come to light that shows just how much demos actually suck.

There was an inkling of this during the 1980s when Arbitron and then Nielsen owner Dun & Bradstreet tried developing the first generation of so-called single-source measurement systems - systems that simultaneously measured exposure to media and the purchase of products.

Arbitron's ill-fated ScanAmerica system even coined a name for this new empirical method of consumer clustering. It called them "buyergraphics" and it was a liberating idea for brand managers and media planners alike. For the first time, they would be able to plan and buy media on the basis of who actually bought their products, or didn't buy their products, or bought their competitor's products. It was heady stuff.

But it was uneconomical, at least on a national scale. So Arbitron folded ScanAmerica after investing more than $100 million, and Dun & Bradstreet ultimately spun off its marketing and research divisions into separate companies - which ironically was reacquired and consolidated under VNU, which ironically is now teaming with Arbitron to launch a new national, and ostensibly better and more economically viable single-source measurement system dubbed Apollo. And people think this one has legs because it is backed by Procter & Gamble.

Asked what they believe the future of TV audience measurement will be five years from now, a panel of well-regarded researchers speaking at an International Radio & Television Society panel on Thursday said they thought Apollo had a good shot. This is partly because it's backed by P&G, but also because it makes sense from a marketer's point of view; it enables TV - and also radio and potentially other media - to be measured on the basis of the actual consumers they reach, not just the demographic surrogates for those product purchasers.

Clearly, Apollo has its limitations. But if similar research is any indication, it is light years ahead of demographics in terms of targeting consumers. Over the past several years, single-source-like data has come to light via another research technique known as "fusion," which mathematically fuses media audience data with product purchasing data. It has shown that demos are weak indicators of what people purchase.

In a now famous analysis conducted by planning guru Erwin Ephron that is based on data from Kantar's MARS database, Ephron found that the difference between demos and the actual targets for products in the pharmaceutical category can be as much as 25 percent. That's a lot of wiggle room, and not the kind that the Six Sigma black belts running marketing organizations and media concerns are about to tolerate.

The only question we have is, now that the demographic species has officially gone extinct, what will the network spin-doctors and TV critics debate during their Nielsen sweeps briefings?