Marty Kaplan: The Naked Nielsens


The metrics are wearing no clothes.

How would you react if you found out that the basis of your business model was bogus?  That’s the nightmare that the television industry is finally waking up to, and I bet that online media won’t be far behind.

The TV business is built on advertising.  Except for premium cable, the money that networks get for selling audiences’ eyeballs to advertisers is the mother’s milk of the industry.  Networks set the price of ads on their shows using demographic information about the age and sex of those shows’ viewers.  And the company that pretty much has a monopoly on furnishing those metrics is Nielsen.

So a few weeks ago, at the Marriott Marquis in New York, it must have felt like pitchfork time when a respected TV network figure in charge of analyzing ratings, CBS Corp. Chief Research Officer David Poltrack, ” target=”_hplink”>Ad Age, Nielsen executives at the convention reported that “ratings demographics by age and sex had a… 0.12 correlation with actual sales produced by exposure to TV ads, where 1.0 is complete correlation and 0 signals no relationship whatsoever.”  Zero-point-one-two! You’d do better using a Ouija board than Nielsen demos. 

It’s particularly ironic that this paradigm-popping confession came from CBS.  From 1955 to 1976, before any network thought in terms of age cohorts, CBS “was the undisputed king of the ratings hill,” writes Neal Gabler in ” target=”_hplink”>Wall Street Journal, “which made them without value to the networks.”  The numbers tell the story: A 30-second ad on Fox’s young-skewing Glee costs $47 per thousand viewers, while a spot on CBS’s The Good Wife, 60 percent of whose audience is 55-plus, costs about half that. 

But now the jig is up.  “Reliance on the 18 to 49 demographic,” Ad Age reports Poltrack saying, “is hazardous to all media and marketers.”  It may be just a coincidence that CBS, which these days runs about even with Fox in overall prime-time viewership, is now being killed by Fox in 18 to 49.  But it’s no coincidence that 80 million baby boomers are aging out of the desirable demo.  To sell air time to reach the fastest-growing part of its audience, the industry needs a new metric. 

So exit demographics, and, just in time, enter psychographics.  That audience-segmentation tool, which collects people into taste and behavior clusters, has been around for a while; if you want to try an online-era version, check out hunch.com. CBS and Nielsen, in what Poltrack calls a “historic move,” have now come up with six audience segments to sell to advertisers instead of age and sex cohorts:  TV companions; media trendsetters; sports enthusiasts; program passionates; surfers and streamers; TV moderators.  The developers of those segments claim that when ad agencies start buying spots on TV shows using these metrics instead of the ones that were fabulous until five minutes ago, there’ll actually be a relationship between seeing ads and buying products.

It can’t be any worse than what they’ve been using until now.  If you talk to network executives privately, and to account managers at ad agencies, doubt about the utility of Nielsens is a poorly-kept secret.  I’m not talking about weaknesses like undercounting racial and ethnic groups, and missing out-of-home viewing in airports and bars, and being clueless about online TV viewing, both legal and not.  I mean the conspiracy of silence about the whole premise of demographics.

Why hasn’t anyone blown the whistle?  Because the whole network-advertising-marketing-research village is in on it, and they’ve been afraid to burn the house down without some new roof to put over their heads.  Poltrack’s salvo suggests that CBS and Nielsen are confident enough about what they’re touting now to admit that their old model was built of straw. 

I suspect that this new metric won’t be nearly as useful as the “taste community” analytics still waiting to be born – a transnational audience analysis that mines all the rich new data available about socially-networked online entertainment consumption.  But for that to happen, the Web analytics that currently pass for measuring engagement – hits, clicks, visits, visitors, pageviews, uniques, repeats and the rest—may also have to bite the dust. 

Marty Kaplan holds the ” target=”_hplink”>USC Annenberg School for Communication & Journalism.  Reach him at martyk@jewishjournal.com