>> Article Archives >> Advertising >> Grabbing Media By The Tail

Grabbing Media By The Tail
By Erwin Ephron

Forget the celebrity smile. This year's iconic body part is the long tail.

Yesterday's Research is Failing Today's Advertising

70 years ago Art Nielsen, Sr. read the future when he had his reports inscribed, ". . . If you cannot measure a thing, then your knowledge is meager indeed." Art Nielsen was predicting media's long tail.

Body Part of the Year

Forget the celebrity smile. This year's iconic body part is the long tail. It describes the scattered distribution of consumers as more options become available to a relatively constant population.

In commerce that means a growing number of small brands and in media many more tiny audiences. Both the Internet and Television are good examples of media's long tail.

Limbs and a Torso

At a recent ARF conference, Scott McDonald of Conde Nast wowed the tail-struck audience by attaching limbs and a torso. A few of Scott's examples:

 From 1,000 hit albums in 2000 to only 400 in 2007 because there are many more internet distributed albums for people to choose from.  From the 100,000 books available at a conventional mega-bookstore, to the 3.7 million titles on the virtual shelves of Amazon. From a top TV show audience share of a 68 in 1953 to a 13 share in 2005, to an average cable rating of less than a 1.0 in 2008 as we top 110+ TV channels. And many, many more brands in every major product and service category.

Boundless and Countless

With all this choice an advertiser's media options seem boundless, but countless might be the better word. Media's long tail is very hard to measure.  The predictable effect on media research, the TV ratings for example, is Apocalyptic. To quote another Media Prophet:
"When the number of channels climbs to Heaven, the average rating falls to Earth and sampling error will blot out the Sun."
The samples tiny audiences require for stable data are beyond even the big purse of Television. It is common knowledge that the ratings of many commercial minutes are smaller than their relative error, which means we can be certain their true value lies somewhere between zero and a real number.

The quick fix approach is averaging the small audiences, but that carries a different price. It ignores the greater targeting value of the many individual pieces of television's growing long tail.

The Survey Is No Longer Enough

Media are learning that sample-based research is no longer adequate to report the size and personality of their many audiences. Yet media research continues with business as usual. Most of the focus is on adding new measurements like engagement or product purchase, not on repairing the basic counts of audience.  That would require a complete remake of how we do the measuring.

The first medium to confront the long tail problem is Out of Home. Outdoor had no choice but to abandon conventional survey measurement. It needed board-by-board ratings but couldn't afford the mega samples (25,000 respondents as a starter) required in each major market to produce stable, unit-by-unit data.

Measuring Out of Home's long tail required outdoor's Traffic Audit Bureau, led by Joe Philport, to build an integrated, multi-layered audience measurement system. For Out of Home the layers are:
  • Near census-level traffic data to provide the robust base count of persons passing by billboards. This measures "opportunity to see."
  • Eye-tracking data to measure the noticeability of different size outdoor units in different roadside situations. This reduces "Opportunity to see" to people who have "Eyes OnTM" the unit.
  • Survey data to gather the demographics and reach/frequency patterns of the Eyes OnTM audiences.  
  • Modeling to integrate the three data streams into a complete unit-by-unit Out of Home database.
In this schema the conventional survey is but one of several data contributors to a total measurement system.

Counting Sand with Your Fingers

Counting grains of sand with your fingers. That's what the long tail asks of survey research today.  There is a better research model and Out of Home has already made the move. Television is next. Measuring its grains of sand is also beyond sample-based surveys.

For TV we can use near-census cable set-top box data as the base count of opportunity to see, eye-tracking to finally measure people seeing commercials and add sample-based research to model viewer demos and reach/frequency.

This would provide precise Eyes OnTM ratings for all TV inventories and allow us to target, package and buy better. It would be a much improved ratings system.

The technique is called data integration. And it grabs media by the tail.

Erwin Ephron is an authority on advertising and the father of "recency planning." His fresh ideas about how ads work today have changed the way campaigns are planned throughout the world. Erwin

More articles by Erwin Ephron
More articles on Advertising