Covering Media's New Metrics

By David Lieberman
February 11, 2008 08:43 AM

Reporters who cover the media can appreciate the surreal feeling that the comics of the Firesign Theater wanted to convey in 1974 with their album of sketches that warned: Everything You Know Is Wrong.

Just about every yardstick we've used to determine how many people read, watch, listen, and buy different media is fast becoming obsolete as the digital revolution reshapes newspapers, magazines, books, television and radio shows, movies and music.

Newspaper circulation figures miss the millions of people who prefer to read news and features online. And conventional overnight and weekly Nielsen TV ratings don't capture viewers who now watch shows at their convenience online or on DVRs, VOD and portable devices.

The changes have cast a spotlight on a part of the media business many reporters take for granted. We're now being forced to look at the often inexact standards used to count audiences, and the opportunities that companies have to manipulate or misinterpret the numbers.

Several recent news events serve as strong examples.

The public got a look at the shaky underpinnings for seemingly bedrock audience measurement figures in late 2007 when Federal Communications Commission Chairman Kevin Martin wanted to expand his agency's authority over cable. A federal law would give him more power if the FCC can demonstrate that cable lines are available to at least 70 percent of all homes and that at least 70 percent of the homes that offered cable service actually subscribe.

Cable long ago passed the first threshold. The second - penetration among homes passed by cable lines - was in dispute.

Martin put the figure at 71.4 percent, using an estimate from a newsletter company that had sent questionnaires to cable operators.

Other analysts said Martin's figure was too high. Research firm SNL Kagan, for example, put cable penetration at just 58.4 percent of homes passed, based on its own surveys of cable operators blended with U.S. Census and Nielsen data.

The FCC ultimately punted, giving cable operators 60 days to report the key figures for their systems.

Yet we already know their answers: Comcast, the largest cable company, said in its latest 10-K filing to the Securities and Exchange Commission that its 23.4 million basic subscribers at the end of 2006 represented just 51.3 percent of the homes it passed. Time Warner Cable reported that its 13.4 million customers in 2006 accounted for 51.4 percent.

Who's right?

We may never really know. That key measure of potential Comcast customers is simply "an estimate based on the best available information," the company says in its filing.

Even companies that have precise customer counts often find ways to make things murky. Consider the case of TiVo.

News stories often mistakenly report that the digital video recorder pioneer has 4.1 million "subscribers." But TiVo doesn't disclose that statistic. Instead, it offers a figure for "subscriptions," which count how many of its DVRs are in use. A single home with three TiVo DVRs counts as three subscriptions.

The challenge of determining what's meaningful in measurements of audience size will become much more difficult as media companies blend traditional standards with new ones.

For example, in November several newspaper companies and advertisers introduced a metric called Audience-FAX. Unlike the Audit Bureau of Circulations' twice-yearly analysis of newsstand and subscription sales - which have been declining since the mid-1980s - Audience-FAX paints a cheerier picture by attempting to count online readers as well as pass-along readers, the different people who check out a single copy of a newspaper.

That sounds great, until you look at what the new metric measures. The audience figures are at least partly derived from surveys in different markets that ask people about their newspaper reading. They're counted as readers if they say that during the period being examined they read something from the paper in question - at home, online or at the barber shop. It doesn't matter if they simply flashed on one page for just a moment, or even if they mistakenly thought that they did.

In television, technology seems to give Nielsen Media Research a better handle on audience habits. Its People Meter boxes, installed in homes of survey participants, automatically record when a TV set is on, the channel being watched, and - when people take the trouble to push the right buttons - who's in the room.

The biggest change this year is Nielsen's decision to include viewers who don't watch a show live as it is broadcast via the airwaves or cable. Weekly ratings now also include people who watch a program any time on the same day it airs - the cutoff is 3:00 AM. And beginning with the TV season that started in September, networks cut ad deals based on the number of people who watch a program up to three days after the air date.

While this certainly provides for a more complete picture, the new statistics still don't tell us how many people watch a show or network in all of its venues. As Nielsen begins to compile those figures - as well as viewing patterns outside the home, for example at the office, in bars, college dorms and hotels - news organizations will have to decide what figures will be most meaningful to their audiences, and how long they're prepared to wait for them.

Even though these and other metrics are flawed, they can't all be ignored. Beat reporters need to have a collection of statistics that they can be sure readers and sources will accept and understand without a lot of explanation. If you doubt that, imagine what market reporters would say about the New York Stock Exchange if they couldn't mention the Dow Jones Industrial Average. Or how baseball writers would cover players and teams if they didn't have batting averages. Or how political correspondents would discuss presidential campaigns without poll results.

What media reporters can do is keep pointing out weaknesses and blind spots in the new statistics. Just because everything we thought we knew about media seems to be wrong, it doesn't mean that everything the data gatherers want us to report is right.

* David Lieberman covers the media for USA Today

Please enter your friend's e-mail address

Please enter your e-mail address

If you would like to include a message, please add it here:

Copyright © 2008 Donald W. Reynolds National Center for Business Journalism