Mastering the Universe

Cable ratings are a double-edged sword for local-cable advertising sales. But new developments on the ratings front are starting to dull the pesky edge of that blade — at least in some markets.

The crux of the problem is that Nielsen Media Research hasn't provided a measurement of cable's universe in individual markets. And because the size of cable's local viewership is open to interpretation, that's led local broadcast to charge cable operators with jury-rigging cable ratings.

The so-called “fusion” of cable meter numbers with cable diary data gives local broadcast a tempting club to regularly swipe at local and regional cable ratings data. But broadcasters regularly overlook the fact that cable has had to self-certify its numbers because Nielsen simply doesn't do it.

Nielsen vice president of communications Anne Elliott says there's a good reason why the company hasn't supplied those numbers. “Our sample size is not sufficient for all 11,000 plus cable head-ends.”

But some relief is in sight. With the expansion of Local People Meters (LPMs) into more of the top 10 markets — Boston LPMs debuted in 2002; New York was added last week, Los Angeles in July, Chicago in August — Nielsen will be able to supply the cable and broadcast companies in those markets with one set of ratings and numbers pertaining to one universe.

With the LPMs, the practice of ascribing missing local diary data by using national people meter numbers, known as the “zero cells issue,” will be gone. The hated, hard-to-remember written diary also will disappear. Objective samples will exist. Yet, lingering objections remain, and the cable research industry stands ready to answer broadcast's complaints.

Asked to take their best shot at what is wrong with cable's ratings, Susan Cuccinello, senior vice president of research at broadcast trade organization the Television Bureau of Advertising, starts right off with the fusion factor.

“We understand fusion helps address the zero-cell problem for cable. And we acknowledge that the fusion formula is based on widely-accepted industry guidelines for estimating ratings in those circumstance,” she says. “But, our biggest concern is that in developing the fusion software, [National Cable Communications] made it possible to easily alter the data imputs.”

Cuccinello suggests that when Nielsen household ratings numbers are fused with the household universe numbers that systems determine themselves, the resulting ratings data could be overstated. (Think Adelphia Communications Corp.) While not charging that cable overstates its audience as a general rule, Cuccinello suggests that the possibility exists, and, in Adelphia's case, did occur.

With cable's local CPMs growing once again, Cuccinello adds the proposition: “If audience estimates are being overstated or inflated, advertisers are paying too much for too little audience.”

NCC senior vice president of research Jeff Boehme is happy to respond to that idea. “We have a decades-long experience of building a mutual methodology that estimates local cable audiences where Nielsen can't or won't,” Boehme says. “Nielsen to this day doesn't provide any direct reportability for any specific cable system or interconnect. However, Nielsen does report a lot more cable viewing than ever before, especially in markets using Local People Meter data.”

He adds, “TVB is losing money because broadcast viewership is declining, so they've concentrated their focus on negative campaigns against cable.” (See story, page 36.)

Yet the amount of fuzzy cable data is real. “You have markets where there is national Nielsen meter viewing for cable that does not show up on local diaries,” notes Jon Sims, vice president of research at Comcast Corp.'s advertising sales division, Comcast Spotlight. “That's what creates a zero cell. But those zero cells can represent 40% of all quarter-hour viewing.”

That's a huge amount, and in an attempt to correct it, Nielsen developed a zero-cell ascription method, according to Sims, using “adjacent viewing” i.e. “ascribing” viewing even though it wasn't really reported. That upset the competition.

But the LPM, which also has some broadcast stations and Fox Network crying, corrects that problem. “Local People Meters get rid of the diaries, get rid of hypo- [broadcast] sweeps periods and gets rid of zero cells,” Sims notes. “Local People Meters will be light-years ahead of what we have now.”

Lyle Schwartz, senior vice president and director of media research at the media-buying firm The Mediaedge, agrees. “The diary is very inefficient and a very ineffective way to measure media,” Schwartz says. “Currently, we get to read the market four times a year, but sweeps programming artificially inflate that view. People Meters allow us to understand how the marketplace is performing with greater accuracy and greater accountability.”

None of the researchers saw any racial or ethnic bias in current Nielsen methodology, despite the recent protests on that front.

“I haven't seen anything in the sample that leads me to believe there's any bias,” Schwartz adds. “Trust me. We'd be screaming if that were the case. But we know from the Boston LPM test that broadcast numbers are going to go down. I hope Nielsen goes ahead with its New York LPM test in June.”

There are things that cable and Nielsen can do to make the numbers more solid.

“NCC will continue to make full disclosure of how its ratings are obtained,” Boehme says. “But Nielsen needs to improve and increase its sampling. It needs to increase the number of LPMs in as many markets as it can. The National People Meters were introduced in 1987. It's 14 years later, and we've only had the Boston Local People Meter test.”

“It's actually been 17 years,” Elliot acknowledges, “because it's always been difficult. Major funding comes from the media companies [broadcast stations and, now, cable systems], and they've been slow” to offer financial help.

Boehme also dismisses the charges that Nielsen is biased against Black or Hispanic households, as various protest groups have insisted of late. “Overall, does Nielsen do a good job measuring TV? No. But do they purposely under-represent minorities? No way. If anything, they oversample the ethnic population.”

“What cable needs to do is support the LPM rollout, as Comcast is doing financially in all 10 markets,” says Sims. “We've borne the lion's share [of the cost], involving millions of dollars. But there is an absolute necessity for having universal estimates for our systems and interconnects replace subscriber counts based on an internally-created method.”