• by Joe Mandese , October 27, 2008

On the eve of meetings this week with its national clients in New York and Los Angeles, Nielsen made two important announcements, including a decision to delay a controversial plan to begin measuring online media usage in some households that are part of its television "currency" panel. The second announcement was a disclosure that Nielsen has been grossly misestimating the household coverage for many smaller cable networks for years because it has not been able to identify when new, "digital tier" channels are added, or removed by satellite TV households.

The Internet measurement plan, which Nielsen originally planned to implement this quarter, has been delayed until April 2009 due to feedback from clients. Nielsen did not disclose the tenor or those conversations, but some clients were concerned that the plan, originally announced at a client meeting in March, was tantamount to a "live test" on Nielsen's television audience measurement panel, which is the basis for some $80 billion in TV advertising spending.

Despite those concerns, Nielsen said it was actually expanding the scope of the test when it resumes in April to include measurement of both Internet video usage, as well as general Web site navigation. Among the concerns some clients have had, is that the type of people who are willing to comply with measurement of their TV viewing behavior, might behave differently if their households Internet usage is also being measured.

The test is separate from the so-called "convergence panel" that Nielsen recently deployed. The convergence panel is a panel comprised of "retired" TV measurement households, who have agreed to have both their TV and Internet usage measured over time, and is not considered a currency panel, or the basis of TV advertising deals.

The Internet measurement test on the live TV sample will initially begin in 375 households in April, but Nielsen plans to ultimately expand the number of households in the fall of 2009.

"Over the past two years, we have conducted research and have worked directly with clients and the Media Rating Council to refine our test methodology," Nielsen said in Friday's statement. "The key elements of this test are: For new People Meter homes, Nielsen Field representatives will recruit simultaneously for television measurement and Internet measurement."

Neither the Internet measurement plan, nor the digital cable universe estimate issues are on Nielsen's agenda for this week's client meetings, but some clients will no doubt be buzzing about, if not discussing the issues publicly with Nielsen executives.

On Friday, Nielsen outlined steps it was taking to improve the accuracy of its digital tier cable network universe estimates. The announcement follows months of disclosures and discussions with Nielsen clients revealing that Nielsen has been grossly underestimating the number of households for some digital tier cable networks distributed over satellite TV systems for several years, and that more recently, its estimates have begun to decline as well. While the universe estimates don't affect the ratings cable networks sell and guarantee to advertisers, they do impact the relative CPMs (cost per thousands) that are the basis of many of those deals.