• by David Goetzl  @dgoetzl, April 7, 2011

Nielsen has completed testing on using set-top-box data to upgrade its existing measurement systems in local markets. The process could ultimately lead to the elimination of the much-criticized diaries used for measurement in small DMAs.

"That's the dream," said Pat Dineen, a Nielsen senior vice president. "And this work enables us to have that kind of conversation with our clients."

With the test involving three markets completed, Dineen said Nielsen is confident it can implement a system that addresses the major issues raised about the reliability of set-top-box (STB) data, including how to measure all homes in a particular market and account for when a box remains on, but a person is no longer watching.

Dineen also said that by evaluating the STB data, Nielsen has emerged confident that its current panel-based measurement is reliable, as the two methods did not produce results that were wildly different. "When we compared the two side by side they looked very similar," he said.

The lengthy test took place in three markets, with each representing one of the three kinds of measurement methods Nielsen uses. St. Louis has local people meters; Greenville, S.C. has set meters; and Reno, Nev. has diaries.

The STB data was culled from homes with Charter cable service in each market -- from 46,000 homes in St. Louis, 12,000 in Greenville and 6,000 in Reno.

Nielsen's aim is to offer a local-market ratings product, derived from a melding of STB data and its traditional methods -- what it is calling a "hybrid" approach.

Nielsen will begin sharing the detailed test results with clients over the next few months, while gauging their interest in moving ahead with further testing and deployment. On a certain level, it ultimately comes down to money: how much clients are willing to invest to fund Nielsen's efforts, including potentially acquiring more data from operators beyond Charter.

"We're going to be guided by what they tell us," Dineen said.

On the vexing matters of how to ensure all homes in a market -- notably those without an STB -- are accounted for and how to ensure that viewing is not counted when an STB is unintentionally left on, Dineen said Nielsen can extrapolate results from its national ratings panel to determine what is going on in a particular market. That type of modeling can "correct for, and improve on, set-top-box data," Dineen said.

With its ability to deal with those issues effectively and attach demographic data to the STB-based results, Nielsen believes it will have a leg up versus competitors marketing STB-based measurement systems. Dineen said its test indicated that its "hybrid" product would produce ratings with greater "fidelity" and sturdiness than Nielsen's current model.

Of course, some would say that is a low bar, especially when it comes to the diary markets that still rely on people filling out forms four times a year. Some station managers and advertisers would argue that any STB data at all in the Lubbocks, the Eurekas, and the Juneaus would be a major improvement.

"This is a step toward working with our clients to replace diaries," Dineen said of the completed test.

Currently, Nielsen panels may include 500 to 650 homes in a particular market. How many homes would be included in one of the panels that incorporates STB data is up for analysis, Dineen said. But the figure could be in the tens of thousands -- 46,000 were used in St. Louis in the test -- and not in the millions.

Nielsen has no plans to deploy STB data in its process for gathering national TV ratings.