• by David Goetzl  @dgoetzl, April 29, 2010

Nielsen plans a test to determine whether set-top-boxes (STBs) can help upgrade ratings in local markets. No timetable is set, but the trial will take place in three undetermined markets.

Those three will be of varying sizes, and will represent a cross-section of the methods Nielsen uses to generate its local numbers. One will be a top-25 market served by local people meters. A second will be a top-60 market with set meters, while the third will be served by diaries.

Nielsen is looking to deploy a "hybrid" model, where data gleaned from the STBs will be combined with information from its traditional panels. That amalgam could provide ratings at the second-by-second level for particular demographic segments.

STB data tracks DVR-enabled viewing and holds the promise of allowing Nielsen to offer commercial ratings for local markets, something media buyers hunger for.

Word of the coming test was revealed by Dave Thomas, Nielsen's president of global media client services, at MediaPost's Outfront event on Tuesday.

Thomas said melding STB data with information from Nielsen's standard panels can "take into consideration some of the vexing things that have caused people to pause when assessing (STB) data as a panacea."

Among them is whether a set-top-box remains on, but a person is no longer watching. "Are people actually viewing as opposed to tuning?" Thomas said.

Also, panels can help make projections for viewing across a full market, not just what takes place in homes that have STBs.

Thomas said Nielsen is "very bullish" on using the boxes to provide "better fidelity around the estimates for the benefit of both buyers and sellers."

Both Rentrak and TiVo have STB-based services in local markets, but only provide ratings at the household level, with no demographic details. Rentrak's fledgling service is being used by multiple local stations, including four in the Columbus, Ohio area.

TiVo launched its service last summer in the San Francisco, Orlando and Tucson markets, although it is unclear whether any stations (or others) have purchased it. The data is gathered anonymously from just TiVo subscribers, using samples of 75,000 in San Francisco, 10,000 in Orlando and 5,000 in Tucson.

Nielsen has been facing criticism for dropping a "live-only" stream for program ratings in local markets. The remaining data takes into account time-shifted ratings and has some buyers complaining they have to pay for skipped commercials.