TiVo Report Sees New TV Data Shaking Up Measurement

TV measurement is going through big changes, according to a new whitepaper from TiVo.

After years of audience estimates being based on small samples or panels of viewer, big data has come to TV in the form of set-top-box data and automated content recognition.

Those new forms of data “are set to radically change the way marketers, networks and broadcasters measure TV viewership, plan media buys and execute ad campaigns,” TiVo says.

While not mentioning Nielsen by name, the report says that using old-school TV measurement—in a digital age is creating skepticism about measurement from networks and marketers.

“The root of this growing skepticism is the continued reliance on small panels – numbering in the tens of thousands of households – to represent the TV viewing behavior of 120+ million U.S. households. As one interview subject with a background in statistics put it, ‘modeling the entire population from such a small sample necessitates lots of assumptions and risks many inaccuracies,’” the report says.

Nielsen stands by its approach.

“To accurately measure anything, you need to be able to directly observe it and sample it in a representative way. That is why our industry standard panel is and will continue to be the foundation of television measurement,” said Kelly Abcarian, senior vice president, product leadership at Nielsen.

“Panels are purposely built for persons level media measurement, while big data is a byproduct of a machine. So not only is this device level data instead of persons level data, but it is also merely an unrepresentative bigger sample. A sample that is incomplete, biased and discriminates against those excluded from its measurement is a problem regardless of its size. If big data comes more from high rises, city dwellers or apartments instead of private residences, meaning you don’t get data from all parts of the market and you don’t have a truth set that can be used for adjustments, you can’t pretend it doesn't matter,” Abcarian said. “As we have seen in digital, when data is not accurate or transparent, this results in under delivery of the ROI goals and advertisers will either lose confidence in television as a medium, or the agency, or both.”

Measurement is a more complicated proposition now than when there were three networks and families watched on TV set in the living room. Now there are new platforms and devices and an explosion of content. Marketers also have experience with digital media, which promises precise measurement, targeted audiences and feedback on how users responded to content and advertising.

TiVo says the new forms of viewing data are being collected at scale, are highly granular, are being built for new viewing methods and can be integrated into marketing models.

Set-top-box data is the largest source of TV viewing data at this point. One downside, however is that because of the fragmented nature of the cable industry combining data from different operators can be tricky.

“The biggest problem – and one that limits the number of companies willing to even tackle STB data aggregation – is dealing with the variety of file formats, data structures and levels of granularity,” the TiVo report said. “Concerns about how well STB data reflects the overall population persist, since 17%of TV households rely purely on over-the-air signals and an increasing number of younger households are cutting the cord.”

Data based on automated content recognition is available from fewer viewers at this point.

“Manufacturers are only beginning to share ACR data with other companies or withholding it completely, limiting the reach of ACR data available today,” the report says. “Vizio is a notable exception, but census-level ACR data availability would require active participation from major manufacturers – and none appear willing to cooperate with others.”

On the plus side, the TiVo report says ACR data collection benefits from its flexibility, allowing the capture of viewing data across linear, DVR and smart TV app viewing, although there may be data gaps in VOD and some long-tail content due to watermarking requirements.

It is tough to combine set-top-box viewing data with commercial logs, the report says.

“ACR technology enables the tracking of ad units at the individual device level without having to rely upon commercial as-run logs. This allows for faster reporting on ad delivery and more accurate data for addressable TV advertising. Nonetheless, taking advantage of this technology requires advertisers to incorporate ACR-ready watermarks during the production and distribution process,” the report said.

The presence of clear, trackable consumer conversion activity is essential, as with all attribution, the report added.

While the new methods of audience data collection have advantages, adopting them faces hurdles, the report says. Those included:

  • Industry inertia and the currency conundrum: “Despite the limitations of traditional, panel-based TV measurement, it is the currency by which billions of dollars of TV advertising is bought and sold. And, it is no small task to shift from today’s gross rating point (GRP)-based TV buying model to the impression-based buying model required to maximize the benefits of next-gen TV data,” the report says.
  • Data quality transparency:For next-gen TV data to become the new standard, data vendors will need to offer greater transparency into their collection approach, modeling methods and data sources.
  • Consumer privacy and regulation: Digital data collection and targeting has long been subject to a set of self-regulatory guidelines, but the nascent nature of TV data collection means these standards are still evolving.

Looking ahead, greater data scale, quality standards, privacy protections and ease of access through aggregation will enable marketers and networks to more seamlessly leverage the strengths of both set-top-box and ACR data,” the report said.

“People-based marketing will become a reality as marketers and networks stop evaluating TV ad exposure in aggregate demographic groups and instead map individual TV viewing behavior to cross-device marketing exposure and attributed outcomes. Now that marketers have these capabilities across other screens, they will also demand them from TV,” the report said.