Set-Top Box
Set-Top Box Overview
The introduction of Live + Same Day program ratings, in the second quarter of 2010, is a significant and important change, and one that is welcomed by most local broadcasters. The public and private dialogue preceding this shift in measurement raised questions about what future changes may be in store for local measurement: namely, the growing interest in utilizing Set-Top Box (STB) data for audience measurement and viewing insights.
The promise of STB data is significant, but so are the challenges. Most entities agree that much work still needs to be done before STB data could be considered a viable alternative to the current TV ratings currency. Several companies are either investigating STB data products, or are already providing STB data to clients; these include Rentrak, Kantar (formerly TNS), TRA, TiVo and Nielsen.
Industry Groups Evaluating STB Data
Concurrently, several industry groups are actively engaged in the evaluation of STB data offerings, and are looking to establish uniform standards, terminology and guidelines while STB data is still in a nascent stage:
- The Coalition for Innovative Media Measurement (CIMM)
- The Council for Research Excellence (CRE) through their Set-Top Box Sub-Committee
- MPG’s Collaborative Alliance Set-Top Box Think Tank
To ensure local market measurement needs are considered along with national measurement, local broadcasters should be an active part of the discussions and decisions about STB measurement. Several O&Os, broadcast groups, and rep firms already have seats at the table and we expect more to step forward in the days ahead.
The Promise of STB Data
Proponents of STB data are drawn, in part, by the sample sizes which are significantly larger than current national and local samples. The hope is that larger samples could improve data stability and reduce sampling error; the truth, though, is that larger samples are not necessarily reliable (more on that in a minute).
In theory, large sample sizes could enable more reliable audience estimates for niche channels and networks, enable sub-DMA geographies to be examined, and make it easier to track viewing behavior as video options continue to expand and further fragment viewing.
STB data can be matched with other databases - such as purchase information, consumer profile information and other segmentation databases. Such metrics can be specific to one advertiser and one campaign, or can provide more general planning guidelines. These measures go beyond simply identifying who is watching and provide insights that can be critical to campaign planning and analysis that go far beyond what is available today. One example: Rentrak, which positions itself as a database-integration solution, can link television viewing in AT&T U-Verse and DISH homes to an auto dealer’s proprietary sales database to deliver a more targeted television viewing analysis.
The Challenges
While the promise of STB data is of interest to both buyers and sellers, the challenges to be overcome are not insignificant. While many of the challenges are not insurmountable and can be addressed in a manner that should prove satisfactory to all interested parties, the work is just beginning and will take time. Some of the key challenges are noted below.
Technical Challenges.
Some of the technical challenges are well known: the fact that 10% of STBs are never turned off, or that 30% of STBs remain “on” around the clock on any given day. Other technological challenges stem from the fact that STBs were not initially created with return-path data in mind, and only about half of STBs currently in circulation are capable of providing return-path data.
There are many different generations of set-top boxes deployed today, even within one MSO or cable system, and each process data differently. While second-by-second data is expected from all STB sources, each box can vary in how often data is processed, even within the same cable system.
Matching second-by-second tuning data with commercial occurrence information will still be a challenge at the local level, where no complete source for all commercials (broadcast and cable) yet exists.
Methodological Challenges.
Large samples are not necessarily representative or projectable. In truth while STB data is often referred to as “census-level” data, it isn’t. Currently, 10% of homes remain exclusively over-the-air (OTA) and are not included in most STB reports. Not only are OTA homes excluded; an additional 16% of households still have analog cable and are excluded from STB measurement. And a significant number of STBs currently in use do not have the capability to provide return path data. It is estimated that about 25% of all television viewing takes place in homes without STBs. Much of that viewing is to broadcast stations, and none should be ignored.
STB data providers and other researcher companies are working on potential solutions to these challenges. Modeling the missing data, or fusing STB data to panel-based data are two of the primary methods being proposed; industry input will be critical in guiding this work.
Business Challenges.
STB data relies on gaining access to data owned by cable operators, satellite, and telco companies. It takes control of the sample out of the hands of the research company and places it in the hands of some of the end-users of the data. Consider as well that the footprint of any MSO, cable system, satellite provider, or telco will not match the traditional definition of a DMA.
Moving Forward
With so many decisions to be made, and with so many companies testing the waters, it is an exciting time in the field of television measurement. Moving ahead, TVB will be providing regular updates on initiatives in the STB field, and will track projects in several local markets in which STB data tests are occurring. We welcome your feedback. Please contact info@tvb.org with comments or questions.