Abstract:

The migration from analogue to digital television is compromising traditional, sample centric audience measurement techniques;- diaries and Peoplemeters. To address this, TNS has invested in a comprehensive programme encompassing research, new metering technology and systems development. A 'census-based' approach, where large volumes of tuning information are extracted from digital STBs using the 'return path' from home, offers new measurement opportunities. This information, referred to as 'clickstream data', presents three challenges. First, the tuning information relates to the STB (not the TV) and after watching, viewers will often turn the TV off but leave the STB on. Second, clickstream data, even when converted to TV tuning or HUT, tells us nothing about the audience. Thirdly, in principle data from hundreds of thousands or millions of homes can be accessed. Analytical tools are needed to handle these data volumes. This paper addresses how these challenges are resolved and describes work being carried out in both areas to validate results. It also describes a new metering approach which provides the necessary inputs to transform and calibrate the raw clickstream date into accurate TV tuning and person's viewing data.