Audience estimates, commonly known as ratings after the main type of measure, permit the valuation of advertising inventory and program and channel carriage rights. Ratings indicate the popularity of a program or a channel. Additional audience measurement data may provide information about the households or individual viewers, their other TV watching, and demographic, geographic and socioeconomic attributes. Ratings help advertisers and media buyers to select the right media at the right time.

Classical TAM (television audience measurement systems) use special equipment to track viewing, including a set meter (which records the tuning and power state of the TV set) or a people meter (a set meter that permits viewers to sign in and out of viewing). The viewing is reported, typically through dedicated phone lines, is combined with channel and program information and aggregated for publication (some companies also release set-level or viewer-level data). Such systems are characterized by small, although not necessarily balanced, samples and high costs. Such systems are typically unable to measure any but the most viewed pay-TV channels reliably. In the smallest markets, less reliable but less costly data collection means, such as paper diaries and phone interviews, may be used.

Aside from high costs, classical TAM systems are becoming less useful as television viewing changes. Their small samples render them unable to measure viewing of all but the most popular pay-TV channels. They are generally able to cope with DVR playback but often not with VOD (video on demand), especially programs never broadcast in realtime. They usually have no means of measuring OTT (over-the-top) viewing. While still important as capturing most viewing in most countries where they exist, and providing the currency on which valuations of advertising space and programming are based, they are distinctly headed in the same direction as realtime viewing—towards near-extinction. A different approach is needed.