Davies Meyer – home
    Data3 min read

    ACR Data

    ACR data is generated through Automatic Content Recognition: automated identification of media being played. On supported smart TVs, video and audio signals can indicate which content ran on the device. This is initially a device-level observation: it establishes neither a particular person’s attention nor automatically anonymous processing.

    ACR Data explained

    A television plays a commercial. recognition can record this where the device, source and recognition method support it. That does not establish whether anyone was in the room, watched attentively or remembered it later. This distinction matters to media planning: recognised playback is a different measure from an attentive exposure or a persuaded person.

    For each dataset, check the devices, countries, sources and periods covered, plus known recognition limits. Not every device provides data and not every playback is recognised. Participating households do not automatically represent the whole population. Population estimates and person-level assignment require additional methods and assumptions. More devices therefore do not invariably replace the information from a carefully recruited panel.

    The data is not inherently anonymous. VIZIO, for example, describes linking its Viewing Data with device identifiers and other data, alongside separate user choices. This is a specific vendor example rather than permission for other providers or markets. Device access, personal-data processing and subsequent advertising contact need appropriate assessment for the intended use. Aggregate reports do not establish that collection itself was anonymous.

    Creative Engineering connects a relevant campaign with an appropriate assessment of delivery. We take responsibility for the concept and quality. AI can summarise patterns; coverage, assignment and uncertainty require qualified review. Website activity observed after a commercial does not establish that the ad caused it. Plan suitable comparisons and assess data licensing, integration and analysis alongside the value for decisions.

    Examples

    Hypothetical application

    A media team uses lawfully available ACR reports to examine recognised playback of a commercial on supported devices. It documents device coverage and adds suitable research on attention and recall. A separate comparison analysis investigates possible additional website activity.

    Key Points

    • Distinguish recognised playback, person-level exposure and attention.
    • Make device coverage and population estimates transparent.
    • Assess personal-data links and purposes of use specifically.

    Practical application

    Ask for an explanation of coverage, recognition, personal-data links and permissible use. Then choose a question the dataset can actually answer and add missing evidence deliberately.

    Useful measures

    Documented coverage

    Share and composition of captured devices and content relative to the particular measurement purpose.

    Recognition quality

    Check correct, false and missing recognition against suitable references.

    Decision value

    Establish the defensible additional insight after licensing, integration and analysis effort.

    Common mistakes

    • Billing recognised ACR playback as proven attention.
    • Describing device-level data as universally anonymous.
    • Treating subsequent website visits as incremental effect without a suitable comparison.

    Sources and context

    Frequently Asked Questions about ACR Data

    No. Recognising content on a device does not automatically establish who was present or paying attention. Additional appropriate measurement is needed.

    Loading related terms…

    All Terms