-
Converged TV Measurement: Currency is the tip of the iceberg
The TV advertising industry is in deep discussion about how ads should be bought and sold as linear and digital converge. While the debate and jockeying makes for “great television” the more interesting story is in the details of how TV’s historically analog marketplace is being transformed by digital measurement and analytics.
The tip of the iceberg
Currency
Below the surface
Measurement
While the traditional TV market has been settled on a demographic audience reach-based approach for decades, digital advertising manages the delivery of the ad all the way to living room couch and handheld device—as such, it can provide a more granular understanding of viewership.
Measurement enablers
- Panels: Observe a subset of a group to be measured (e.g. the whole TV viewing audience) and use math to extrapolate.
- STB data: Generated from households that use set top boxes, from certain TV providers including cable and satellite.
- ACR data: Generated from automatic content recognition software in smart TVs that recognizes anything on screen.
- Impressions: Directly counted video views, as measured by one or more digital ad servers.
Identity and audience
In addition to view counts, marketers need to know who they reach and how often.
- First-party: Advertisers and agencies maintain their own audience/customer data that they can associate with individual views.
- Third-party (Vendor): Audience and data specialists aggregate data on viewers and connect it to views.
- Third-party (Publisher): Content owners with major reach collect their own info via subscriptions and other means to sell more valuable audience-specific views.
Identity and audience enablers
Various mechanisms exist to identify viewers and persistently associate them with audience attributes.
- Household data: Consumer information can be associated with households using purchase, subscription, public, and other data.
- IP address: Households can be uniquely identified by numeric internet protocol addresses used to connect homes to the internet.
- Login data: Consumers can be required to log in and provide unique identifiers such as emails/phone numbers, as well as agree to viewer data sharing policies.
- Device ID: Devices can be uniquely identified by temporary IDs such as cookies, or more persistent hardware or platform IDs.
Attention and outcomes
Marketers are working to get information on the quality of views and downstream business results.
- Attention metrics: Not all views are created equal. Co-viewing (multiple viewers) and viewer focus are factors in audience attention.
- Online conversions: Digital traffic reporting allows marketers to understand if online purchases or site views happen after ad views.
- Offline conversions: For non-digital purchases and real-world activities, various data points support correlation with ad viewership.
Attention and outcomes enablers
These datasets can directly indicate or be used to extrapolate viewer behavior.
- Attention data: Some solutions incorporate cameras (typically via opt-in panels) to understand how intently viewers are watching.
- Location data: Information about the viewer’s location can be translated into specific (e.g. movie-watching) or general (e.g. vacationing) behavior.
- Purchase data: Credit cards, loyalty cards, and other transaction-related components provide data to confirm viewer purchases.
- POP data: Data collected at points-of-purchase, such as emails or phone numbers, can be used to correlate views with conversions.
Fraud and other measurement disruptors
More exact digital measurement comes with intentional and accidental data corruption issues.
- Fraud: Criminal enterprises reap rewards with relative anonymity and low risk by setting up schemes to generate and get paid for non-existing views.
- Misrepresentation: The systems involved may accidentally or neglectfully report details of viewing data, including audience attributes, location, viewing device, inventory source, and more.
- Screen off: Both CTVs and set top boxes have the potential to report viewership based on data that reaches a device connected to a screen that is turned off.
