“On TV and Video” is a column exploring opportunities and challenges in advanced TV and video.
Today’s column is written by Jane Clarke, managing director and CEO at the Coalition for Innovative Media Measurement (CIMM).
This upfront, the focus is on outcomes.
As presentations continue, the role of outcome-based measurement – whether for attribution or campaign optimization – is the key concern of buyers and sellers.
A+E is an early adopter, offering outcome-based guarantees and real-time campaign optimization based upon attribution. But everyone from Turner, which is using set-top box data from AT&T video subscribers, to most recently Hulu, which also announced plans to help sponsors assess outcome impact, is jumping in.
NBCU is also working with advertisers to guarantee that its commercials will drive sales. And in a sign of the growing importance of attribution, Horizon Media recently announced that it is aiming to make more than half of its buys guaranteed against outcomes within the next three to five years.
But while the attribution train is heading swiftly down the track – and much optimism and enthusiasm abounds – there is also much confusion, uncertainty and need for education in the space.
The time is now for buyers and sellers to work with the broad array of attribution vendors to ensure there is understanding not only in exactly what attribution is and how it is conducted, but also to gain insight into what approach is best for the KPI being assessed and for individual businesses.
That need was made abundantly clear in a comprehensive review of the state of TV attribution, which was commissioned by the Coalition for Innovative Media Measurement (CIMM) and 4A’s Media Measurement Task Force.
While the industry expresses excitement about attribution, there is awareness that not every attribution solution solves for every campaign goal.
When evaluating attribution vendors, marketers must ask critical questions to ensure that the attribution measurement matches expectations.
What forms of TV are measured, and how?: Determine if all forms of TV and video will be measured – linear, video on demand, addressable, short-form video on connected TVs, computers or mobile and long-form premium TV on connected TVs, computers or mobile – and whether the cross-platform impact of TV on brands will be taken into account.
Dealing with a knowledge gap: Make sure that the provider you work with has a general knowledge of TV and understands the industry’s unique language and practice. Also, be attuned to the overselling and overpromising of what an attribution study can deliver.
Data questions: Make sure you understand how a provider is disentangling data integration issues and overcoming the challenge of creating representative data sets.
Implementation: Make sure you understand how to get the right answers in a standardized data delivery platform as compared to a custom analysis. How easy will it be to implement a customized solution?
Applications: Clarify with the provider how to know which answer is right when results don’t align with past learnings. Do you understand what factors drive other factors and how to accumulate learning across studies rather than learning new lessons each time a study is conducted?
Since attribution as a practice is still developing, it’s important to dig deeply into those issues that are most relevant to your business and your objectives. Ask questions and request sample input data until you are satisfied.
For example, there are many different television measurement techniques in use and sources of viewing data available. You need to be sure that the data being used for your study includes all of the platforms that are important to you and that the biases in the data don’t yield a materially distorted view of your audience.
You also need to be sure that the approaches the providers use for isolating the impact of television are relevant to your business. Some techniques involve statistical models like regression; others involve test and control studies. Both approaches are valid, depending on your objectives. But there are different factors that may influence results. In the regression models, are all the drivers of marketplace outcomes included in the study?
Are all the marketing tactics, external influences and non-television media included? In the test and control approach, are the control groups set up in a way that all other factors are equally represented? Is the only thing that’s different the presence of television advertising in the test group?
Ask questions. Trust, but verify.
And don’t be shy about sharing your experience. The industry needs to know how the practice is evolving and whether any of these key concerns are rectified. A continuing dialogue between marketers and their providers is critical for everyone to ensure that the attribution train does not come off the tracks.
Follow CIMM (@CIMM_NEWS) and AdExchanger (@adexchanger) on Twitter.
