Let’s Replace TV Panels with Real Data About People to Reflect Changing Demographics
With more viewers cutting the cord and streaming, Nielsen’s traditional panel methodology is having trouble keeping up with fragmented consumer behavior.
The prevailing thought in the industry is that panels are needed for calibration, due to biases in some big data sets. But much like the third-party cookie, panels are proving themselves to be a legacy approach in a new digital era. That’s because they largely rely on tracking demographic information like gender and age, but not the nuances of both a consumer’s background and behavior.
As a result, Nielsen panel measurement has recently faced a ton of backlash in the industry: losing MRC accreditation, public statements to cease use from the Video Advertising Bureau, and suspension of use from major media companies due to undercounting viewers, including Black and Hispanic individuals, during COVID-19.
Because the industry has yet to create a new replacement standard, media companies are continuing to use Nielsen panel data. But the solution might be simpler than we think: more accurate, ethically-sourced data that reflects changing demographics.
The big data vs. panels debate
Nielsen has argued that big data “has limitations” and can “increase the probability of waste and fraud.” I respectfully disagree.
Data can be adjusted for its biases. Panels, on the other hand, can create biases, as they don’t fully consider the nuances of consumer behavior, especially as it relates to audiences that are underrepresented. One example of this issue arises in political polling. The wide discrepancies happen due to not having accurate representation from people who work multiple jobs.
Analytics need to reflect changing demographics, and panels have never been effective at understanding growing minority audiences. Consider this fact: When Nielsen began using panels some 50 years ago, the U.S. was 88.6 percent white. Today, the U.S. population is much more multiracial and more diverse. As of 2020 Census data, the U.S. was 61.6%, 18.7% Hispanic or Latino, 12.4% Black or African American, and 6% Asian. What’s more, the percentage of people who reported multiple races changed more than all of the race-alone groups, from 2.9% in 2010 to 10.2% in 2020. Overall, more than 50% of all TV viewers ages 18-24 are diverse.
It’s impossible for such an outdated approach to keep pace with the rapid growth of both multicultural and diverse U.S. audiences and their consumption of CTV.
The untapped power of CTV and mobile data
With third-party cookies going away, there’s plenty of opportunity for brands to use ethically sourced first-party data to build custom audiences that improve their targeting accuracy. But there’s an important caveat to the best first-party data to leverage: It should be based on mobile and CTV devices.
When you inform CTV targeting with app behavior data from mobile devices at scale and then match this data to millions of validated CTV households, you create more accurate, relevant, and unique audience segments and insights. Mobile-based signals let you determine viewer action from your ads, increasing the ability to scale multicultural audiences across CTV. When you layer this data with other customer information such as interests, purchase signals, POI, program type, and, yes, even Census data, you get a more robust—and reliable—profile of your target viewer.
A path forward for more precise CTV targeting
As the world continues to change, so must the mechanisms we use to understand TV viewers. Using a mobile data-based approach in CTV helps unearth more comprehensive audience insights than the one-dimensional panel methodology that has been used since the 70s. This data can help reach viewers not represented effectively on panels, increasing message effectiveness in the way media is consumed today: simultaneously on mobile and TV. The result is a more inclusive approach in which every viewer counts.
Aziz Rahimtoola is the CEO and co-founder of Sabio.