Does it matter if you measure performance with one Smart TV brand, or many Smart TV brands? The answer is yes.
Jeffrey Silverman, PhD
Director, Data Science & Analytics
Mar 10, 2020
Today’s media landscape is more fragmented than ever. Consumers spend hours each week navigating an ever increasing array of content channels, creating new challenges for both marketers and publishers seeking to understand true audience reach and engagement.
As more and more providers begin to enter the television data and audience measurement marketplace to help answer these questions, Samba TV partnered with a major fitness company, using the brand’s national TV campaign to assess the efficacy of audience measurement from the individual television manufacturer’s point of view.
Twenty leading global TV manufacturers have Samba TV’s ACR technology integrated at the chipset level, enabling our researchers to compare results for viewership of the ad campaign across nearly every major TV brand. We reviewed the results for six of the largest Smart TV brands individually, as well as the total viewership for all Smart TV brands with Samba’s ACR technology, to arrive at an aggregate score with the viewership for each individual Smart TV brand projected to the entire US population.
The Samba study found massive disparities in the individually reported data. Because different manufacturers cater to different demographics (discount televisions versus high-end 8K, for example), it is inherently difficult for any single manufacturer to accurately generate a statistically relevant audience forecast of the US.
The different consumer segments create the same challenge for the TV manufacturers that legacy measurement panel providers face today. Both are trying to extrapolate the viewing behaviors of 209 million adults across the most fragmented media environment in history without a complete picture of the audience.
In the chart below, you can see Smart TV Brand #5 projects the test campaign reached 38M households in the US. This is twice the reach compared to Smart TV Brand #1, which projects the campaign reached 18M households.
These projections were done after adjusting for demographic bias. Therefore, the differences seen here can only be explained by Smart TV Brand loyalty bias. To arrive at the statistically most relevant assessment, Samba TV combines data from 20 Smart TV brands to smooth out inherent biases (ex: brands with only low-cost TVs) to deliver a more accurate projection of actual campaign reach. The aggregated data shows that the campaign’s projected reach was actually 34M households in the US.
The key takeaway for those navigating today’s data and measurement landscape is clear: If you are using a data set with only one Smart TV brand, you simply are not getting the entire picture.