TV viewing behavior varies greatly by individual. Without person-level insights, household data fails to represent who is actually watching TV. TVision’s calibration data helps measurement platforms understand what actually happens with households tune in.
Ensure your measurement reflects what individuals, not households watch.
Understand assumptions and account for bias in combined models.
Drive innovation when your teams have the flexibility to lean into your own UVP.
“TVision’s panel enables VideoAmp to enhance our dataset with information about the actual humans in front of the TV."
EVP of TV Strategy & Currency
Census data, by nature of its scale, has the potential to be a more accurate measure of TV viewing behaviors than solutions based on panel data alone. But likewise, census data on its own, without person-level insight, falls short of the detail that media buyers and sellers require of any future currency. The future of TV measurement will be based on a combination of census data calibrated for accuracy with panel data.
Here’s what makes TVision the calibration panel of choice for our industry:
Individuals in the same household often don’t watch the same content. Future currencies must reflect who is watching, and what ads they see.
With the increasing popularity of shorter ad lengths, second-by-second data is critical for reporting on ad delivery and engagement.
Audiences transition fluidly between linear and CTV. Measurement providers need a comprehensive view of how people consume media.
As people move in and out of the room, passive measurement ensures an accurate understanding of who’s watching without the risk of human error.
If adjustments to panel data are made for statistical significance, they need to be clearly identified so they can be addressed as the data scales up.
We don’t believe in a one size fits all approach. Innovation will be more likely if companies can manipulate our data on their own.
This example shows what household-level data looks like for ESPN, compared to person-level viewing data. Household-level data reflects an outsized share of female viewers, and younger age groups. Calibrating with TVision’s data lets measurement providers adjust for differences in individual viewing preferences.
HOUSEHOLD-BASED VS. PERSON-BASED AUDIENCE DEMOS FOR ESPN
"With TVision we are able to add a deeper layer of insights to our already robust measurement offering."
SVP Business Development
We support measurement providers who are solving for a variety of use cases, have varying availability of data science resources, and have different timing needs.
This is the quickest option. TVision provides person-level probabilities for pre-identified scenarios such as demographic profiles, household type, time of day, and channel. Calculated probabilities overcome the challenges of building models from the ground up.
This is raw, sessionized data without any probabilities or assumptions built in. Sessionized data is flexible and transparent so your data engineering teams can develop their own probabilities and assumptions, and choose what is important for your solution.
Get a nuanced understanding of second-by-second TV viewing behaviors that enable ad-level measurement. With this output, our panel informs your model so you can recognize and apply person-level viewing information to larger census data sets.
TVision gathers second-by-second data from a nationally representative panel of households. Our in-home technology is capable of measuring everything that comes across the glass (both content and ads), which of our panelists are present in the room when the TV is on, and if they are paying attention.
WE REPORT ON:
Solve for knowledge gaps, and unlimited use cases with flexible data on how, when, and what people watch on TV.
TVision gathers person-level, second-by-second insight on viewer behavior across both linear and CTV. We deliver flexible, transparent, and privacy-safe data feeds so your engineering teams can develop and support their own solutions.