Our friends at TVREV recently interviewed Inscape GM, VP of Product Management, Data Science and Analytics Charbel Makhoul about the new National Representative Panel and how it helps increase the value of ACR Data.
The interview, conducted by TVREV founder and lead analyst Alan Wolk, digs into the hows and whys behind the NRP launch. Read on to see how we're calibrating viewership data against census data to create a more refined, currency-grade data set:
Alan Wolk: What exactly is involved in creating the NRP, how are you making all those calculations?
Charbel Makhoul: It’s a fairly complex calculation, but what we do is we look at the distribution of VIZIO TV owners from whom we are collecting viewership data. Let’s say, for instance, that the census reveals that 6 percent of the population is between the ages of 18 and 24.
But then we find that 10 percent of the people in our Inscape data set are 18-24. So we calibrate the numbers in the National Representative Panel to take that into account.
Similarly, if 7 percent of our Inscape panelists live in the Boston area, but only 5 percent of the U.S. population does, then we can calibrate using that too.
We use a number of metrics, not just age and location, for the NRP. These include gender, ethnicity and household income and whether the household has a cable subscription or uses an over-the-air antenna. And then we calibrate against all of them in order to create our NRP.
Alan Wolk: What is it about your data that makes it “currency grade?”
Charbel Makhoul: When we model the viewing data we collect from our opted-in viewers with US census data, we are able to eliminate many inconsistencies. Those inconsistencies are not intentional—they occur because the profile of who owns a VIZIO TV does not match up one-to-one with U.S. Census data. By eliminating these disparities,we are best positioned to support our currency measurement partners, by providing data and solutions that are nationally representative with a high degree of alignment with the US population.
Alan Wolk: Who is this product designed for—who would want this sort of data?
Charbel Makhoul: For the most part, we feel that the audience will be our current client base of measurement companies, ad agencies and TV networks. While some of these companies have the resources to do their own analysis to achieve a national representative sample, many do not. They have used Inscape data in the past or they’ve used Inscape data via third party companies that make use of Inscape data.
Either way, they are always looking for more accurate data and by creating the NRP, we will be able to accelerate the adoption of currency grade data across the television industry.
ALAN WOLK: Mapping your data sets to the U.S. census seems pretty basic. Why aren’t your clients all doing this themselves?
Charbel Makhoul: It only seems basic. It’s actually a very complicated set of equations to pull off and requires a lot of resources and a lot of data scientists.
What’s interesting is that you can always keep on refining your data. You may decide you want to get more specific about gender, so you throw more data about gender on, only that skews your data about household income, so you need another data set to account for that and then you’re already at four data sets.
The thing about adding more datasets is that it makes the process far more complex than it needs to be, while adding in a much greater opportunity for error.
So one of our key goals is to reduce the number of datasets our partners and clients must rely on.
But what’s important to note here is that we are not trying to compete with our partners or with other measurement companies.
What we are doing is playing a catalyst role and trying to accelerate innovation around measurement accuracy by providing our partners with the key building blocks they need to make this happen. Our goal is to be a resource for our partners by giving them the tools and the means to achieve these sorts of measurement innovations.
Alan Wolk: Final question: what are the advantages you get from being a part of VIZIO?
Charbel Makhoul: VIZIO is the second largest TV manufacturer in the U.S. market right now and we have 21 million opted-in households to collect data from. The distribution of those households is not dramatically different from the U.S. census distribution. So the adjustments we are making to refine the data are not massive and that gives everyone even more confidence in the results.
To learn more about VIZIO’s National Representative Panel download the full white paper.
