We often work with clients whose brand name is asked in questions in our survey and listed in our taxonomy. And sometimes we'll get questions as to why our survey data can differ from your 1st party data or Tag data. Most of the time, this can be attributed to different methodology.
Let's go through two examples.
Tagging your Login Page
You've tagged your login page, and when you build an audience with your Tag, you see there are 3 million people. But when you build an audience in Resonate using our corresponding survey attribute value, the audience size shows 4 million. What should you do?
In this case, rely on the login Tag date for the addressable audience size rather than the projected audience size for our corresponding survey attribute value. While the population sizes do vary due to differences in methodology in how the data is captured, the insights derived from the survey are reliable.
For competitive analysis, Resonate can provide a good estimate for the size of your competitor's audience, and the insights found in the platform can help guide you in knowing how their audience differs from yours.
Known Audience Size via 1st Party Data
Let's say you're XYZ cable provider, and you know that you have 4 million cable subscribers. But when you build an audience in Resonate using our corresponding survey attribute value, it shows there are 8 million subscribers to XYZ cable. Why the discrepancy?
First think through whether your data is reported at the household level or the individual level. In this case, it probably is household level. Our data is based on individual devices, and we see an average of 2.3 devices per person.
In this case, you should rely on your number for sizing the audience but it could still be helpful to look at the audience size from our corresponding survey attribute value, as that will be at the individual level, not the household level. As mentioned above, insights derived from our corresponding survey attribute value of consumers who subscribe to XYZ cable are still valid.
It All Comes Back to the Survey Question
It's also always a good idea to take a look at the underlying survey question itself, which is shown in the platform both when you're building an audience and when you're viewing insights.
As an example, let's look at the survey attribute value for Wall Street Journal. When we read the survey question - noted in the red box below - we can see we're asking whether people have read the Wall Street Journal in any format (print, website, mobile/app, tablet/e-reader) at least 3 out of the last 7 days. This number will most likely not line up with any number that the Wall Street Journal publishes due to the unique nature of the question itself.
So, just a helpful reminder to keep the survey question in mind.