A quick guide to affinity analysis
The goal of an affinity analysis is to determine how much overlap there is between the audiences of series.
Let’s say that we have 100 people who watch Show A in the USA. If 20 of those 100 also watch Show B, and 10 of those 100 also watch Show C, then the affinity of Show A to Show B is twice as strong as the affinity of Show A to Show C. Note that these groups are not exclusive so the same person could be one of the 20 and one of the 10.
Put simply, the higher affinity means that a viewer of Show A is more likely to also watch Show B than Show C.
So how do we get this information?
Parrot Analytics has an industry-unique dataset of consumption from the daily tracking of free streaming/downloading we perform as part of our television demand measurement. We can run market-specific analysis on this consumption dataset to see the relationship of a show to any other show, provided of course that there is enough streaming/downloading activity! For example, this means that we cannot get affinity directly for pre-release series as there is nothing released to stream or download. (Note that we can still do this indirectly using genome modeling, however.)
As affinity analysis is based on actual consumption data for consumers in this specific market, we are confident it is an accurate way to answer this question.
In DEMAND360 Lite, affinity is surfaced on each Show Details page as the collection of shows that "People are also watching." You can access these collections by scrolling down to the bottom of a Show Detail page.