Well, for the record, before working on the article I've been liaising with Parrot Analytics for about the last two weeks. They walked me through some of their methodology. Basically, they measure demand as the desire for a particular TV show as expressed by audiences on a variety of platforms. They capture and combine billions of digital touchpoints including:
--> Social media engagement from platforms like Facebook, Twitter, Tumblr, etc.
-->Research and access to informational sites and critic blogs
-->Public streaming on YouTube and DailyMotion
-->Global file-sharing and peer-to-peer traffic from sites like uTorrent and Popcorn Time (they own the patent for this type of data collection)
The output is carefully weighted to reflect the level of engagement - so, for example, a file-share is a heavier weighting than a tweet. Essentially, they're assessing the level of engagement people have online with each individual TV show. Meanwhile, they also run a per capita adjustment to allow country-by-country comparisons, otherwise countries with higher populations (e.g. China) would dominate.
.
Parrot seem very reputable - to the extent that networks use them to assess the strength of their brands or to identify levels of demand in potential TV shows. Meanwhile, the Guinness Book of Records consider their methodology solid enough to use them as a data-source for their official "Most In-Demand TV Shows."
.
Interestingly, the data they've passed me on the Marvel Netflix correlates perfectly with that of other analytics firms I've spoken with. It does have one gap as regards Marvel Netflix; you see, based on some previous analytics I've worked with, my suspicion is that people engage with the Marvel Netflix shows as a discrete brand in its own right, not as individual shows. I can't prove that, however, but that's my suspicion.