Thanks Rodney, I’m not sure what you mean by “preliminary”, but if you’re referring to the sample size I definitely agree that this needs to be taken with grain of salt. The issue for the studios isn’t as much the lack of available data as it is the lack of a “sandbox” through which to test its models. Without direct access to the consumer, studios aren’t able to do much A/B testing on their own models, which slows everything down considerably. The other issue is the very irritating survival of linear, demo-based research (panels), where it’s been shown over and over that there’s way too much complexity in audiences for it to work. Research methods in Hollywood are largely broken, but studios are doing a really good job at innovating in that space.
I checked both values you pointed out, and there are indeed identical.
Re. Netflix, I know them quite well, and without having their log files, I am very familiar with their methodology. This is another conversation … .