Streaming
A Streaming View Is Whichever Number Flatters The Service
Viewing figures depend entirely on how a view is defined, and because services set their own definitions, comparable-looking numbers can describe very different amounts of watching.

Streaming services publish viewing figures that appear to be counts of something definite. They are counts of whatever the service has chosen to define as a view, and the definitions differ substantially.
The definition does most of the work
A service can count a title as viewed when a couple of minutes have played, when a defined proportion has been completed, or when a household finishes the whole thing. Each produces a different figure from identical behaviour.
Hours watched divided by running time is another common approach, which converts total consumption into an equivalent number of complete viewings that nobody necessarily performed.
None of these is dishonest on its own terms, but they are not interchangeable, and a figure quoted without its definition cannot be compared with anything.
The counting unit is not a person
Accounts are shared, profiles are ignored, and a television in a living room may represent one viewer or several. Services generally count accounts or devices because those are what they can observe.
That makes per-viewer comparisons unreliable, particularly between titles that skew towards communal viewing and those watched alone on a phone.
It also means the same household can appear as multiple viewers or as one, depending on which devices were used and whether anyone signed in separately.
Windows are chosen after the fact
Figures are usually reported over a defined period after release, and the length of that window has a large effect on which titles look successful.
A short window favours titles with concentrated demand, while a longer one favours those that accumulate steadily through recommendation.
Because the reporting window is selected by whoever is publishing, it tends to be the one that presents the result most favourably.
Internal numbers are different from published ones
Services track completion rates, drop-off points, repeat viewing and whether a title brought in or retained subscribers. Those measures drive commissioning decisions.
Published figures serve a separate purpose, supporting marketing and negotiation, and they are released selectively rather than as a complete record.
The gap explains why a title can be publicly celebrated and quietly not renewed. The number that was announced is not the number the decision was made on.
Independent measurement is partial
Outside firms estimate viewing through panels and device data, which provides comparison across services but covers particular territories and screen types unevenly.
Those estimates typically miss viewing on phones or outside the home, and they arrive later than the services' own data.
The practical consequence is that no complete public picture exists, and most confident statements about what is being watched rest on partial measurement of a partly defined thing.
Questions readers ask
Why can I never find older films on a service that has them?
Usually thin metadata. Older licensed titles often carry only a synopsis and a genre, so the recommendation system has almost nothing to work with.
Are recommendations based on the film itself?
No. They are based on structured descriptions of it plus viewing behaviour. The system has no direct access to what the film is actually like.





