When Netflix releases their quarterly datecdotes (Click here for my explainer on that term, my phrase for any Netflix data point about households watching 2+ minutes of a film in the first 28 days), they tend to come in two flavors.
Like, for example, this tweet:
72 million households are betting on dead.
ARMY OF THE DEAD has been the #1 film around the world and is projected to be one of Netflix’s most popular films ever in its first 4 weeks. pic.twitter.com/85foTPFAny
— NetflixFilm (@NetflixFilm) May 28, 2021
Or this quote in a press release:
Note the words “projected” or “on their way”. Every so often, Netflix releases datecdotes early, and I categorize these as “estimated” datecdotes. Sometimes these get updated later. For example, in Netflix’s quarterly earnings reports:
Note the words “chose to watch”. If it’s past tense, that’s an actual number. Future tense are estimates.
The suspicious among you may wonder: why does Netflix update some titles and not others? For example, Thunder Force never got an “actual” update like its fellow titles Army of the Dead and Fatherhood this quarter. Well, my answer is my “visual of the week”:
Hmm, Netflix only updated their estimates when they were higher than initially forecast? How many times has Netflix not updated their estimates? Actually about 15 times, six for TV and nine for film. Here’s a table showing the updated titles and the missing film estimates:
Insights and Implications
– Of the film or TV series that have been updated, the average bump upwards has been about 17%, or 9.3 million, in total households.
– Yes, it is fair to assume that missing datecdotes would have shown smaller numbers. Unless Netflix projections are deliberately conservative—and good forecasts get as much over as they do under—the misses should be evenly spread out. In fact, Netflix often emphasizes this point when they provide subscriber forecasts each quarter:
– The missing film updates are more glaring than the TV series updates. All of the “estimated” TV series came in Netflix’s quarterly earnings reports (called 10Ks), which is usually where Netflix provides updates. So my gut is these are probably pretty accurate. (Some were only missing a handful of days, for example.)
– That’s not the case with film, where only two estimated film titles were in earnings reports. The other seven were in tweets or press releases. Netflix has routinely touted films like The Midnight Sky, Enola Holmes or Thunder Force that later could have been updated in their 10Ks but suspiciously were not. I do think it’s because they ended up lower than Netflix anticipated. (And I speculated this could be the case with Army of the Dead, but I was wrong.)
– If you’re curious, my data set now includes 124 total “datecdotes”, 62 for film and 62 for TV. Five of these are “unofficial”, being leaks from talent or something along that line.
