Yesterday, Netflix had their quarterly earnings report and dropped ten new datecdotes—my term for Netflix’s measurement of the subscribers who watched 2 minutes in the first 28 days—on us. Add to that the Snyder Cut coming out on HBO Max, a new streamer on the “acquired film list”, a new superhero show on Disney+, and we have a ton to cover!
Television
Long term, this streaming ratings report will summarize multiple ratings sources every week to get the best estimate for what streaming content is doing well. Anything that meets my standards—article to come—will be included.
But not all data sources will be ranked equally. Any data about actual “viewership”—meaning measuring how many people are watching how much of a given title—is/will be more highly valued in my system. Because they’re “sources of truth”.
Actual viewership is the goal for any TV show or film. Lots of data points—like IMDb Ratings, Google Trends interest, top ten rankings, social media conversation—are related to viewership. In statistical terms, they’re correlated with viewership. (The more folks who search for a TV show, the likelier they are to watch that show.) We can use these data sources to predict viewership.
But viewership is the goal. What we are trying to predict. The “y” value. The “actuals”. The source of truth.
That is what makes every “Netflix earnings day” so exciting for the EntStrategyGuy. I get to add a bunch of “sources of truth” to my data set. (And since they come from Netflix itself, they’re *almost* as accurate as you can get.) Currently, the Netflix-released data set is up to a whopping 110 data points. Here is the updated quarterly look at how well Netflix’s content is performing, both “total viewership” and “percentage of subscribers”:
Even better, the two sources of truth we have—Nielsen and Netflix—generally match. There’s some international variation—which makes sense—but for the most part, the series Nielsen’s data says would be the biggest match the series Netflix said were the biggest.
To note: Netflix and Nielsen measure different things (subscribers versus minutes) in different geographies (global vs. the U.S.), but together, they tell roughly the same story.
How did Netflix do for the quarter? Well, the Nielsen data told us that Netflix started very, very strong, then declined as the quarter went on. (Read last week for details.) That’s what Netflix’s data shows: The new U.S. series were good but not elite. Here’s a table with just season 1 launches for context:
What about Cobra Kai? Compared to other season 3 launches, it was fine, but you can see the gap between it and an “elite” series like Stranger Things:
There is one outlier, though, which is Lupin. Its five episodes are a global juggernaut, especially in Europe, whereas it failed to resonate in the U.S. Why didn’t it drive more usage? It’s short, only five episodes long, averaging 46 minutes per episode. That just doesn’t sustain interest as much as a 10 episode series like Bridgerton.
The big question going into earnings—which I asked last week and in this thread—would be if the decline in viewership would hurt Netflix’s subscriber growth. And if the decline in viewership was driven by a shortage of new content due to Covid-19 production pauses.
Indeed, Netflix said in their report that this indeed was what was happening. Mystery solved!
So when will Netflix’ rebound? Look towards the second half of the year. That’s when Netflix said they will be past the Covid-19 slowdown. We’ll see new seasons from Money Heist and The Witcher—which are two of their top five series all time—and the film Red Notice, the Rock/Gal Gadot/Ryan Reynolds aspiring blockbuster. The fun question for viewership in 2021 will be if competitors like Hulu, Disney+, Prime Video and HBO Max can take advantage of Netflix’s lighter Q2 to grab customer usage/mindshare.
Other Quick Notes on TV
– Speaking of Disney+, the latest season of Falcon and Winter Soldier is out and it looks like it’s continuing the success of The Mandalorian and WandaVision. Like those other weekly-released series, we should wait to judge performance until all the episodes are out, but this is a launch in the right direction, with a higher total and per episode viewership than WandaVision’s launch.
– Also, WandaVision stayed on the “originals” top ten list, even without new episodes, so clearly, some folks did start that series late and are bingeing to catch up.
– This week, Law & Order: SVU made the “Nielsen Acquired TV” list, having been watched for 5.3 million hours. This is noteworthy because it’s the first time a non-Netflix title has made the “acquired” TV list. In the past, Amazon and Disney originals have made the “original” TV list but not the acquired list. (“Acquired” titles are series streamers license for a “second run” after their initial run on broadcast or cable.) This directly relates to the thesis that Netflix is seeing a general decline in usage, since non-Netflix library titles like Law & Order SVU are showing up in this viewing list.
– Not only that, but since Prime Video is still represented with Coming 2 America, also for the first time, the Nielsen Top 30 features four different streamers (Netflix, Hulu, Disney+ and Prime Video)! (Even crazier? Potentially a fifth streamer made the list, if Nielsen was tracking their releases. Which I’ll discuss in film next.)
Film
On March 18th, HBO Max released a new, four-hour cut of Zack Snyder’s Justice League. Fans had long advocated for this new version, and HBO Max spent $70 million to make it a reality. Since Nielsen’s data is delayed by four weeks, I’ve been waiting to opine on the Snyder Cut until we knew if Nielsen would provide a ratings estimate. That would be this week’s data (starting for the week of March 15th).
Alas, Nielsen did not provide a data point.
Which leaves us to use the proxies I mentioned earlier. In this case, knowing the top film titles in the U.S. from both Nielsen and Netflix, we can compare using Google Trends.
If Google Trends is to be believed—and wait a moment for why it may not—then arguably the Snyder Cut was the fifth buzziest film released since Christmas, only lagging Wonder Woman 1984, Soul, Coming 2 America and Godzilla vs. Kong. It was more popular than I Care A Lot (Netflix’s biggest U.S. film this quarter, according to Nielsen) and about as popular as Raya and the Last Dragon.
The limitation with Google Trends is that it measures search traffic, which can be artificially high for newsworthy titles. Of which the Snyder Cut certainly was buzzy. So interest may not translate to actual viewership.
However, a few third party analytics firms tried to measure how well it did. According to Antenna, reported by Bloomberg, Justice League drove an 8.9% increase in HBO Max app downloads. SambaTV, reported by The Observer, said 1.8 million folks watched The Snyder Cut on its opening weekend, behind Soul and Wonder Woman 1984, but ahead of WandaVision and The Falcon and the Winter Soldier. In general, the consensus from third party firms is that the Snyder Cut “won” its opening weekend. The one outlier is TVision, which said the Snyder Cut had the smallest share of streaming of HBO Max’s launches this quarter. (Brandon Katz has a good round up of sources, and Scott Mendelson covered some as well.)
That said, the positive news is for Godzilla vs, Kong, which did even better in search interest in the U.S. It will be interesting to see if we get a data point for it next week, as HBO has already said it’s their most viewed title of the year but hasn’t provided a specific number.
Other Quick Notes on Film
– For another week, both Raya and the Last Dragon and Coming 2 America stayed on the feature film list, but both decayed even further. Tracking how long films stay on the top ten list in 2021 will be a fascinating project!
– Yes Day had the typical bigger second week in total viewership (9.9 million hours) but decline in viewership per day (1.4 from 2.2).
– Speaking of, The Dark Knight fell off the top ten list and was roughly replaced with Savages as the latest licensed title to do well when it’s added to Netflix. Even with these licensed titles, we see that their shelf life is fairly short, even for a title as huge as The Dark Knight, which is a heavily rewatched title.
– For space, I cut the discussion of Oscar titles and Netflix’s feature film datecdotes. I’ll cover those in future articles.
