Netflix’s Own Data Shows That They Had a Bad Quarter

Share on facebook
Share on twitter
Share on linkedin
Share on whatsapp
Share on email

– Netflix only released four “datecdotes” for film for Q2, one of the lowest totals on record.
– In TV, Netflix released 8 datecdotes, and they were generally lower than past quarters.
Luca has the most viewership, according to Nielsen, of any film through its first two weeks, with 28.6 million hours in week 2.


I was as indecisive as a Netflix viewer without a “Play Something” button trying to find a headline this week. Consider these three options:

– Through two weeks, Luca is the biggest film on streaming since March of 2020!
– Bosch becomes the first “season 7” in my data set!
– Manifest strives to become the biggest show of the year!

And yet those stories were overwhelmed by Netflix’s quarterly earnings report and the onslaught of “datecdotes” they provide. Or should I say “trickle” since Netflix was fairly stingy with their self-reported data. Given that Netflix has now provided about 120 of these gems, there’s a lot to unpack.

(Reminder: The streaming ratings report compiles data from Nielsen’s weekly top ten viewership ranks, Netflix datecdotes, Top Ten lists, Google Trends and IMDb to determine the most popular content. While most data points are current, Nielsen’s data covers the weeks of June 21st to June 27th.)

On The Website

In case you missed it, check out my coverage of Netflix’s second quarter earnings report. First, I published a “Visual of the Week” on Netflix datecdotes, specifically how often Netflix updates their “estimated” data points. Second, on Twitter, I covered Netflix’s quarterly earnings from a subscriber, financials and strategy perspective:


Here’s Netflix’s TV datecdotes from this quarter:

Last week, I speculated that streaming ratings might be down for Netflix. Well, this earnings report confirms it:

This might be Netflix’s worst quarter for original content since early 2019, based on the Netflix datecdotes.

(A datecdote is Netflix’s self-reported calculation of the number of households/“members” that watch 2+ minutes of a given show or film in the first 28 days globally.)

Crazy, right? Well, not that crazy when you think about it. The entire world shut down last spring. For broadcasters, cable channels, and streamers—who make shows and launch them immediately—this meant an immediate dearth of entertainment, starting in the spring and really impacting the fall. Since Netflix produces shows and then can spend up to 9 months (or more) in post-production, dubbing and editing, they didn’t have a content crunch last fall.

(That’s right: folks love the binge release model because they don’t have to wait for new episodes. This is an interesting psychological phenomenon: customers have to wait, they just don’t now they’re waiting.)

Here’s my high-level summary of Netflix TV series.



(This chart uses “season” datecdotes, filtering out any series below 20 million households. I’ve updated it from past looks to include both international and U.S. series, but the trends are roughly the same.) 

If you count by “total households” watching a show, Netflix is basically flat since they started releasing series. If you track by “percent of households” watching, then you can see that Netflix’s content is progressively reaching fewer households. A few more trends stuck out to me too…

First, the second seasons of earlier 2021 series didn’t hold up. 

Netflix released two second seasons of originals launched in January and March, and both had big drops. Specifically, Who Killed Sara? and Lupin had dramatically smaller second outings. (Technically, Lupin was a “Part II” of season 1.) Here:


Indeed, this quarter, it seems notable that Netflix is touting subsequent seasons that were down from previous seasons. In previous quarters, growth was the story, for example, with The Crown from S3 to S4. 

Long term, the most important shows (for streamers or broadcasters) are series that “grow” over time. Each season powers more and more viewers. Right now, Stranger Things and Money Heist have proven they can do that for Netflix. Bridgerton and The Witcher look like they will too.

Already, Lupin and Who Killed Sara? have shown they’re “good” shows, but not “growers”. Interestingly, Netflix also included Too Hot To Handle, and it had a drop off from its 2020 Q2 season 1 last year. Meaning the driver may be larger than individual titles…

Second, overall streaming on Netflix may have been down.

That’s what I speculated in my last “visual of the week” in case you missed it. 

That’s using the Nielsen top ten data. But the Netflix datecdotes seem to tell the same story. Here’s a chart comparing “similar” shows to each other:

If you try to compare shows that are similar to each other (either both reality or both fantasy or what not), essentially, most of the content from 2020 was simply bigger than its comp in 2021. In other words, another sign that overall usage of Netflix was down in Q2.

Likely everything from reopening to a bad content slate drove this. And one other thing…

Third, Netflix’s new “tentpole” didn’t deliver.

Reviewing the Nielsen data from the last year—and based on my personal experience in streaming—having one big series can help drive viewership for everything on a streamer. So in some ways, viewership is “correlated” with usage. Meaning one or two big shows bring in new or lapsed users, these folks start using a streamer again, and hence, the ratings go way up for everything.

Again, this concept isn’t revolutionary: shows that aired after Friends in the 1990s had a ratings bump and so did some shows that aired after The Big Bang Theory this century.

Thus, if Netflix doesn’t have a huge hit, it doesn’t drive overall usage. In Netflix’s long-term plans, the hope was that Jupiter’s Legacy, or maybe Shadow and Bone, would fill that hole. They didn’t deliver:

Only two new series made it into the top 20. In Q1 of 2021, for comparison, they had five. In Q4 of 2020, Bridgerton became the largest season 1 launch of all-time and The Queen’s Gambit took the sixth spot. Emily in Paris would have made the top 10 too. So this is one of the first quarters where Netflix didn’t launch a new top ten series. Meanwhile, Netflix didn’t have a major returning tentpole series either.

I could add to those two looks some other factoids:

– The U.S. shows performed fairly poorly. The Sons of Sam, The Circle and Too Hot to Handle S2 were three of the four lowest datecdotes Netflix has ever provided.
– A few potentially big shows were ignored by Netflix, including Jupiter’s Legacy, Lucifer season 6, Start Up and Manifest. Since they don’t own two of those shows and cancelled one, this makes some sense, but also shows that Netflix is having trouble touting their original success. For example, why not shout out Lucifer?

Increasingly—and in news that shouldn’t be too surprising—it looks like Netflix is as dependent on “hits” as any network historically. Yes, streaming changed the distribution and look, but the core principle of entertainment since it was born as an industry holds: hits help you win.

There is some upside for Netflix, though: this is probably as dark as it gets. Covid-19 shut down production for everyone. The problem is it impacted Netflix later than its streaming rivals. Going into the second half of 2021, especially Q4, their content should be back on track. So it probably won’t be this bad again, though competition may ramp up to meet them.

Quick Notes on TV

– Premiere: Bosch. Amazon’s Prime Video’s Studio’s longest running series (68 episodes, 7 seasons) happens to be a police procedural. For the third time this report, I’ll note that, maybe, the streamers are relearning the lessons of broadcast TV: cop shows are popular. I checked to see if Bosch is the longest running streaming series period, but it looks like a few Netflix titles have it beat by total episodes (Orange is the New Black at 91 episodes and Grace and Frankie at 78). Still, seven seasons and 68 episodes is a lot. (After Bosch’s second week, we’ll compare it to other Amazon Originals.) Meanwhile, Bosch is getting the spinoff treatment on IMDb TV, so yes, the fourth lesson of TV: make spinoffs of police procedurals.

– Premiere: Elite. The Spanish Netflix Original got the Netflix “datecdote” treatment, with 37 million global households tuning it. But clearly most of those were not in the U.S., as it earned a fairly low 2.9 million. In context…

– Premiere: The World’s Most Amazing Vacation Rentals. Yeah, this reality show premiered to 4.2 million hours of viewing in the U.S. Not that that’s bad; cheap reality can be very cost effective. But it does indicate that international content may not travel as well to the U.S. as Netflix sometimes implies or hints off the record to journalists if some random reality house hunting show can beat it.

 – Binge Release Decay Curves: I wanted to call out two series that had almost perfect “binge release curves”. (Remember, big first week, huge second, then a huge third week drop off.) They’re also probably the two most popular series until Manifest took over Netflix. (And, yes, Sweet Tooth earned a 60 million datecdote call out in the last earnings report.)

Manifest and Downton Abbey continue to lead an “acquired” titles boom for Netflix, combining for 43.9 million hours. Again, we’ll check in with Manifest in a few weeks when it loses steam, which won’t be until the week of July 19th at the earliest, based on the daily top ten data. Downton Abbey’s viewership is declining too, already in half from last week (9.5 this week from 21 million the week before. The binge release curve in action!)


Here are the datecdotes from Netflix for films in the last quarter:

While the datecdotes numbers were down for TV, in film, they are just sort of missing.

Look at that table. Four datecdotes! That’s all we got. That’s low:

And of those four films, none really stood out. (In this look, I filter out films below 40 million households.)

For film, two titles made the 9th and 10th spots, which actually isn’t a great performance. In previous quarters, at least some films made the top. 

Moreover, the missing titles feel glaring. It would have been interesting to see what The Woman in the Window, Dog Gone Trouble, or Stowaway did in the global viewership, given that they didn’t provide any data, presumably not great.

(And yes, for as much guff as I give Netflix for selecting random datecdotes each quarter, at least they give us some! Hulu only provides vague “most watched” titles, and Prime Video, HBO Max and others provide nothing. So at least we have 100+ examples of Netflix being somewhat open.)

Quick Notes on Film

– Premiere: The Ice Road. The Liam Neeson-helmed original debuted to nearly 12 million hours, good for 20th on our list. Based on the top ten data, it looks primed to follow most smaller Netflix titles, with a smallish launch, a stronger second week and then a drop off in week 3.

Luca: What a hold! 28.6 million total hours in its second week. This is the biggest first two weeks of any film in my data set. (Two potential exceptions would be Spenser Confidential, which I don’t have week one data and Wonder Woman 1984, which didn’t get a second week of data for Nielsen.) Still, Disney has their biggest film launch yet. (Will Black Widow beat it?)

Anecdata of the Week 

Normally, we skip linear TV ratings in this space, but this story is fun: Formula 1 has seen a boost in ratings over the last year.

Here’s some data from Hedgeye Communications (link to subscribe here):

Why are ratings up? One explanation is Netflix’s docuseries, Drive to Survive, boosting F1’s popularity. Anecdotally, I’ve heard this too. The only thing holding me from going all in on this theory is that the ratings were surprisingly shallow in the Nielsen ratings. Formula 1: Drive to Survive only made the top ten list for one week for Nielsen, with a low 4 million hours viewed. Does that alone explain the boost in linear ratings?

So what are the takeaways? F1 ratings really are up, and likely the buzzy streaming show helped somewhat.

Not to mention, it’s funny that we talk about “cord cutting” so much when it comes to streaming/Netflix, but this would mean that folks are watching a streaming show, then somehow watching F1 on linear TV. Hmmm…

Coming Soon! 

– So Netflix had a bad quarter. Does Q3 look better? Slightly. Money Heist is a performer in every non-U.S. territory. It will help with subscriber numbers in the EU. After that, here’s Netflix’s quote about content: 

Never Have I Ever season 1 had 40 million households tune in during Q2 of 2020 (a big quarter for streaming) and Sex Education season 1 did 54M in Q1 of 2019. (We didn’t get a datecdote last year for season 2.) So do those move the needle? I don’t think so. That said, the weekly top ten data shows both Never Have I Ever and Virgin River taking over the top spots from Manifest this week. We’ll see how much viewership that translates to.

– The Olympics start today. Will those take a bite out of streaming ratings? It will be fun to see. (It will in my house. We watch non-stop.)

– There are a few straight-to-streaming shows we’ll probably never find out how well they did. Schmigadoon and Ted Lasso season 2 on Apple TV+, Dr. Death on Peacock, and Making the Cut on Prime Video too.

The Entertainment Strategy Guy

The Entertainment Strategy Guy

Former strategy and business development guy at a major streaming company. But I like writing more than sending email, so I launched this website to share what I know.


Join the Entertainment Strategy Guy Substack

Weekly insights into the world of streaming entertainment.

Join Substack List
%d bloggers like this: