(For Part I of this series on Netflix versus Crazy Rich Asians, click here.)
If you can’t tell by my article last week, I had a lot of fun with my comparison between Netflix romantic comedies and Crazy Rich Asians. Unfortunately, I had a lot of ideas for that article that hit the cutting room floor.
Some because they were too speculative, some to save room, and some to make a tighter narrative. (I had tried a long shot publication at a bigger outlet.) And some because I couldn’t prove them. So for a respected publication, it didn’t make sense.
But this is my website. I’m free to make all the speculation and ask all the tough questions I want to here. Since Netflix only provided me one number—80 million customer accounts watched an original romantic comedy the last summer—well, I want to ask that number a lot of questions. I want to interrogate that number to within an inch of its life. So that’s what I’m doing today. Asking—and answering—all the other questions inspired by this comparison.
What other “circumstantial” evidence did you leave out?
A few pieces, but one major one. Essentially, the major studios stopped making romantic comedies for two reasons. First, they don’t have a high enough “ceiling” in that they don’t ever tend to have billion dollar movies. Second, and crucial for our math, is that they also don’t tend to perform well overseas. This applies generally to all comedies. Comedy is a local phenomena so it’s rare for comedy films to do well overseas unless they are very, very broad. (Some of the broad sitcoms like The Big Bang Theory or Simpsons do travel. Others, I’ve heard, don’t.)
We’re seeing this right now. Aquaman is the number one movie in the world…and it didn’t open in America. Crazy Rich Asians, meanwhile, flopped in China. To show this effect, here’s some data. For my series on Disney’s Lucasfilm acquisition, I made a data set of 50 “franchise” movies. These provide a good set of comps for comic book movies and their ilk. As you can see, franchises now see 63% of the total box office come from overseas (and even this still includes a lot of old Star Wars and Indiana Jones data.)
Now compare that to romantic comedies. I don’t have as large a list, so I pulled some sample romantic comedies. The trend is clear…
Four recent romantic comedies that did “well”, had over 70% of their box office come from the US market. Crazy Rich Asians, notably, only had 22% of its total box office come from overseas. Compare that to massive blockbusters like Avengers or Pacific Rim, where over 66% or 75% of their box office came from overseas.
This has implications for Netflix. Mainly, three facts collide that can’t all be true simultaneously:
– Netflix had 80 million customer accounts watch an original romantic comedy last summer.
– There are 60 million US customer accounts. (Rounded up slightly.)
– US romantic comedies tend to have 60-40 splits in US to international viewing, sometimes as high as 70-30.
This puts us in an awkward place when it comes to the Netflix number. Based off Crazy Rich Asians and other romantic comedies, I could easily assume 60% of the viewership was US based. That leads to some really tricky “consultant math”. Go with me on these assumptions:
– Assume 60% to 40% domestic to foreign split on Netflix romantic comedies
– Assume 1.4 “viewers” for every Netflix customer account.
– Black Panther sold 76 million tickets in the United States.
– Assume 15% rewatch rate for Black Panther.
Here’s those assumptions, now it table form:
Now, even worse than Netflix claiming it beat Crazy Rich Asians, if we take some conservative assumptions, more people watched a Netflix romantic comedy than the biggest movie in the US last year (Black Panther). Do you honestly believe that?
Moreover, it seems like just a massive amount of Americans watched these Netflix romantic comedies. If the skew were more in line with the recent romantic comedies (say 70% US customer accounts), well then nearly 56 million Netflix customers watched. That’s over 96% of Netflix US customers and 1 in every 6 Americans. That’s so high that it makes me think something is wrong.
Either romantic comedies have started traveling well globally—which I doubt—or nearly every Netflix customer watched a romantic comedy—which I doubt. You see where this leaves me?
Any other circumstantial data?
I should define circumstantial. In this context, I’m describing that we know certain things about how entertainment works. Evidence that flies in the face of this is inherently suspicious. This is my version of “forensic accounting”. That’s going through a firm’s books looking for irregularities that don’t prove a crime in and of themselves, but indicate that something suspicious is going on. (Or money is being stolen.)
In this case, I return to one of the core principles of entertainment: the “logarithmic distribution of returns”. Or “winner take all economics”. Do you numbers match up with a principle that proves itself repeatedly in every part of entertainment?
In Netflix’ case, it is phenomenally unlikely that each romantic comedy was equally successful. Or if they were all equally successful, that means that really they didn’t have a true hit. That just isn’t how movies work. The most popular is not the leader by a bit, but by huge margins. In other words, it tends not to just be popular with its core demographic, but expands to the culture at large. We saw this with Black Panther and Crazy Rich Asians. They both became cultural phenomena, doing multiples better than other films in their category.
We have hints from Netflix—well more than hints, they told us—that To All The Boys I’ve Loved Before was their most popular show. (As I was writing this, they also said The Kissing Booth was highly rewatched as well.) But by how much? They didn’t tell us, which makes me think that it wasn’t multiples more popular in terms of customers, except in maybe rewatch rate. But if it were truly a global mega-hit, we’d have seen it in more popular measuring methods like Google Trends. Instead, it was the biggest Netflix show, but just by a bit. Here’s a reminder of the Google Trend data:
So let’s assume To All the Boys I’ve Loved Before and The Kissing Booth were twice as popular as the rest of the rom-coms, then would it make sense that all the rest of the original romantic comedies evenly divided the 80 million customer accounts?
Not at all!
I haven’t written about customer segments, yet, but we know Netflix absolutely believes in them. They claim to have over 1,300. So does it make sense that romantic comedy fans tend to watch romantic comedies? Obviously. I mean, if I watch a superhero movie, Netflix offers me more big budget superhero movies, not teen romantic comedies.
To get to a huge number like 80 million customers, Netflix is either saying each show performed equally well with different audiences—which is incredibly unlikely—or that they had a base level of romantic comedies, with one or two “hits” that brought in a ton of extra people, that we didn’t see in the popularity data.
We can either believe the firm who is keeping their data secret, or core economic principles. I tend to side with economics.
Well, fine, tell us what you do think does explain the situation?
Well, the “inadvertent view hypothesis”.
This would be that a large number of the “customers” were not actually interested in the romantic comedies, but somehow engaged with them while scrolling through Netflix. These engagements get counted as “customer count watched”. For example, count trailer views as views of the original. Or auto-play.
It actually explains all the numbers. To All the Boys I’ve Loved Before could still be the most popular, but still have very reasonable numbers. Meanwhile Netflix easily has 80 million customers constantly browsing its platform, and since they prominently display newly released movies, it would be easy for customers to inadvertently view romantic comedies and count. Especially with the auto-play feature on some trailers.
Again I can’t prove this, but it makes more sense.
How could you prove it? Or how could Netflix disprove it?
One of the keys to datecdotes is to limit the number of categories you release. Essentially, the most important two numbers are customers and then the hours they spend watching a show. Netflix gave us the former but not the latter. If we had both, we’d be much closer to understanding the true popularity of Netflix original romantic comedies.
Take this chart that I made up myself:
These are three potential “total hours viewed” of a Netflix Original. And they all tell a completely different story for how “well” these films did. Assuming each film was about two hours in length, if everyone watched a full movie, you’d get the high number of 160 million total hours. My money would be that if Netflix knew that the total hours viewed was this great, they would have told us.
Or, let’s say Netflix did have 80 million customers watch an original, but if you added up all the “hours viewed” it came to 20 million hours viewed. In that case, the internet would tear the streaming company apart. Who cares if 80 million accounts watched part of a film and then stopped?
Most likely, the number is somewhere in the middle. That means that it’s complicated, which is likely where the number landed and why they didn’t tell us. Just how low is the key question. And rewatch rate would complicate this further.
(For those who are curious, viewerships by total hours tends to look like a reverse “J” with streaming shows. Lots of people watch the beginning, but drop out within the first ten minutes. Very few people drop out at the midpoint. Then a lot of people watch the whole thing, just less than who started it.)
Next time, I’ll dig into some business lessons we can learn from Netflix, along with some suspiciously recurring numbers.