Back in the halcyon days of April, Netflix had just crushed another quarterly earnings report and it was riding high. In Decider, I said their report had something for both sides—for the haters and the lovers, skeptics and the supporters, bears and the bulls.
Well, Netflix finally had a bad earnings report.
The most fascinating thought, to me, was this one by Gene Munster:
“As much as I love the company, I just think its best days, unfortunately, are in fact behind it…I think we’re going to look back at this quarter as one of the pivotal moments in the Netflix story.”
If the laws of entropy are indeed correct, well at some point, every company’s best days are behind it. Unfortunately, we hardly ever realize this in the moment. This doesn’t mean the companies go out of business a la Blockbuster—IBM is well past it’s high water mark, but it’s still around and publicly traded—and it doesn’t even mean the stock price will decline—since stocks in general have gone up in general even faster than inflation. But at some point everything declines.
So is this the moment of Netflix’s high water market? Honestly, it may be. But we won’t know for sure until years from now.
To figure it out, I’m going to dig through Netflix’s last earnings report for the strategic insights I can find. As a reminder: I’m not here to give you stock advice. I’m here to critique strategy and Netflix’s quarterly reports are the best time to update my priors/data on Netflix’s strategy. Today, let’s discuss meta thoughts and content strategy; tomorrow I’ll go over strategy, subscriber and financial thoughts.
At Least Netflix Gives Us Financial Data to Parse.
Let’s praise Netflix for one thing to start: producing this document in the first place.
If Apple had bought Netflix in 2015, Netflix would have become an operating segment, which means that Apple could pick and choose selected numbers to release about their performance. Likely they would have hidden as much as possible, they way they now hide iPhone sales. So I’d have much less data to judge them on.
To get a feel for this, take a gander at AT&T. We used to get a lot of HBO data every quarter—even as part of Warner-Media—but since AT&T acquired them, they went back to not reporting on HBO specifically. Meanwhile, if HBO were a standalone company, we’d have even more data than both previous reporting situations. The current situation leaves us guessing about their revenue, operating income and subscriber totals. We only get little tidbits if AT&T deigns to give it to us.
If we had to power rank the streaming platforms based on data released, right now it looks like this:
4. CBS All-Access
6. Amazon Prime/Video/Studios
And all of them pale compared to the networks and TV channels of old who had TV ratings released every day and provided us financials. To Netflix’s credit, they give us their financials to make columns like this possible.
What is a “Netflix Killer” Anways?
Alan Wolk had a good article at TVRev clarifying that Netflix won’t actually disappear anytime soon, which is a statement I wholeheartedly agree with. Why, then, do so many headlines have “Netflix Killer” in them?
Well, fuzziness in definitions. For a lot of folks, Netflix is one of the most over-priced companies in the world. They’re usually reacting to folks who think that Netflix is destined to conquer all of television. So you could reasonably say that any of the following end states is the “death of Netflix”, depending on your point of view:
1. Netflix suffers a few bad quarters and ends up with a price-to-earnings ratio around 20-25. (To show the gap, Netflix is currently at 123; most media firms trade between 15-20; Disney is currently a 20.5.)
2. Netflix is acquired by another larger digital company. (I recommend Facebook in this article.)
3. Netflix becomes the 3rd or 4th most subscribed OTT platform in America and/or the world.
4. Netflix goes out of business.
This is how I can think that Munster may be right—Netflix’s best days are behind them—and that Alan Wolk is right—there is a no “Netflix killer”. It depends on the definition. My personal opinion is that option 3 above is exceedingly likely, which means Netflix should valued like HBO, not like Amazon. Netflix is here to stay, but maybe not one of the most highly valued companies in the world, which may be death depending on how much stock you hold.
How do you evaluate the biggest spender in Hollywood’s performance when they dole out so little data? By my count, they’ve released 17 “datecdotes” going back to the Q3 2018 earnings report. They’ve doled out a few more to news outlets over time, like this one to Reuters, this one to Variety or this tweet for Stranger Things last week.
(Meanwhile, Netflix now has a “top ten” list in the UK, which I haven’t even begun to parse for data yet!)
Well, we work hard to find some “signal” that we can trust in this paucity of noise. But before we go any further, let’s combat the biggest risk in this enterprise, which is coming to the wrong conclusions.
Combating the Availability Heuristic
Is little Netflix data worse than no data? I’d argue yes, because it skews our perspective because of the “availability heuristic”, a concept pioneered by behavioral economists Kahneman and Tversky. Here’s the Wikipedia article, and my key quote from a few weeks back I’ve been sharing repeatedly:
I’m about to go deep into Netflix content data, filling your brain with easily available datecdotes about how great Netflix is doing. To combat that, I built you this table to study. It’s a selection of the shows you may have heard of that Netflix released since then June 1st going to July 4th. The goal is for you to study it, and now you can recall exactly how many shows Netflix DIDN’T release data for:
The green is the two data points we did receive since June. The yellow is five notable series that I’m surprised didn’t get data points released. And I didn’t even include every show or film or documentary, just the content that I believed Americans would have heard of.
Did Q2 Really Under-perform in Content?
Netflix said it did, and that the lack of big hits hurt the subscriber numbers in the US and overseas. With Stranger Things 3, Orange is the New Black and The Crown coming in Q3, they won’t have this problem again. Well, at least for one quarter.
Does this hold up to their datecdotes? Here is the same data I’ve been using, just reframed by quarter:
Sure, they’ve released slightly more datecdotes, but it still seems like the current crop of content did better than Q1 of this year. The conclusion, then, is that we need more data to answer the question. If that’s the case–that Netflix really didn’t have good content this quarter–then why does it matter how many people watched Dead to Me and Murder Mystery?
Putting Q2 Movies in Context.
A key challenge when we hear a Netflix datedcote—take that 48 million watched The Perfect Date—is to understand what that means in terms we understand, like box office. Would that have been a smash hit? (Ted Sarandos himself made this claim about The Christmas Chronicles.)
Well, according to Google Trends, um, no. Take The Perfect Date. In this table I shared last week, I couldn’t even get that film to register in public interest in the topic.
So maybe a ton of people watched it, they just didn’t care to find out anything else about the film afterwards. That’s odd.
I should tamp down expectations just slightly on the Google Trends tool a bit given how much I hype it. While it matches fairly well with what I expect on interest for the majority of movies and TV shows—and Google has said it can use it to forecast box office four weeks out—it is correlated with success, not causal. There will absolutely be examples of films and series that underperform this one measure of interest.
That said, at this point I feel comfortable saying that Bird Box was Netflix’s one true hit. That would have been a Get Out, Quiet Place or It in terms of global box office. Triple Frontier and Murder Mystery were probably $100 million box office films. Everything else? Way less.
Putting The Q2 TV Series Into Context.
TV is much trickier to put into context, which is why I’ve focused on films. Mainly, I just don’t accept Netflix’s TV completion definition of “70% completion of one episode” as a unique viewer for a TV series. Sure, if you watch 70% of a movie, you finished it. But 70% of one episode of TV? If you didn’t like it, you’d cancel it. That’s why most TV series see their ratings fall from their premieres.
Again, we can go to the Google Trends to see what 30 million unique viewers globally for a TV series means compared to say, Game of Thrones, that HBO says has 44 million unique viewers in just the US plus digital. Here is the worldwide view:
In other words, despite the big global numbers, nothing comes close to Game of Thrones. Or Stranger Things, for that matter. So let’s pull out the two biggest shows on TV to see if we can get a clearer picture:
My takeaway here is that most of the trendy critically acclaimed TV series right now are all about the same in popularity between When They See Us, Big Little Lies, and The Handmaid’s Tale. So when Netflix says When They See Us is a hit, it’s about as big as those. (Though no matter what show I run this with for Amazon, either Fleabag or Good Omens, they lag well behind HBO, Netflix and Hulu.)
Of course, the key difference between HBO/Hulu and Netflix is the issue of how many shows. Netflix is making many, many more shows, but still getting about one hit. Let’s explore that.
The Real Impact of Content Loss is Less Efficient Spending By Netflix…Which is Bad
If you’ve been following Disney’s inexorable march towards box office domination, you know how well their films are doing. This success, though, has in the past redounded to Netflix’s benefit. Netflix had quite a great Disney quarter in Q4 of last year when Black Panther, Avengers: Infinity War and Incredibles 2 all dropped on its service. Those were the three biggest US movies of last year.
Netflix said this won’t be a problem. Here is their specific quote about how well Pay 1 movies do on their platform versus originals:
This is my current candidate for the most misleading statement in Netflix’s last earnings report. Note they don’t define their terms. Is it in terms of hours consumed? In terms of demand and likelihood to renew? By unique viewers? Does this mean original are bigger than their biggest licensed film, or just the average licensed film?
But it doesn’t matter, because even if Netflix has more films that are genuinely popular, the question is about total volume of content but how much it costs to get there. My working theory is that Netflix is losing the most efficient content on its platform by a large margin.
The Disney/Netflix deal may have been the best use of content licensing spending by any company in contemporary media. The Disney/Netflix deal was suspected to be between $200-400 million per year for Disney’s whole output. When the deal was extended in 2015, neither Disney nor Netflix realized the run Disney was about to embark on, so the price wasn’t reported to be much higher. Disney’s been the top studio for the last four years, and had 3 of the top five films every year.
Getting all of Disney’s output was a steal for “only” a few hundred million dollars out of Netflix’s $15 billion content budget. On the open market, The Force Awakens by itself probably would have commanded nearly 9 figures in pay 1 licensed value. (That film helped Disney sign a massive TNT Star Wars deal through 2022.) Then consider that Netflix got all of Star Wars AND Marvel AND Pixar AND Disney animation AND Disney live-action…
(Arguably The Office and Friends were incredibly undervalued too.)
All this content will be incredibly difficult to replace. So Netflix is going to go from a deal where the hit rate for top movies was incredible—literally only Disney itself in the 1980s or maybe Warner Bros in the 2000s had the same hit rates—to one where they have to prove they can make the movies. And as my updated feature film table shows, Netflix is average at making films.
You would need the chart above to be heavily skewed at the top to prove that Netflix was above average in hit rate. But they aren’t. In the last year Netflix has only had two films get over 80 million viewers. And again, the totals are less important than the relative size of each bucket.
(Both the table and chart were updated from the initial Twitter versions after some proofreading. My key assumption is that, since Q4 of 2018, any film or TV series without a leak yet that got over 30 million viewers would have made it into an earnings report, which seems more than fair. If you want my data to cross-check my work, email me. Or if I missed a data point, send me a note.)
I’ll repeat this: Netflix had a licensing deal that generated multiple blockbusters per year and has so far NOT shown the same ability to replace this content at the same rate. As a result, they will have to spend much more to achieve the same level of viewership.
That’s inefficiency at its finest.
If you have $15 billion to spend, making 10 blockbusters is more efficient than making 5. Making 20 is even better. Netflix—and really their development teams led by Ted Sarandos—hasn’t show the ability to perform better in Hollywood than their peers. Indeed, this is why they’ve had to spend so much on content in the first place, and that’s what is driving both the cash flow losses and that’s what is driving the price increases and that’s driving the customer losses.
So maybe losing all that licensed content will matter.