The Los Angeles region, and the entire basketball universe, is reeling from the death of Kobe Bryant, the legendary Lakers basketball player. If you’re looking for the “Hollywood” connection, I have two. First, the Lakers and “showtime” basketball have always been an influential part of the entertainment ecosystem in Los Angeles. A place to go to see and be seen. Second, Kobe was an emerging film producer who won an Oscar. His contribution to his passion for film was tragically cut short.
As a long time Lakers fan–read here for some insight on this–this death is shocking and hurts.
Most Important Story of the Week – Facebook Watch Decreases Investment on Scripted Originals
This news is two-fold for Facebook Watch. First, two big series–Limetown and Sorry For Your Loss–were not renewed for subsequent seasons by Facebook. Still, cancellations happen. When you pair that news with reporting from Deadline that Facebook is generally pulling back from scripted original content, well you have a new story.
Mostly, though, this story seemed to pass by in the night. But it’s the perfect story for my column because the significant doesn’t seem to match the coverage.
So let’s try to explain why Facebook may be pulling back on scripted originals. And we have to start with the fact that Facebook is a tech behemoth. Facebook resembles the cash rich fellow M-GAFA titans (Microsoft, Google, Apple, and Amazon) that throw off billions in free cash each year. Really, companies minting free cash have three options to do with it:
Option 1: Give it back to shareholders.
Option 2: Invest it in new businesses.
Option 3: Light it on fire.
Well, as Matt Levine would note, Option 3 is securities fraud so don’t do that. Of course, we could just change it to…
Option 1: Give it back to shareholders.
Option 2: Invest it in new businesses.
Option 3: Enter the original content business!
They’re the same thing anyways. Companies come in with grand ambitions, realize the cash flows in don’t match the cash flows out, and they leave the originals business (or dial back their investment). Facebook follows on the heels of Microsoft and Youtube in this regard. Heck, even MoviePass had started making original content at some point.
The key is how the original content supports the core business model and value proposition. With that in mind, let’s explore why Facebook Watch is leaving the original scripted business, floating some theories, discarding others and looking for lessons for other entertainment and tech companies. Since I’m not a big believer in single causes, I’ll proportion my judgement out too.
Theory 1: Ad-supported video just can’t scripted content.
If this theory were true, woe be to the giant cable company launching a new ad-supported business!
Let’s make the best case for this take. The working theory is that folks just don’t want to watch advertising anymore, so they just can’t get behind a video service like Facebook Watch that is only supported by ads. With the launch of Peacock, I saw this hot take a bit on social media.
Of the theories, I’d give this the least likelihood of being true. From AVOD to FAST to combos (Hulu, Peacock, etc), advertising is alive and well in entertainment. Despite what customers say about hating advertising, they end up putting up with quite a bit. It’s not like Youtube is struggling with viewership, is it?
Judgement: 0% responsible.
Theory 2: Scripted content is too expensive (or doesn’t have the ROI).
If this theory is true, woe be to the traditional studios getting into the scripted TV originals game.
This is the flip side of the above theory. It’s not about the monetization (ads versus subscriptions) but about the costs of goods sold (the cost to make and market content). What I like about this theory is, if you’re honestly looking at monetization, it’s not like entertainment has seen booming revenue in the US. If anything, folks pay about what they always have.
So what’s fueling the boom in original content? Deficit financing and super high earnings multiples.
Worse, deficits are financing a boom in production costs as everyone is fighting over the same relatively limited supple (top end talent) so paying increasingly more. Consider this: in 2004, ABC spent $5 million per hour on it’s Lost pilot, up to that point the historical highpoint. Most dramas cost in the low seven figures. Now, word on the street is that Lord of the Rings, The Falcon and Winter Soldier and Game of Thrones could cost 5 times that amount. Meanwhile, each of the streamers, I’d estimate, would have double digit shows that cost $10 million plus. Did revenues increase five times over the last fifteen years? Nope.
Thus, Facebook may just be on the cutting edge–with Youtube–of realizing that scripted originals aren’t the golden goose Netflix and Amazon make them out to be. It’s not that they can’t make some money on them, just not nearly enough to support the skyrocketing budgets.
Judgement: 25% responsible.
Theory 3: Facebook Watch needed more library content.
If this theory is true, woe be to the giant device company that launched a streaming platform sans library.
The best case for this is that after you come to watch a prestige original, you need to find something else to occupy your time until the next original comes. That’s library content. While I josh on Netflix for lots of things, I do absolutely believe that Reed Hastings is right when he says he’s in a battle for folks’ time. But I’d rephrase it slightly in that you’re also battling for space in people’s mental headspace. When they decide to watch TV, they then pick a service to watch. Library content’s purpose is to keep permanent space in people’s mental headspace. Having loads of library content makes it more likely that you’re folks’ first choice to find something.
The problem is Facebook Watch doesn’t have this. Fellow ad-supported titan Youtube clearly does. It’s purpose was videos first and foremost, so there is always something else to watch. Netflix has it. Even Amazon has it. Facebook has socially generated videos, which aren’t the same ballpark as scripted video.
Judgement: 20% responsible.
Theory 4: Social video can’t support scripted content.
If this theory is true, woe be to the social platforms producing tons of originals.
Here’s my favorite theory, because it directly ties to the customer value proposition and behavior. To start, you have to understand that most TV viewing still happens on living room devices. Meaning folks come home, flip on the big TV and sit on the couch. This even applies to kids!
The challenge is that Facebook has really, really struggled to get their TV app into living room TVs. Not that they haven’t tried by advertising with the Muppets.
I could go further. Scrolling through a Facebook feed is an inherently personal activity. (Which is how I think most people feel about their social feeds.) It’s not necessarily private–if you’re in a relationship, hopefully you would let your spouse see your feed–but it isn’t something you share either. For a huge majority of TV viewers, TV is social. Again, the TV is the centerpiece of your living room. That’s why, when I laid out my “map” of streaming video, I split “social video” from “entertainment”. They serve different purposes.
As a result, Facebook’s original content efforts were doomed from the start, as long as they tried to combine original content with user’s videos. (The equivocation? I can’t decide if Youtube is “social TV” or “ad-supported”. It honestly acts like both.)
Judgement: 40% responsible.
Theory 5: The idiosyncratic reasons. Or “No one watches video on Facebook watch.”
If this theory is true, woe be to Facebook.
The theory here is that there are always idiosyncratic problems with businesses and we can’t always draw cross industry trends. With Facebook, that could be more true than anything else!
To start, Facebook has a privacy and customer trust issue. So the issue isn’t just that social video has problems, but everyone has a problem with Facebook. And PR stumble after PR stumble after privacy violation after privacy violation means no one is eager to buy new Facebook devices or download new applications to follow you across the internet and suck up all your data. Throw on top the (over-hyped) idea that the kids hate Facebook and you have a perfect storm to crush use on your site.
Combine this with sub-optimal UX for their application, the struggle to woo creators in this climate and a lack of ability to market content with the studios, and you have the perfect storm that they just couldn’t overcome.
Judgement: 15% responsible.
The future? Well, let’s not go too far.
Facebook isn’t leaving video. And given that they still have billions to spend on anything they want (including lighting it on fire), they’ll still have a place in original content. They just signed a deal with Steve Harvey for his show, it’s not like their video ambitions have been quelled that much. They just aren’t investing in pricey scripted originals.
Still! The perception that Netflix and Amazon have crushed it in scripted originals has influenced the boom in “Peak TV” more than any other phenomenon. The failures of Google, Microsoft and now Facebook should inform that analysis just as much.
Other Contenders for Most Important
Disney+ Has All The Dramas and Drama
Disney+ has had showrunners leave, writers get fired, and pilots moved to the Disney Channel in the last week. Which is par for the course for any channel or streaming service, but Disney is the buzziest streaming platform after Netflix and buzziest company for super-engaged fans, so every stumble gets picked over.
In this case, though, I think the problems are real. And to be clear, if any execs at Disney+ are reading, this isn’t your fault! Instead, it’s tough to make a brand new streaming service from scratch. Amazon and Netflix went through the same thing, with show cancellations or bad show premieres or showrunner chaos. Apple TV+ is going through it now. It should have been expected.
But I think Disney though they’d ramp up easier than this. First, Disney+ probably thought the entire Disney production system would help provide all the content they’d need. The problem with this theory is that movie studio timelines aren’t the same as TV timelines. And even if you have all this content, setting up a new network is still it’s own challenge. (Even ABC never cracked how to get the rest of the Disney apparatus to make them new content.)
Second–and really this is the issue in particular–the content for Disney+ was wholly new, even if they didn’t think it was. They’ve made TV; they’ve made feature films; but they hadn’t made TV series with the production budgets of feature films.
More to come on this, I promise. (Meaning by the end of the year!)
Netflix is releasing some DVDs of original movies via the Criterion Collection
We should monitor how much this actually impacts Netflix’s business model, u. After a certain point, if someone isn’t subscribed to your service, they never will be. Alternatively, if someone really wants to buy a DVD, just let them.
Universal and Warner Bros Make a Home Entertainment Joint Venture
This feels a bit like Bill Murray in Ghostbusters, “Dogs sleeping with cats” and what not. If the studios are selling DVDs together, well then you know business is tough. The main driver, based on my experience, is that home entertainment requires a lot of manpower to pull off. More than you’d guess. By consolidating, it boosts the ROI on the declining sales (which are still more than you’d guess).
Data of the Week – Netflix Earnings Report
I have to admit: I didn’t see this datecdote news coming. As we’re all speculating about what shows Netflix may dribble out information on, no one predicted that Netflix would change the rules of their game!
But we should have! Especially after Netflix changed the rules for their the year end top ten lists. Besides just ranting about how little Netflix is telling us, I’d like to tear down one number Netflix told us.
“35% on average”.
That 35% is the gap between the old metric (70% completion rate) and the new one (2 minutes of viewing). Meaning the 2 minutes of viewing are “on average” 35% higher than the 70% completion rate. Rephrased, “the 70% completion metric” is about 74% of the number of folks who watch 2 minutes. The problem, frankly, is someone like me making a table like this and putting it on Twitter:
But why? That’s a great table that helps inform everyone on how Netflix is changing!
Well, because I hate, despise and loathe averages, and you should too!
To understand data, at least superficially, I think you need at least three pieces. Literally not having one of the three means you are more likely to be misled then led on a topic. The first piece is the average. It gives you the center. If we’re asking about heights in the US, then the “average” is 5 foot 9.
The second piece is the “standard deviation”. That basically tells you how far apart or close a data set is. For height, for example, this is super useful because it lets you know that, for example, there aren’t a lot of 8 foot tall folks running around the US. If the standard deviation was two feet, there would be! (It’s around 4 inches.) Without the standard deviation, you know hardly anything.
The third piece, that isn’t captured by standard deviation or average is the shape of the distribution. And I wrote a whole article about this here. For instance, if you take male or female height, that’s a normally distributed shape. But if you combine them, your shape actually has two humps in it. Interesting!
For Netflix, the point is we don’t know anything about their standard deviation or shape of the data. Do any of their films have 50% difference between the data points? Probably! A 10% standard deviation is fairly reasonable. (Meaning some 2 minute viewing stats are 50% higher than the 70% completion.) You could update my table from above that the numbers being presented now could be incredibly misleading.
Further, for something like the conversion of 70% to 2 minutes viewing, the shape actually has a much longer tail “to the right” if you will because there is a floor of 100% where everyone who watches 2 minutes finishes the show. (In other words, you can’t watch 2 minutes but not watch 70%.) But we have no idea exactly how that lean looks like.
In other words, this data could still be way, way off!