(For the last few weeks, I’ve been debuting a series of articles answering a question posed to me by The Ankler’s Richard Rushfield: Will The Irishman Make Any Money? It’s a great question because it gets as so many of the challenges of the business of streaming video. Read the rest here, here, here and here.)
On Monday, I explained the grand plan of Richard Rushfield and I plan to estimate the value of future Netflix films, starting with The Irishman, out earlier this month in limited theatrical release, but coming to the world’s biggest streamer next Wednesday. For traditionally released theatrical films with normal second windows, we have a robust model we can employ.
What about streaming only? Well, that’s where it gets tricky.
A lot of folks do some back of the envelope math for this, and this can be a useful way to look at the problem of valuing streaming. Take Richard’s approach from a few weeks, back looking at Disney films that had been in theaters after 3 weeks (when Netflix pulls them from streaming). Of all the films, Disney earned roughly $310 million after 3 weeks of theatrical distribution. That’s the equivalent, Richard noted, of 4.3 million customers subscribing for 12 months on Disney+. If that number seems big, it should be, which shows the value of theatrical releases for studios.
Could we just take that approach and just apply it to The Irishman? Unfortunately, it has some flaws, mainly double counting subscribers. We need a different method to employ in “The Great Irishman Challenge”. Unlike traditionally released films, in steaming there are a few ways to value a given title’s performance, and each method has its own pros and cons, ranging from crippling to merely difficult to over come.
Which I’ll (re)explain today, along with describing which ones are fine, which ones are incorrect, and which ones I prefer. At the end, I’ll explain which one we’re using.
Four Ways to Value Streaming Video and One Way NOT To
I’ve previously valued streaming video in two articles. First, back in January, I looked at Disney’s decision to keep theatrical windows for Star Wars films. Second, back in May, I explained streaming video models in order to put a value to HBO’s Game of Thrones. Today’s article will explain all the models from those two articles and add a new method I figured out how to calculate last month. (I had employed this method at a previous employer, but needed a key piece of data, as you’ll see.)
For each of these methods, I’m going to assume that Netflix had a feature film that was seen by 40 million subscribers in the first 28 days, divided evenly between the US and international. The film cost $115 to make and Netflix spent $50 million to market it. As for box office? Let’s say it had $120 million in the US and $80 million globally.
Sub-Optimal Method #1: Multiply Customers by Month by Price
This is the most common method of “back of the envelope” valuations I’ve seen for Netflix films. Usually, you hear folks do a version of this on podcasts, and I’ve seen it for Lord of The Rings on Amazon a few different times. Also, you could do “customer years” the way Richard did above. Here’s how the model for this approach would look:
The problems? First, this is one-quarter of Netflix’s subscriber base attributed to a film in one month, which would probably be one-third of their active users. In other words, if Netflix had two other properties getting similar ratings, then every other film released that month would “financially” be a net loser. Second, this approach doesn’t account for “customer lifetime value”, which is really the better approach to valuing customers, versus the one month or 12 month view. Third, this approach doesn’t distinguish between films and TV series total hours of viewing (because it is just subscribers) so it’s tilted towards films, which are shorter and easier to finish.
Still, you can use this to ballpark how long a film would need to make its money back. It’s just sub-optimal because of double counting.
Sub-Optimal Method #2: Attribute Customers by Usage
One of the interesting ways to look at content is to think about what percentage of viewership a title makes up of all the viewership on your platform. If 10% of all hours watched are your platform are Friends, that has to mean something. The challenge is knowing how much people actually watch on Netflix. Netflix has helpfully told us twice (twice!) that they stream about 100 million hours daily in the United States. That means I can calculate potential usage of a TV series or film! That would look like this:
The problems? First, getting the usage data is really tough. For films, we’d have a pretty easy time, but for TV series, we really don’t know how many people watch how many episodes. And getting usage numbers for Amazon or Hulu may be nearly impossible. Second, it also doesn’t factor in “customer lifetime value”. Third, it over-weights TV library content because there is just a lot more to watch, hence it’s “usage” is much higher, if you can get that viewership data.
Still, you can use this to compare the usage of various shows and movies. It’s just sub-optimal because it’s tilted to shows with much longer runs.
The Bad Method: Multiplying Subscribers by Customer Lifetime Value
I’ve seen this mentioned in places, though running them down is tough on Twitter. So this may be a strawman, but it’s worth pointing out in case it hits anyone to do. Twice, I’ve criticized valuation methods because they don’t use customer lifetime value. You may be tempted, then to just take the number of subscribers and multiply by CLV instead. Like this: