(This is an “Appendix” to a multi-part series answering the question: “Who will win the battle to make the next Game of Thrones?” Previous articles are here:
Part I: The Introduction and POCD Framework
Appendix: Licensed, Co-Productions and Wholly-Owned Television Shows…Explained!
Appendix: TV Series Business Models…Explained! Part 1
Appendix: TV Series Business Models…Explained Part 2
Appendix: Subscription Video Economics…Explained Part 1)
Today, we continue our journey explaining how a TV show makes money for its producers. Last week, I started describing the model along with describing costs and the first set of revenue. Today, I continue explaining the revenue piece, touch on why I didn’t build this model for the upcoming Star Wars series, describe some of the fees involved, and provide some final thoughts. Finally, I link to my two main sources.
The bulk of the value of a TV series—especially non-hits or mega-hits—comes in the first window. Especially when you talk about value as “money”. For 90% of TV series, the initial license fee usually represents most of the money a show will ever make. But if a TV series takes off, then the other windows will have much more value. And let’s start with one of the newest of those windows.
TV came to home entertainment late, and may be out of it soon. This window couldn’t take off like movies in the 1990s because the number of VHS tapes required was unwieldy. When DVDs made entire seasons the same size as a VHS, right as the quality of top series drastically improved (I’m thinking of The Sopranos here), then the TV home entertainment market took off. With the rise of digital you could buy series without any space at all. (This is a great example of technology enabling a business model.)
Yet, as soon as digital sales started, they ended, didn’t they? So the lifespan of this window was fairly small. I mean, why buy or rent a TV show episode if Netflix will buy the rights and stream them all for you? Still, I’m including it in the model because home entertainment is still something, but more importantly, it was REALLY a thing for Game of Thrones. Franchises that inspire super-fans—which is a short list—can still generate home entertainment sales. At the height, these can be in the millions of units sold per season, at least when GoT started. That can add up and really offset production costs.
So including the first-run licensing revenue, here’s our model so far.
Licensing – Renting Your Show to Other People
I don’t like the word licensing in TV. It sounds too close to selling either naming rights or toy rights. But it is used constantly in television. Instead, I’d call it “TV renting” because that’s really what you’re doing: renting the rights of your TV show out to various players. Still, licensing it is.
For all the sturm und drang about streaming video, in a lot of ways, they were just another window for licensing. Instead of counting the number of runs and paying broadcast residuals, you sold them for unlimited runs to a streaming network, and because it was digital (and not covered by guild agreements), the studios also had to pay less in backend compared to syndication.
The variations in licensing come down to who you licensed to, when they get that license and where they are. Typically, production companies expend the most energy on the first, domestic window and everything else is a bonus. To summarize:
International Sales – The US TV model built up around selling to American domestic broadcast. Once you did that well, you could sell the rights of a TV show to the international market. That’s the big form of potential revenue. However, the shows that fit international market tended to be broad comedies, crime procedurals or genre series. (Many prestige shows don’t travel at all.) This has also blended in with digital sales as streamers can now buy out global rights instead of just US versus international.
Syndication – Again, another legacy of US TV model, syndication was selling a broadcast show—and sometimes a cable show—to TV broadcast network groups to run in off-primetime hours. This still goes on for shows like Friends to Mom to Law and Order: TBD. Since this was done for whole of series runs, often, this could be worth literally billions for a handful of shows. (Again, the kings are Friends, Seinfeld, Cheers, Simpsons, Everybody Loves Raymond, Two and a Half Men and other broad comedies.)
Second Window Licensing – I’m using this to refer to cable channels getting in on the game. You could take your broadcast series, and instead of selling them in syndication, sell them to cable channels in deals that looked roughly similar, but had less impact on residuals and participations. And premium channels could get in on the game too (HBO sold Sex and the City, The Sopranos and Entourage). Recently, even streaming series (Bojack Horseman on Comedy Central) have been sold into syndication. Really, what you need to know is that a lot of TV producers made money by selling their shows to a second channel after the first window ended. And then…
Digital Sales – Digital sales can happen in the first, second or library windows. And many times it doesn’t even conflict with the syndication window above (unless the streamer pays for exclusivity on that window too) which is why Friends is on Netflix and TBS right now. (Friends is such a mega-hit in the TV business I use it a lot.) Since so much money is involved in digital, sometimes the streamers buy out global rights simultaneously, which is a change in business driven by the rise of global streamers.
This gives us four new revenue lines, the three categories above plus “library” which is the placeholder value for time periods after a series is fresh, say five years from the last season and older.
Tax credits are a great euphemism. We could just call them, “government bribes to movie studios”. What else do you call a straight cash payment to a business to come to your state or country? It’s the utter opposite of the “free market”, and states of all political stripes take advantage of it. (*Cough* Georgia *Cough*)
I didn’t calculate this explicitly for films because at the size I was working—hundreds of millions—it would wash out in the production budget. (Though lots of big budget films, including Marvel films take advantage of these credits.) I’ve also seen these accounted for as “negative COGS” as opposed to revenue, if that makes sense from an accounting perspective. (It isn’t money you make, but money that offsets your costs.)
That said, for TV shows in the low millions of dollars range, a couple million dollars can really hit the bottom line. Some TV producers—I’ve heard—have the goal to breakeven on everything else, and their profit is the tax credit. Starting research for this series, I found this Variety article where Penny Dreadful managed to convince California to pay them $25 million. That’s a hefty bribe, er, paycheck.
Smaller Pieces – Merchandise, Product Placement
For 98% of shows (my assumption) licensed merchandise will never be a thing. People just didn’t buy NYPD Blue shirts or Cheers hats. Yet, that has changed as nerds have taken over the world. I own not one but two Game of Thrones shirts. Breaking Bad—another megahit—had two iconic shirts in both the Heisenberg and Pollos Hermanos t-shirts. So I’ll include this line item because if any of my fantasy series becomes a “megahit”, they’ll make some money off of that.
As for product placement, this a real line item, but I only bring it up because I knew someone would mention it. Except for big reality series—think Top Chef—product placement isn’t really a major driver revenue. It helps in the way that it offsets production costs, but I won’t be included it in this model, but I thought about it.
Why fees? And whose fees?
Let’s start with the first question: because Hollywood will always find a way to lay on a fee for the work it was doing anyways. This applies to television production as well, but not always as blatantly as with feature film distribution, where in my experience studios tend to use both distribution and marketing fees all the time. In TV, if the networks isn’t the producer, they usually pay a flat fee, so “distribution fees” don’t come up. But yeah, if the studio is doing a lot of syndication and international licensing, they’ll charge distribution fees.
(Also, I’ve heard that the streamers charge distribution fees on their initial fees, making their 130% initial fee less by 30%, which is just classic Hollywood nonsense accounting of making something seem artificially higher than it is.)
Oh, and package fees.
We all know what these are, but I don’t consider agents “talent” so these go in the fees portion. Package fees are the fees collected by agents for putting together a package of talent of a TV show. (And if you’re reading this in the future, the WGA went to war with the talent agencies as this article was being written, hence why everyone knows about them.) Since they are taken out before talent gets paid, this makes it even less like talent participation.
This is the final piece of the model. Basically, the key talent has a claim on some proportion of the profits of the TV show once it breaks even. This is probably the piece I’m looking forward to calculating the least. It can have extremely wide variance. I’m sure nearly every major character on Game of Thrones has back end participation; I have no idea what each individual agents was able to negotiate. How could you? But I have to factor it in.
So putting in the fees, talent participation, and room for residual payments and you get a model that looks like this, for a traditional “first run licensed” series:
The key is that for the network, studio and talent, I’m just going to subtract and add different pieces to see who gets what. You may also notice I added in some “network” specific costs and revenues. Those are complicated enough that I’ll explain them in my next article. But first…
Some Other Thoughts on and Caveats to My TV Model
Thought 1: Broadcast vs Cable vs Premium vs Streaming
Initially, I debated if I was going to build a model for each of these different platforms/types of network. But I didn’t. Instead, I left room for “first window revenue”. When I write about the different platforms, I’ll clarify their different models, but for now what matters is that each “network” has a “revenue” amount it (hopefully) generated greater than the license fee it paid for the content.
Thought 2: Each show needs its own model
I wish I could build the smooth waterfall of profits a la movies, but with TV shows they each have individualized distribution models. Some shows go into syndication; some don’t. Some can sell merch; most don’t. Some have multiple windows; many have just one streamer. So each show needs its own model.
This is also why I didn’t build this complicated model for the Star Wars series. At the time of that deal, Disney didn’t really know if it was going to launch its own streaming platform, and probably didn’t think it would. (It likely planned to sell Star Wars series to the streamers for huge fees.) To model this, I just used an equivalent of a “imputed license fee” for each TV series. It was just easier than building this model for each potential Lucasfilm series.
Caveat 1: This model is wrong
Inevitably. I’m trying to create a “platonic ideal” TV business model, if you will, for every type of scripted TV show. Individual studios will have their own models, usually bespoke for every TV show, and/or many additional budget lines. This is my shell model that in general tells the story. (Also, my model shows in general how you add and subtract lines.) There’s no model I could build that could capture all other models perfectly, though if you want to send me a model, I’m happy to update my priors if I missed something.
Caveat 2: Scripted series only
Speaking of differences, reality TV in its many incarnations is a different ball game. Especially since they are usually not union shows. These models are for half-hour and one-hour scripted series (dramas and comedies) only.
Caveat 3: Oh, technically this model is per season
I should have mentioned that earlier. Technically, you build one of these for each season, paying our participations and residuals separately. Which would mean for a show like Game of Thrones, you need eight. Since my goal is to use this for a series, I’m going to factor that in.
Caveat 4: US-only models
Good qualification here: if you’re making TV shows or films in India or Japan or Korea or Nigeria—all booming film communities—my models do not apply nearly as well. These are based on the US system, especially the talent participation models, focus on international sales and union residuals.
Like my film models, I used two main sources for reference, in addition to building/learning about similar models in business school off case studies and lectures and my own personal experience in a streaming studio. (The latter of which meant I reviewed and built these models.)
But there are two great sources if you want to learn more about film and TV accounting. First, Harold L. Vogel’s financial tome Entertainment Industry Economics. Which was my gold standard text book. If you want to get smarter on the entertainment biz, just buy a copy of this. I “acquire” my best ideas from this book. Also, Jeffrey C. Ulin’s The Business of Media Distribution doesn’t have example TV models, but a good history of the TV biz and some of its economics too.