The Real Economics of Games

Every so often, another story surfaces on YouTube and in the blogosphere about how expensive games are to make. And, clearly, games, especially AAA games, cost more than ever. When the modern game biz emerged in the early 90’s, the production cost ratio between what’s now called AAA and a top shelf, big budget film was roughly 1:10, that is, a big film cost about ten times as much to make as a big game. Now that ratio is right around 1:1 and it’s pretty clear that GTA VI could flip the ratio to something like 2:1 or even higher. So what happened? How did games get so expensive? And does it even make sense that they are? Before we try to answer, let’s take a look at where the money actually goes in a game project.

Marketing

When games came exclusively in physical media - plastic cartridges, floppy disks or CD-ROMs, big game marketing was completely focused on day one of release. This was for very good reasons. In those days, a game did roughly 90% of its business in the first ninety days of its existence. A lot of that was driven by retail. Stores live on very thin margins. They have to be super efficient about their use of shelf space. If something doesn’t move quickly it gets replaced with something that does.

There’s a story the science fiction writer, Halan Ellison, used to tell about waiting for a flight at O’Hare in Chicago and noticing that there was this guy making the rounds of those little carousels filled with currently trendy paperbacks. He was refilling them with books. But Ellison noticed something strange. He’d been waiting long enough that he saw the guy at the same carousel twice, about two hours between visits. Any books that hadn’t sold at least one copy got pulled and replaced with books that had sold. When Ellison asked the guy about it, he confirmed that, yeah, a paperback had exactly two hours to sell at least one copy or it was coming off the shelves for good. Basically, the same thing was true in every game software store in the world. The time limit wasn’t two hours, but a high profile game had about a week to start selling. After that, the stores would pull it unless the publishers paid them to keep it on the shelf.

So what marketers learned to do was something called ‘surge’ marketing. The same style of marketing used for decades by the movie industry. Build as much hype as possible to line up with launch in hopes of driving a bunch of units off the shelves as quickly as they could. The thing about surge marketing is that it’s high risk and high cost. High risk because any shift in the release date can blow up expensive media buys, essentially just throwing money on a bonfire. High cost because, for a very long time, games were in lockstep with the last three months of the year, when everybody is screaming for attention and the cost of ad space is ten times higher than it is during the summer.

Digital distribution and social media changed a lot of the terms in the marketing equation, but it also drove the noise level in the games market a couple of orders of magnitude higher. Worse, a lot of ad buys aren’t pre-negotiated anymore. Marketers have to pay whatever the going rate is on the day the ad gets served. What that does is make enormous amounts of money for Google, Meta and Amazon. It also drives the cost of marketing higher as a percentage than it’s ever been. In 2000, the rule of thumb for marketing allocations as a percentage of projected Net Revenue was 15% across the entire industry. Now, it’s 25% but, in AAA, it’s 50%. Sometimes even higher. Yes… you heard that right. Sometimes selling the game costs more than making it. Which clearly means they’re making the wrong product, but that’s a conversation for another day.

Manufacturing, Distribution and the Cost of Sales

In 1992, the average Sega Genesis cartridge cost about $5 to make and another $1 to get it to the US. The retail markup on that cart was 2x, so, a $20 game had to wholesale for about $10, which meant Sega would get a $4 gross margin on every cartridge they sold. IF they sold it. You can see that, back in the day, with physical goods, the inventory risk was enormous, especially with ROM cartridges that had to be built on one side of the Pacific and shipped to the other. Oh, and just for fun, they had to show up at exactly the right time for launch.

Then Valve blew up the entire industry with Steam and, suddenly, digital distribution was the way games got to customers. In reality, it wasn’t that sudden. It took almost fifteen years for digital to genuinely dominate, but that’s still pretty fast. Digital took away that enormous physical goods risk and left more money for the publishers right? Well… maybe. But not really. First, there’s the platform fee which, for Steam and the consoles, is 30%. Still, that $20 game now costs you basically nothing to produce and you get $14 in gross margin instead of $4. Cool… except there is a raft of new costs in the mix.

It turns out that what was true about those carousels at O’Hare is just as true with digital storefronts. You want placement? Give me money. You want spot ads. Sure. No problem. Oh… but you want to associate those ads with keywords your competition is using? That’ll cost extra. The current bid for the top of the main page is $8 per click. Oh… it didn’t convert to a sale? So sorry. Give me another $8 and you can try again. That extra $10 in margin disappears very quickly for any but the highest profile, most popular games.

Royalties

As the noise level in entertainment has climbed, it’s become increasingly tempting to the big publishers to try to cross-collateralize the properties they own. If you’re Warner or Disney, you have every reason to do this, even if consumers are understandably wary of the result. Probably the one sector where renting IP has clearly turned out to be worth doing is in professional sports. This isn’t just about American football and Soccer. It’s basketball. It’s baseball. It’s racing. It’s extreme sports. The whole gamut. But what most people don’t realize is that the cost of renting the official branding of those sports ranges from around 15% to as much as 30%. So, now, you’re burning 50% of your net revenue on marketing and another 25% on a license. Suddenly that $20 looks like 20 x .7 x .25 = $3.50. So here we are 30+ years after the bad old days of ROMs. To just break even against inflation, we’d need to see at least $10 after the costs of royalties, distribution and sales. But, instead, we’re seeing about a third of that.

Development and Support

Then, of course there’s the piece of the puzzle everyone loves to talk about. Laura Fryer, arguably the finest producer XBox ever had, pointed out recently in a post on her YouTube channel that the original Halo was built by around 100 people. Not long after she posted, she corrected that number when Marty O’Donnell and a few others from the original team told her that, in reality, the core team was only about 48 strong. The last Halo offering, which was almost universally reviled for being stupidly light on features, had a credits list containing more than 2000 people. To be fair, at least half of them were contractors, but that doesn’t necessarily mean they’re cheaper. In fact, the opposite is often true.

In North America, the average loaded cost of internal game developer teams is around $15K per employee per month. The average for contractors is lower, but that’s largely because those jobs tend to be lower paid gigs like QA. For easy math, we can use $10K per month and be pretty close. Let’s take a hypothetical AAA project and assume we’re looking at an average of a couple of hundred people working on the game for three years. That’s an internal cost of $104M. Let’s add on another hundred contractors at a cost of $36M. So here we are. $140M in direct development cost, assuming the game isn’t delayed and only takes three years to make which, as we’ve seen numerous times of late, is far from a sure thing.

But then the curse of digital distribution hits you square in the bank account. In the 90’s, whatever you shipped is what got sold. There was never any expectation that you’d have to maintain the game after release. Now? It’s pro forma. What that means is shipping doesn’t shut down your costs. Sure, if your game is a one and done, if you’re only doing bug fixes and maintenance then maybe you can trim that team of 200 down to 50 and that list of contractors down to 20. But you’ll be paying for them for years. Assuming no DLC, assuming no content updates, assuming a lifespan of only five years, you’re still looking at another $5M to $10M in dev cost. So now you’re at $150M and that’s for a game that, in the world of AAA, would be considered near the bottom of the cost range.

The Bottom Line

Ok… let’s see where this all nets out. At modern game publishers, the min bar for an ‘adequate’ game is 2x their combined dev and marketing costs. A ‘success’ is double that. In our little example project that means we’re looking at having to move 44M units of the game to be considered a ‘success’. In 2025 a tiny handful of games managed numbers even close to that threshold. Many of them didn’t even make it to breakeven.

And there, in a nutshell, is the problem with contemporary AAA development. This happened with filmmaking 75 years ago, when the cost of making movies in Hollywood, with internal staff, got so high that the studios simply couldn’t make money. What came out of that was an evolution of the business that favored hybrid productions relying mainly on contractors, offshoring and a host of the kinds of changes we’re seeing in games right now. It also led to an explosion of clever, smaller outfits that operated independently and found their own ways to reach an audience. The same thing is happening in games. Indie doesn’t mean toy-like anymore. It means not bound to the old models that the big publishers have such a difficult time letting go.

At TLG we’re evolving our own models that rely on small, highly flexible teams and games scoped to the gamers we want to reach, but with a relentless focus on quality. We think that will be a winning combination in the new landscape of games that will emerge from the ashes of the old studios.