Monday, September 13, 2010

Some Accounting on the Cost of Making Games

Sorry about the delay in updating. PAX was fantastic, but exhausting. I was talking to another Hothead at PAX about the cost of the show and the return it actually has. While exhibiting at an event like PAX is certainly expensive, my claim is that it's probably the best investment you can get for your marketing dollar. And getting that attention from a community is very important, especially for a smaller developer.

I like Frictional, and not just because they're one of the few studios left making interesting horror games. For those not aware, they made the Penumbra series of games and just released Amnesia: The Dark Descent. They're also pretty transparent about their process, including a very interesting post Kieron linked in this week's Sunday Papers. It's from January of this year and discusses how much it costs Frictional to make a game. Specifically, how much development $100,000 gets them. Spoiler: it will cover about six month of development time for Frictional's five employees (and contractors).

Games, and software in general, are interesting because there is almost no materials cost for the products created. Maybe a little bit for hardware, but the cost of development is almost entirely labour and overhead. The cost comes from bodies on the project. Which means a project being even a little late can quickly send a viable budget into "maybe we'll break even" territory. Aside from actually producing a quality game, keeping your schedule will determine whether or not a game is profitable.

As an aside, one interesting thing to note from the Frictional post is that it actually costs the company about $2200/month to pay an employee $1500/month. That extra 30% comes from government taxes and vacation pay. Sweden is a pretty socialist country, so I assume that figure is higher than in the US, but it's probably not far off from here in Canada (and with our/Sweden's awesome socialized health care, the employer doesn't have to pay for medical insurance for employees, so the end cost in the US might be about the same anyway). I didn't really consider this before, but I suppose that's part of the reason why hiring contractors (sometimes repeatedly and contiguously, creating perma-contractors) is an increasingly common practice in the industry.

Here's the interesting part the Frictional post didn't really discuss. The cost of making Amnesia is only half of the equation. The other half is how many copies does Frictional need to sell to make that money back? Frictional estimates that $100,000 gets them 1/6 of the way to completion, so Amnesia cost around $600,000 total. The game sells for $20 on Steam, so after Valve's cut (30%), it's $14 per sale for Frictional. If you buy it directly from Frictional's website, they keep all $20. It will take about 43,000 sales just to cover the cost of development. If Frictional wants a comfortable buffer to cover the cost of developing whatever's next, they'll probably need twice that. Just guessing, 90,000 copies seems like a solid success.

That's both enlivening and intimidating. A digital distribution-only indie horror game with next to no marketing could be hard pressed to hit almost 100,000 units. But it almost means if just 4 of every 100 people that bought Dead Space (2.2 million worldwide) also buys Amnesia, Frictional will be doing just fine. That makes it seem like a more attainable goal.

It's also god damn hard to make a game, even as an indie, for less than a million dollars. Another person here, another few months there and a budget can easily hit seven figures. Most people seem unaware of just how expensive even seemingly small independent games are. Frictional's post mentions a conversation with a friend outside the industry who guessed their budget might be $25,000. I imagine even many serious gamers wouldn't be much more accurate. Even an order of magnitude increase to that guess is less than half the actual budget.

To be clear, I don't care about profit for profit's sake. If you're in games to make money, you're probably an idiot, because there are lots of ways to make way more money with way less risk. But profit is important because developers need to eat and pay their rent. Financial success means talent creators can keep doing what they're doing. That's what matters.

The point of looking at this balance sheet is I think it emphasizes just how much influence the community of knowledgeable consumers have over what titles do well. A community getting fired up about some indie title and sending a few hundred more sales their way is actually a big difference. It's a rounding error to EA or Activision, but for Frictional, it's significant.

Part of the reason why I actually like the space I work in so much is that there's room for weirdness. When a big publisher has to move literally millions of copies to just break even, they're not going to take any big risks, creative or otherwise. But if you only need 100,000 sales for a comfortable success, you can make something weird that a lot of people might not like. But if even a small subset do, an indie studio can do just fine.

Every time you recommend some wacky indie title to a friend, you're helping out maybe more than you think. Gamers have far more influence over weird little games than they do tentpole blockbusters that get millions of advertising spend, so wield that influence wisely. Reward developers for doing weird, interesting things. In that spirit (if nothing else that to thank them for the frank discussion of their financials), go check out the demo of Amnesia. If you like it, or know someone who might, spread that word. A little kindness and we might help Frictional hit that comfortable 100,000. I think that would be great for all of us.

Labels: ,


Blogger Darius Kazemi said...

That 30% figure is about right for the US too. Although once you factor in the cost of office space and training time and other things, a good rule of thumb is that employees end up costing about 2X their salary.

September 13, 2010 at 2:30 PM  
Blogger Nels Anderson said...

@Darius Ah ha, thanks. I'm sure that isn't a novel realization for a lot of people, but for someone that doesn't peer into the business side of things that much, it was definitely interesting.

September 14, 2010 at 6:52 AM  
Blogger David Carlton said...

Yeah, 2x is what I've heard as well. And I certainly know people who prefer to work as contractors - if you have a spouse who has health insurance and like working on new stuff frequently, it can be a good match. (Which isn't to say that it doesn't get abused, too; and having health insurance tied to your employment status is barbaric.)

$1500/month, wow. I hope the cost of living where they are is a lot lower than it is where I am. Though, when I was a grad student, I was quite happy on less than that; depends on your context.

Getting 4% of Dead Space's sales sounds attainable, but that sort of thing can also be dangerous, because it can lull you into a false sense of security. I'm sure there are zillions of companies that have had the idea "if I could get 1% of the iPod's sales, I'd be rich, and how hard can that be?", only to discover that the answer is "hard enough to cause you to lose a lot of money". Fortunately, these folks do seem to be doing a pretty good job of getting the word out, based on the amount of buzz I've been hearing, so that's a start, at least.

September 14, 2010 at 10:05 AM  
Blogger bob said...

"If you buy it directly from Frictional's website, they keep all $20."
Actually it looks like they're selling through an e-commerce site that takes at least 12%; better than Steam I'd imagine, but still not all $20 of the sale.
Sadly, according to their blog, they're aren't getting the sales they'd like (and sales have already dropped off), so I'm guessing not even close to 90K. Without money to spend on marketing, getting those kinds of sales numbers is hard.

September 15, 2010 at 5:32 PM  
Blogger Nels Anderson said...

@David Yeah, I wasn't dogging on people who opt to be contract. I'm a little less fond of places keeping someone that wants to be FT on contract literally for years. Granted, those people probably should just not renew their contract and go find somewhere more suitable to work. And yes, privatized health care contingent on your employer is awful (and once again, woooo Canada!).

It's definitely true too that "get tiny % of huge thing" can be real dangerous thinking. I mainly mention that for the purposes of scale, not that I think that alone is good enough reason to make something ;)

@bob Oh really? I pre-ordered months ago and forgot there was an separate e-commerce site. Still, 12% is better than 30% from Steam (which itself is way better than most console margins).

Yeah, I saw their blog post about Amnesia's launch. It sounds like they're near the safe point, so that's good at least. For a zero-marketing horror game, it's hard to imagine blowing the door off. But it sounds like they'll be able to fund their next project, so if that's a metric for success, I'm glad that was achieved. I'll keep banging the drum for Amnesia though, maybe it will have more legs than they expect.

September 15, 2010 at 11:28 PM  
Blogger bob said...

@Nels Anderson:
Yeah, that 12% assumes there are no other fees or costs associated with the hosting or sales. I've never worked on a console title, but I have a vague idea of what the console fees are like for boxed games; are the fees for the console downloads - e.g. XBLA - any more reasonable? I should hope so, given the smaller profit margins on the sorts of games that are sold as downloads.

I had assumed (hoped) that download sales would allow for a big improvement in how long a game would sell. Boxed copies on shelves have traditionally had, what, a month at best to sell all the copies they ever would before being removed to make room for a newer game. (Huge hits being the exception.) Instead of extending that window, online sales appear to have diminished it, instead. Certainly online sales give a "tail" to game sales that they never had before, but if sales drop off exponentially after the first week, subsequent sales aren't going to be significant. Perhaps this only applies to developers without any real marketing resources - people buy the game during the short review period, after which the game loses all visibility (oddly, despite being prominently featured on Steam). Hopefully there's a second wave of sales when word of mouth kicks in. I do find it rather disheartening, though. At a time when development costs for AAA games have become completely unsustainable, I had held out hope that small indie developers would have an easier time staying afloat. Exponentially lower development costs seem to correspond with exponentially lower sales, unfortunately, regardless of game quality.

September 16, 2010 at 11:30 AM  
Blogger Nels Anderson said...

@bob MS has a weird sliding scale based on units sold for XBLA, I forget what Sony's has for PSN but it's also kind of weird. Valve (and Stardock w/ Impulse, etc.) are definitely the most fair with a flat 30% for everyone (as far as I'm aware).

Digital dist is definitely a huge boon for small indies and the boost you get from, say, a weekend sale are a lot more than you'd think. Frictional wrote that whole "how far to do we get for $100,000" piece because I think that's about how much they picked up during on of the Steam sales. It's definitely way better than retail where there's first week and that's basically it.

It's unfortunate that people aren't more open with the sales figures (makes sense, I guess) because I'm really curious about the drop-off rate for game sales. We can sorta compare w/ certain XBLA/PSN titles thanks to leaderboards. I've been tracking that for DeathSpank and it's definitely interesting. The problem is I wish I could compare it against other folks rather than just having to guess.

September 16, 2010 at 10:59 PM  

Post a Comment

Subscribe to Post Comments [Atom]

<< Home