Dude, Where's My Money? Part One: The Science of Steam

This is the first in a series of articles dedicated to an examination of the entire process of publishing PC games, from beginning to end. We won't be getting into development here - that's a whole other kettle of fish.

We'll be talking about Steam a lot, since it's the largest open game distribution platform, but we'll also take a look at third-party distribution and alternative ways to sell games. Before you start trying to find your own answers, let's take a critical look at how the gears of publishing turn nowadays. We'll start with prices, discounts, and how Steam wishlists work.

Let's start in the middle

That's right, the middle. Take your rose-tinted glasses off and arm yourself with a calculator, some basic math, and a bit of patience. We're going to try to figure out how much your game is going to cost.

Before we can answer this question, we need to ask another one: how many sales do you need to break even?

Let's assume you plan to spend $100,000 to develop and market your game, then sell copies for $20 apiece. Here's the easiest way to calculate this: $100,000 / $20 = 12,000 copies until your investment begins to pay off. What, your math says 5,000? You're doing it wrong. Let's find out why.

Let's say you're planning to release your game in North America, Western Europe, Russia, and China. For regions with sizable currency exchange rates, instead of your twenty bucks you'll have to deal with their "average price," since the game will be priced differently there.

For the countries we're interested in, prices are not directly converted from US dollar into the national currency. That's just the way it works there. Steam's equivalent of a $20 price tag for China is Ұ70 (about $10); for Russia it's 465 rubles ($8).

Assuming that 50% of copies sold are purchased in the US ($20 each) and Europe (17.2€, or $20 each), 25% are sold in Russia ($8 each), and 25% in China ($10 each), your average price ends up being $14.50 per unit. The number of copies you need to sell in order to break even just increased to 6,896.

The more sales you get in regions with higher prices (North America and Europe), the closer your average price gets to the one you set initially.

Consider that most large digital distribution platforms offer refunds, and their customers do use them. No game on Steam has ever been refunded zero times. There are always some dissatisfied customers or users who experience technical issues. You have to plan on 2.5-5% of units sold getting returned. Let's be pessimistic and go with 5%. Now you have to sell 7,258 copies to break even.

For now, we won't bother calculating precise taxes (such as VAT and GST). This is a complex issue that largely depends on which country your company is registered in, the customer's country of origin, and any agreements for the prevention of double taxation that might exist between your country and that of the store.

Based on my own experience, a company registered in Cyprus loses the equivalent of 4.5% of its worldwide sales on taxes, bringing our adjusted target to 7,600.

Finally, we have to account for how revenue is split up. The distribution platform usually takes about 30%, so now you need to sell at least 10,875 copies to stay in the black.

And don't forget about any revenue taxes you'll have to pay in your country of residence. If your company hails from Germany, enterprise revenue is taxed at 15.825%. That means that you now have to sell 12,919 copies at full price just to break even.

Final score: Copies = your budget / (average price * distributor's cut * refunds * taxes)

But that's not all! Your average price doesn't stay the same throughout the year. It goes down with discounts, sales, and special promotions. After the first few months of full price sales, the average price for the year will drop by 25-30%.

Discounts and Sales

Like it or lump it, we've got to accept the fact that modern PC game distribution is driven almost entirely by sales. Now let's take a look at Steam.

Steam has two major sales per year (the summer and winter sales), along with the option to participate in Weeklong Deals, which are weekly sales that start every Monday. You can also decide to hold a sale of your own and choose the time, discount, and occasion for the sale.

When scheduling your release, make sure it doesn't coincide with the summer or winter sale. Think carefully about when to launch your first discount. It will cause your full price sales to drop afterwards, and you'll have a hard time bringing them back up to the pre-discount level.

Remember that Steam discounts add up. If you want to release on a Friday, think again - you could end up in the middle of a Weekend Deal with a major publisher, such as Ubisoft selling off a large portion of its back catalog for peanuts. I wouldn't recommend competing for customer attention with AAA games being sold for 70% off.

Then there are special promotions you can negotiate with your personal manager at Steam. These include Daily Deals, Weekend Deals, Midweek Madness, or exclusive sales that happen throughout the year, such as the best games from a particular genre or specific country.

The option to participate in these special promotions is based largely on the luck of the draw. This is one reason to consider partnering with a publisher. If they've been working with Steam for a long time, they'll have a better chance of providing you with these kinds of opportunities.

So those are Weeklong Deals. But what about the Summer Sale? When Steam launches its discounts (other than the ones you set), you gain a lot of visibility for a little while, along with plenty of traffic. Every major sale will give you a bump in revenue and increase your customer base while significantly expanding your game's wishlist presence (as you can see from the graph below).

A lot of people think that a discount of 20% is enough to get the job done. If you ever get to 50%, forget about having as many full price sales as you used to.

Remember that you can only launch a Weeklong Deal once every 60 days. All things considered, you're only going to participate in five or six sales (probably five) a year. So, release window aside, you'll have five or six growth opportunities. Then, after your release window closes, the price of the game will correlate with the average discount spread over the year. We're expecting Flash Sales to return soon, which would allow for sales lasting six, eight, ten, or twelve hours. However, keep in mind that larger discounts also increase the percentage of refunds.

Steam has another sweet tool besides sales: coupons. You can hand out coupons offering a 14-day discount on a game to everyone who owns another game. These coupons provide a discount of 15-25%, and they do a great job if you already have a sizable customer base. You can't distribute these coupons on your own - you'll have to ask your Steam manager to generate them.

Steam managers aren't especially fond of this tool, since issuing the coupons requires some manual labor on the part of Steam's engineers. If your game has an install base of over 200,000, Steam is likely to greenlight it. By the way, this is another reason to work with a publisher - they might already have some releases with a large audience or experience dealing with coupons.

You should also consider Steam bundles - not just bundling your game with DLC, but also with other games, potentially even games released by other developers and publishers. Think about colleagues you'd like to team up and sell games with, and try to discuss the possibility in advance.

Steam Wishlist

We've already calculated that we need to sell at least 13,000 copies of our game at full price in order to break even (hopefully within the first month) before discounts kick in. Let's see how Steam Wishlist can help us. Or, to put it bluntly, let's see how the number of people who have a certain game on their wishlist corresponds to the future sales of that game.

All the companies I know have their own ideas about Steam Wishlist's inner workings and how it converts to sales. I'm going to speak from my own experience and that of colleagues I've discussed the issue with, as well as Steam's own statistics as displayed in the developer toolbox.

Can you find the release date and sales on the chart above (click to make it larger)? Steam provides a decent amount of data about wishlists, but there's only one statistic we really need: the conversion rate of wishlist entries into purchases.

This parameter is constantly in flux. At the time of this writing, Steam's partner tools show that a 5-7.5% conversion rate is okay for the first three months after release, 12.5% for six months, and about 15.2% for the whole first year.

Certain games and companies can reach conversion rates of 19-30%. I suggest staying on the pessimistic side. Then if you make it, you'll have cause to celebrate, and if you don't, well, there's no use getting upset.

There's no such thing as a free lunch on Steam. The top 100 games generate 90% of net revenue, and this is not going to change - these games sell well, and everyone follows the money. Once you've used up the ad displays you get when your game launches you'll have limited options within the Steam ecosystem. In fact, you only have two real tools at your disposal: discounts to attract new customers and mailing lists for wishlisters.

You also get several ad display cycles to accompany major news and updates: five cycles of 500,000 displays, to be exact. These ads are only shown to users who already own the game... or have it on their wishlist. The resulting conversion rate ranges from 1.9% to 4.2%, so don't expect to get a lot of sales out of this tool. You can also earn some additional display cycles if your game sells well enough and your announcements are rated highly.

The common consensus is that in 2018 you need to sell at least 2,000 copies a day to stay on Steam's list of top-selling games and at least 700-1,000 to keep getting sufficient organic ad displays.

A notification of your game's release will be sent to all users who have it on their wishlists. So the more people who have your game on their wishlist, the better - you'll get a spike in sales within the first few days and make it onto the list of top-selling games, which will get you additional organic traffic, which will lead to even more sales. And so on and so forth.

Steam will provide you with about one million initial ad displays immediately upon release. Based on your sales, Steam will draw a conclusion regarding your conversion rate and the expediency of granting you additional traffic. Good statistics result in more traffic, and more traffic results in better sales. Investing heavily into community involvement and brand advertising often pays off - companies can cultivate large, loyal fanbases and manage to stay on the list of top sellers for months or even years. Their success is based not on bribing some Steam manager, but on maintaining fairly high sales figures for a significant period of time.

Consider Early Access. There are a lot of articles explaining how it can be a useful option for developers. Sure, Early Access isn't early money, but it does give you a unique opportunity to fix your game's technical and design issues and test-drive its marketing materials.

Steam remains tight-lipped on the inner workings of its rotation mechanisms, so all developers and publishers can do is try to use empirical and analytical methods to figure them out. Once in a while Reddit users will dig into how the whole thing works and how it might be exploited.

Let's grab our trusty calculators again. Assuming a 5% wishlist-to-sales conversion ratio for the first month, we'll need to have at least 40,000 wishlisting users in order to sell 2,000 copies a month. This simple math helps us predict a fixed sales volume regardless of any other factors at play.

Opinions about how wishlist entries convert to sales after the first year vary wildly. I know of at least two perspectives that totally contradict each other. One view claims that you can convert up to 70% of your audience within two to three years by taking advantage of Steam sales (both seasonal and individual). So if you have 40,000 wishlisting users right now, you'll end up selling about 28,000 copies.

The other perspective predicts that you'll never sell more copies of a game within its entire lifecycle than the number of wishlisting users on release day. This view is certainly more pessimistic. Ultimately, it's up to you which one to believe.

The diagram above shows the number of wishlist entries and the final outcome (the number of entries minus purchases and unwishlistings). Keep in mind that, for many games, wishlist entries never decrease.

Any activity from your side can result in many users adding your game to their wishlist in anticipation of a price that meets their expectations. But what can you do to convert these wishlisters into buyers? That's something you'll have to figure out for yourself.

The next article in the series will take a closer look at more aspects of Steam, including engaging with your community and the benefits of good store page design.

Nikolay Bondarenko is chief technical officer for Protocol One and owner of Ash of Gods creator AurumDust.