Sebi wrote this. It’s a really good read, based on many of our evening conversations, and triggered by a great talk on photography which emphasised the fact that you need 18 to 20 impressions in order for a random stranger to recognize your name as a photographer. We both listened to that talk this morning, and that’s when it hit me: 18-20 impressions are an amazing number that can validate all the (frequently very diffuse) advice giving to game developers (and photographers, and writers, since we are at it) to market themselves early. As early as when the idea of a game (book, photography) actually hits them.

An impression is an occasion for a random stranger to hear about you. An impression is a tweet you notice, an article you read, a Facebook post that draws attention. An impression is also meeting someone at a convention, or otherwise. An impression is the number of times you see a game suggested on Steam. The hunt for impressions is the reason you see so many [random big brand here] ads on TV, on the street, everywhere. They are doing everything they can so you see them enough times to remember them.

Now let’s do some calculations.  As a game developer/ photographer/writer/creative doer who actually wants to make a living out of your passion, it is wise to estimate your development time and associate some numbers which will give you a rough understanding on how many copies you have to sell in order to recuperate your investment.

Now I’m gonna go really conservative and assume you are a one man gamedev band who makes a premium (not free to play) game in hers/his spare time over 2 years. Given the extraordinary tendency of game developers to crunch and exhaust themselves (which I deeply condemn), I will consider one full time year of development for economic purposes. According to Gamasutra, a US based employee in the games industry makes about 85000 dollars per year (I used the last numbers listed in the gender gap section). Assuming the absolute impossible idea that you have no other costs, if you spend a total of a full time working year to make your game, you must sell for 85000 dollars to recoup your investment. These numbers are extremely conservative and do not include profit or any external spendings whatsoever (which is another way to say, they are unrealistically low, but also debatable. You could also theoretically relocate to Rwanda in order to make your game, since it is one of the poorest country in the world. But then again…).

If you sell your game with 15 dollars a piece, completely by yourself and preferably by cash paid in hand or sent by snail mail 🙂 (so you have no Steam fees, no PayPal fees, no nothing fees; well, this excludes snail mail, too, because that is also a cost), you must sell 5666 (heh!) copies to get your 85000 dollars back. And of course, never do discounts.

Now I don’t know much, but I think there is a pretty fat number of games on Steam that sell less than 5666 copies. I just Twitter-spawned the amazing Steam Spy (sorry, Sergey!) for some exact data, but meanwhile I just browsed the All New Releases tab on the Steam frontpage, and with surprisingly little persistence, I found a bunch of examples that support my claim. I did not search much and I did not go far in the past because the games industry is hit driven, right? But here are some older examples, some of which very close to my heart. And for those examples that are close to my heart, I know how absurdly small the numbers I am playing with here are.

To ensure 5666 sold copies you must impress 5666 people 18-20 times. This means about 101988 to 113320 actions that you take in order for people to get a chance to know about your game. Sure, you have friends who are already impressed by you, you might be lucky, you might get to know the most famous publisher or journalist at some game convention (dammit, no game conventions, those Cost Money!), you might strike gold like Notch did, right? But there are more chances that you might not, so why not start taking those 100000 actions early, and do your best to make people remember you.