A community blog post over at GamaSutra by Regency Solitaire developer Jake Birkett is a quick study on the ratio between week-one and year-one game sales on Steam. By gathering an (admittedly small) sample of ratios, Birkett gives some insight into what developers can expect to sell based on their first week’s sales figures.
The range of the 30ish developers who submitted data lies fairly tightly between 2x and 10x (that is to say, year-one sales tend to run between 2x and 10x those of the first week), with a mean of 5.1x and a median of 4.5x. Despite the small number of data points, this gives developers at least a ballpark of what to expect from their games on Steam. Keeping in mind a 5x growth over the course of the year feels like a conservative estimate that leaves plenty of room to be pleasantly surprised. Though, as with all things, there’s room for disappointment, too.
Other factors, such as marketing, buzz and simply having made A Good Game will all have an effect, of course, though Birkett suggests that “the release date of the game didn’t seem to make any difference” and that he “could not detect an obvious correlation between game size/fame and ratio”.
Check out the full blog post for more details. In lieu of a more comprehensive study, this seems to be a good baseline for developers to act against when planning their Steam release and their workloads post-launch.