Game Career Guide is part of the Informa Tech Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.

Get the latest Education e-news
  • Postmortem - Starcom: Nexus

    - Kevin Lin

  • Entering Early Access

    From a sales and marketing standpoint, entry into Early Access needs to be taken seriously. For almost all games on Steam, EA or otherwise, their launch week shapes their entire return. Almost all games' first year sales are a single digit multiple of their first week, generally clustering around 3 to 6 times. E.g., a game that sells 100 copies in the first week could expect to sell 300 to 600 copies in the first year.

    From my analysis of visible Steam data, many EA games receive some kind of bump when graduating to full release (sometimes even a big one) but if a game hasn't found some footing in the market at the start, it will almost certainly fail.

    I had personally committed, both to myself and players, not to abandon Starcom: Nexus during Early Access.

    But my savings were not inexhaustible. By this point I'd spent 3800 hours over the equivalent of 16+ full time months and roughly $10,000 of my own money. This is excluding the larger cost of, you know, living. I'd been working 60 to 70 hours a week for several months. Combined with the uncertainty about how the game would fare, this was producing a large amount of stress.

    Cumulative hours to Early Access

    There is a scarcity of hard data to predict a launch success, particularly for EA titles. One analysis done by Jake Birkett estimates that during its first week, a game will sell 0.5 copies for each player who has wishlisted it before launch. But this was a limited survey and there was a range of results: from 0.14 to 1.8. Moreover, the games surveyed had mostly been released at a time when Steam was seeing fewer releases per month.

    I felt a worst-case scenario would be the game selling only 400 copies in the first week. Combined with a pessimistic tail model, that could be as low as $30,000 lifetime, and that's before Valve took their revenue share. If it did that poorly, my plan was to do an aggressive but abbreviated Early Access and start sending out my resume in a few months. I did not believe this was a likely outcome, but it was a possibility that had been magnified in my anxiety-wracked mind, which had begun contemplating "even worse than worst-case" scenarios.

    A slightly more optimistic model was 800 copies the first week. With a reasonable tail and a moderate bump on graduation, this could justify the "bygones" decision I'd made, at least in the sense that I would have made more than a living wage from the time of that decision to the time of full graduation. Which is a very forgiving criteria, but also excludes the value of an established audience for future games. In this scenario my plan was to spend 6-9 months in EA and after launch evaluate the possibility of a sequel.

    At the most optimistic end, I estimated as high as 2000 copies in the first week was possible based on wishlists and followers.

    After nearly 4000 hours of development, I pushed the launch button that officially triggered the game's entrance into Early Access.

    The Button

    After an incredibly stressful first day, it became clear that the worse-case scenario wasn't going to happen: the game sold over 500 copies in the first 24 hours. The sales rate dropped rapidly after the initial rush, but I had expected that. Like the first week making up a significant percentage of annual sales, the first 48 hours usually make up a majority of the first week sales.

    One week after entering Early Access the game had sold over 1500 copies and grossed a little over $20,000.

    First three months of Early Access

    At this point, I'd like to draw attention to a convention in indie PC game development: when developers publicly talk about gross revenue, we are not usually referring to the money we receive before expenses, which is the standard accounting definition of gross revenue. We are referring to the money Valve (or another distributor) receives. Why? I don't know, probably because it sounds more impressive. But that amount never appears in my bank account. It's like including part of your boss's salary when someone asks how much you make. So when I say the game grossed $20,000, I mean that Valve collected $20,000. After various deductions like VAT, returns and Valve's revenue share, Wx3 Labs LLC (which is just me) grossed around $12,000. Which still sounds pretty good for one week, except that a) I had spent $10,000 in development and b) sales drop off a LOT after the first week.

    Even though I had only net $2000 in profit for my 16 months of labor so far, this was a good start and I could at least hold off work on my resume for the moment and focus on finishing the game.

    Early Access Development

    The beta testing paid off in the sense that nothing went catastrophically wrong on launch. Players were in general very happy: the first few dozen player reviews were all positive, and after 100 reviews it had a score of 94% positive.

    But it was still obviously a work in progress. After several patches to hot-fix a few potentially serious bugs, I started work on the first post-release milestone, adding features and content to extend the game from a polished beta to a full game.

    I tried my best to be accessible to the players via multiple channels: I checked the Steam discussion forums every day, was always logged into Discord and read the "F8" feedback logs regularly. And, as painful as it was, I read every single negative review on Steam.

    Listening to players' pain points gave me the best insight on where the game needed work. At the same time, I had to resist the temptation to pivot the overall design based on the strongly expressed opinion of a handful of players.

    Over the next ten months the game grew. From a few dozen star systems to over 150. The median time for players to complete the game increased from six hours to well over twenty. I added features that players asked for like attacks drones and quick-saves. Towards the end I added an element that I had conceived of and teased in the very first piece of concept art: a mysterious Dyson sphere that the player could enter.

    The game continued to be well-received and saw spikes in customers with sales and Steam events, but after the post-launch cliff, average revenue continued to slowly drift downward: it looked like the game was experiencing a fairly typical Steam release life cycle. There was plenty more I could do on the game, but the longer I spent the less overall return I would see, and the less budget I would have for my next project.

    I also wanted to make sure that the game graduated from Early Access before I was too burned out to properly support it. The game had a beginning, middle and end. The core features were all in, tested and well-polished. In the Steam Store description I had pitched a 6-12 month Early Access period and that was rapidly running out.

    So in late October I announced a December 12th release date, exactly one year after the initial launch.

    Graduation Day

    In the lead-up to the game's graduation from Early Access, I had frozen development to ensure that the first version non-EA players experienced was as bug-free as possible. There was plenty for me to do: market outreach to potential streamers, more analysis on the analytics data and even starting preparatory work for the next project by improving my 3D modeling and texturing skills.

    I also spent a lot of time trying to guess what would happen when I pushed the "launch button" for the second and final time.

    By this point, the game was doing "great" by indie game dev standards but just "okay" by the standards of someone trying to start a business:

    After one year in early access it had sold 6500 copies and grossed $98,000. After Valve's cut, VAT, and other deductions that left $60,000.

    During Early Access I had spent another $10,000 in external costs for a total of roughly $20,000. That left $40,000 in net revenue. Divided over the entire development lifecycle so far, that worked out to $16,000 per year. That's more than US minimum wage but just barely, and not even close to a living wage.

    But there was the fact that unless graduating from Early Access actually caused sales to drop to zero, the game had expectations of earnings that would continue to come in over its lifetime, albeit at an ever-diminishing rate. My ballpark guess was that with zero graduation boost, the lifetime return would work out to around $32,000 net per year of development after external costs, before taxes. That's for the full development lifetime, including both the "pre-commit" time when I had yet to really analyze the prospects for the game, plus some amount of post-graduation support.

    On top of the financial return, I had gained a lot of skills, both in technical and design areas, as well as a much better understanding of the business aspects. I had a base of several thousand players with a positive experience who might be interested in a sequel.

    And there was the hope, expectation even, that graduating from Early Access would produce a measurable sales boost.

    Conventional wisdom among devs is that you only get one launch and you should treat entry into Early Access like it's your only shot. Anecdotally, some indie developers have reported little visibility boost on graduation, but others have seen a significant increase in sales.

    During pre-EA market research, I had been tracking a large cohort of Steam games that launched around the same time as mine. Some of those were also Early Access titles and a dozen of those had graduated by this point. Most of those showed a spike in review counts after graduation that suggested a healthy increase in sales, in some cases apparently matching or exceeding the EA launch over a slightly longer time period. A few of the titles, however, showed no increase at all around the time of their graduation. And again, review counts are only a rough proxy for sales. Guessing from limited data (which was something I did a lot of while wearing my indie market analyst hat) I thought there was a good chance the game would sell at least 30% of its initial launch numbers in the few weeks after graduation.

    In my marketing journal, the day before full launch I wrote: "My best guess is that first month of graduation will sell between 700 and 3000 copies, but that's a wide range."

    A few hours later, apparently gripped with pessimism, I added: "I have now convinced myself after looking at Steam that it will get no boost above a normal sale."

    Those late fears turned out not be realized. Quite the opposite: initial sales were well above my projections. In the first 24 hours of full release Starcom: Nexus sold over 1000 copies, and another 1000 in the next 24. Numbers began dropping rapidly after that, but after a week the game had grossed almost as much as the entire year of Early Access. That sounds a lot bit better than it really is because of how front-heavy game sales are, but it was still fantastic.

    Early access launch vs Full launch

    So I was wrong in my graduation predictions. It's the kind of wrong I'm delighted to be, but it was still wrong and it would be helpful to understand why.

    Looking back, there were several major causes:

    1. In my examination of recent Early Access graduates, I think I had subconsciously "rounded down" some of the graduation ratios (ratios of graduation week review counts to launch review counts), possibly in a sub-conscious attempt to avoid severe disappointment.
    2. A number of graduates had their graduation review counts spread across a wider time frame. I.e., while most launches have the bulk of their first reviews in the first week, many EA graduates saw the reviews spread over 2-3 weeks.
    3. During Early Access I consistently had around 1 review for around every 30 copies sold. After graduation, that ratio changed to about 1 review for every 40-45 copies sold. If that shift holds true for other Early Access titles, it would explain a good chunk of my underestimation.

    As of writing, a little under six months after entering full release, the lifetime results for the game so far are:

    • Steam gross revenue: ~$460,000
    • Copies sold: 29,000
    • External development expenses (music, art, licenses, etc): $26,000
    • Total development time: equivalent to 30 full time months

    The net result has been that the game has so far earned me a bit over $8000 per month of development (before income tax).

    That result from creating the game I've always dreamed of making is a feeling I can't describe. I am tremendously fortunate to have had this opportunity.

    Starcom: Nexus is now available on Steam.


comments powered by Disqus