Even when I think I’m done, I still end up spending half the day tweaking the demo build. While working on it, I finally completed the event where F6E’s consciousness chip gets transferred. Previously, I had temporarily handled it by giving the player a suit program. If I’d focused solely on this event, it would’ve been quicker, but juggling it alongside other tasks took time. Next, I’m planning to build the terminal room on Floor -3 of the Archetype Server to access the Threshold stage, and set up the transition into that area.
I recently got a DM from a dev I don’t know asking how I handle game marketing. I don’t really know what I’m doing either—it’s all trial and error—so I just shared what I’ve learned so far. After sending it, I figured I might as well document it here. Some of this might be specific to my situation, and a lot of it overlaps with advice you can find online.
- Open your Steam page early: If you already have materials ready, it helps to start bringing in traffic as soon as possible.
- Release the demo early and start collecting feedback early: If you already have feedback channels, it may not apply, but in my case, I had almost no opportunities for feedback. I’d heard this advice before but never fully understood it—now I do. For me, players are essentially producers, and the demo was almost the only way I could gather and apply their input.
- Post-demo “dead zone”: A few weeks after release, when the demo loses its freshness, traffic can stop. GPT warned me not to get discouraged at this point. It was disorienting at first, but I’ve learned to accept it as part of the system and now balance feedback implementation with main content development.
- Use Steam visibility tools: I gave GPT things like my click-through rate and traffic graph, and it recommended changes to tags and thumbnails. Nothing dramatic happened, but I was able to experiment and make small improvements—and at least understand how the system works.
- Reach a certain wishlist count before applying to Next Fest: Obviously, the higher the better, but GPT suggested a baseline number to aim for before applying.
- Grow wishlists through demo/Next Fest instead of pitching: I’ve learned that publishers use bots to track games with certain metrics, so it’s more effective to let them come to you through public visibility than by pitching directly. I used to assume publishers would see the “potential” in a half-finished game, but that was a mistake. Pitching might work for games that are either totally finished or wildly unique—but mine was neither.
- Reach out to sites like Lurkit or Keymailer around Next Fest: They seem to offer discounted indie support programs specifically for indie (solo?) devs using the service for the first time.
- Look at other games when morale drops: Everyone’s different, but seeing how difficult this process is for other indie devs helped me realize it’s not just me—it’s the system. (and it’s brutal)
- Use Featured Announcements wisely: I’ve heard you should only use them every 2–3 months. I’m not sure exactly how it worked, but I think I got a small bump in traffic from using one. I’m planning to write another in mid-August, when the new boss is added.
- Keep updating your community and social media:
- Steam seems to be unlikely to bring in traffic on its own unless something special happens—like sustained positive reactions or a Featured Announcement (and even then, nothing is guaranteed). So in my case, I’ve found that I need to drive traffic from the outside. That usually means posting devlogs regularly on social media and community platforms. A single post tends to stay visible for about a day—two at most. If possible, you could try posting once a week with a two-day gap between each post, across one social media platform and two communities. It felt important not to make people think I was spamming, so I tried to adjust the content to fit each space and avoided posting too frequently in any one place.
- Thinking about how to present your game: In my case, working solo meant there were clear limitations on visuals. And while I put a lot of effort into the story and worldbuilding, the game’s genre doesn’t allow those elements to be front and center during gameplay. But surprisingly, those story elements played a bigger role on social media than I expected. Even when I posted simple or unremarkable gameplay clips, if I added a short explanation about the story behind them, people showed much more interest.
- Ask AI often: I don’t know if ChatGPT or Claude can actually give me a strong, coherent plan, but when I fed them my current Steam data and development context, I got some surprisingly specific and useful suggestions. But lately, their responses have become overly accommodating and emotionally attuned, so I’ve had to filter that out manually—which can be a bit of a chore.
This is the first time I’ve actually tracked the numbers and consulted AI step by step. It made me reflect on how reckless I was last time.
While going over Steam with AI—and after spending the last 3–4 years using social media for the first time in my life—I’ve come to feel that exposure works a lot like bubbles rising underwater. No matter how large the bubble gets, nothing happens on the surface until it finally breaks above the water. And when it does, it usually gets noticed for just a day or two at most—then things go quiet again.
But underwater, there are all kinds of unknown obstacles—ones you can only vaguely guess at. So even if you manage to create a big bubble, it can disappear halfway up, or sometimes you can’t even tell whether it surfaced or not. For someone like me, who lacks social instincts and just knows how to work, this system feels especially difficult to navigate. Steam, at least, feels like one of the more fair and structured platforms.
And even when people do see the bubbles and try to check what’s behind them, if what’s down there isn’t interesting, it doesn’t matter how many or how big the bubbles are—nothing really happens. That’s how it feels to me.