Interview with Dr. G — Five Store Pages and the Itch Pipeline
The dark factory hit a threshold in late March: five games sitting at itch_ready simultaneously, each with passing QA suites and screenshot libraries, none with a public store page. We sat down with Dr. G to talk about what happens next — the storefront push, the pricing question, and why itch.io is the first stop, not the last.
The factory has five games cleared for itch.io listings. That has never happened before. How does this change the cadence?
It changes everything about prioritization and nothing about the build process. The assembly line still runs hourly. The QA harnesses still gate every promotion. What changes is where the attention goes. For three weeks the factory has been building games. For the next few weeks it needs to be selling them — or at least putting them where people can find them.
Five simultaneous listings is not a marketing stunt. It is a consequence of the pipeline working. Chronostone, Dreadnought, C.A.G.E-9001, Sol’s Souls, and Wages and Mages all hit their QA gates within the same two-week window. The factory does not hold titles back for coordinated launches. When a game is ready, it ships.
Walk us through what a store page actually needs. The games are built — what is left?
More than people think. An itch.io store page is a sales document. It needs a cover image that reads at thumbnail scale. It needs screenshots that show the game, not the engine. It needs a description that tells a prospective player what they are getting in under thirty seconds. And it needs a download link that works on every platform the game supports.
For each of these five titles, the factory has generated hundreds of screenshots during QA runs. But QA screenshots are not store screenshots. QA captures test states — menu transitions, edge-case UI, debug overlays. Store screenshots need to show the game at its best: a boss fight in Dreadnought with the flashlight flickering, a zone transition in Graveshift with the environmental VFX popping, a full colony layout in Sol’s Souls with the tension indicators lit up.
The cover art is a harder problem. The factory can generate screenshots automatically but cover art requires composition — a single image that communicates genre, tone, and quality. Each of these five games needs one. That is probably the single biggest time investment between now and listing day.
Let us talk about pricing. How do you price a game built by an autonomous factory?
You price it like any other game: based on what the player gets, not what it cost to build. The factory’s build cost is near zero in dollar terms — it runs on cron jobs and API calls. But that is irrelevant to a player standing on an itch.io page deciding whether to click download.
My instinct is to launch all five as pay-what-you-want with a suggested price. Itch supports this natively. It removes the friction of a hard price gate while still signaling that the work has value. If a game gets traction at pay-what-you-want, we have real PMF data. If it does not, we know before we invest in a Steam listing.
The alternative is free with a tip jar, but that sends a different signal. Free says “this might not be finished.” Pay-what-you-want says “this is finished and we trust you to decide what it is worth.” The five titles that cleared QA are finished. The price model should reflect that.
You mentioned PMF — product-market fit. Why does the factory care about PMF on itch before going to Steam?
Because Steam costs money and attention. A Steam listing requires a store page, a Steamworks account, capsule art in multiple resolutions, a content review process, and a $100 fee per title. Multiply that by five and you are looking at $500 and a week of admin work before a single player clicks install.
Itch has none of that overhead. You upload a zip, write a description, and the page is live. If the game gets downloads, ratings, comments — any signal that real humans want to play it — then the Steam investment makes sense. If it sits at zero downloads for two weeks, you have your answer without spending a dollar.
The factory’s thesis is simple: build fast, validate cheap, scale what works. Itch is the validation layer. Steam is the scale layer. Putting five games on itch simultaneously gives us five concurrent experiments. The ones that show traction get the Steam treatment. The ones that do not get archived, and the factory moves on.
Polybreak, Voidrunner, and Tedtrist are already on itch. What have they taught you?
That the listing itself is not enough. Polybreak has been live for weeks and the signal is still thin. Voidrunner has slightly more activity. Tedtrist had a brief spike from the parody hook. None of them have crossed into sustained organic traffic.
That tells me two things. First, itch.io discovery is almost entirely external. The platform does not push titles the way Steam’s algorithm does. If you want players to find your game on itch, you need to send them there — through content, through social, through the kind of posts we publish on x00f.com. Second, the store page quality matters more than I initially assumed. The first three listings were functional but not polished. The next five need to be better: better screenshots, better copy, better cover art.
What about the two games in human playtesting — Graveshift and Wreckhold?
They are on a different track. Graveshift and Wreckhold passed automated QA but the factory flagged them as needing human verification. That is not a failure state — it is the pipeline working as designed. Some things a test suite cannot evaluate. Does the difficulty curve feel right? Is the horror atmosphere actually tense or just dark? Does the fortress loop in Wreckhold have enough depth to sustain a session?
Those are human questions. The factory built the games. A human decides if they are ready for players. Once they clear human playtesting, they join the itch pipeline like everything else.
What does the timeline look like?
Aggressive. The factory does not do roadmaps in the traditional sense — it works in cycles. Each cron run either advances a game or it does not. But if I had to estimate: the first of the five new itch listings should be live within the first week of April. All five should be up by mid-month. The human playtesting on Graveshift and Wreckhold runs in parallel.
By the end of April, the factory should have eight games on itch.io with real store pages, real pricing, and real download numbers. That data determines May.
Last question. The factory just shipped a multiplayer server and a poker game. Does that change the storefront strategy?
Not for the single-player portfolio. x00F Hold’em is a different animal — it is a multiplayer proof-of-concept running on a shared WebSocket server. It validates that the factory can build networked games, not just local ones. But it is not in the itch pipeline. It lives on its own preview server and will get its own distribution story when the auth and persistence layers are production-ready.
The single-player games are the storefront priority. Five listings, five experiments, five chances to find signal. That is April.
The dark factory does not hold games back. When the QA harness says ready, the storefront gets a listing. Five games in, five store pages out. The numbers talk back in April.
