JOURNAL
2026: Can Indie Developers Still Make Money?
The Pitfall Field Guide
Okay, so five ways I've watched this go wrong, over and over. Not abstract warnings. Patterns. Each one's a failure mode you can spot from a mile away, usually right after you've stepped in it. I've stepped in a few myself.
Building the answer before knowing the question.
You know the canonical case, right? The all-in-one platform: AI tools, paid courses, plugins, integrations, menus branching into menus. Dreamed up in a day, built for a year, and on launch day zero users show up. Not a slow trickle. Zero. The whole building went up before anyone checked if people even wanted to stand inside it. Validation would've taken a week: a landing page, five conversations. Instead you pay for the lesson at full price, one year at a time. I hate how familiar this feels. (╯°□°)╯︵ ┻━┻
The ceiling before the floor.
The open-world game, alone. The all-in-one platform stuffed with features. Same mistake, different coat. Animation pipelines, vehicle physics, dynamic weather, all designed for a version of you with infinite time and a team, when you're one person with a day job and a weekend. I've been that person. Everything for everyone is nothing for anyone. The ceiling sits so high the floor never gets finished.
Planning that disguised itself as work.
This one's quieter and honestly more dangerous for me, because it feels like diligence. Months. Whole months spent designing products in my head, in notes, in idea lists. A graveyard of a dozen imagined products, each with a name and a feature list and a target market, and not one of them ever gets a line of code. Planning feels like progress, and it isn't. It's the itch of "thinking about building" scratching the spot where building should be. You tell yourself you're being careful, doing research, waiting till you understand the problem. Meanwhile a year slips by and you've got a very organized notebook and nothing shipped. Ouch.
The classic gravestone I keep picturing: a note-taking app for students that syncs everything everywhere and uses AI to summarize lectures. Three weeks on the feature list. The editor never opened. It didn't need three weeks of planning. It needed one afternoon of building and one real student to hand it to. I need that reminder taped to my monitor.
The gold rush.
Ugh, the AI-wrapper wave. A hundred thousand people shipping the exact same thin wrapper around the exact same API in the exact same week. Zero moat, zero distribution. When the ground's that crowded, you don't win by being one more body in the pile. You win by being somewhere the crowd isn't. Building what everyone else is building feels safe, and I've learned it's the least safe thing possible. The gold rush works for exactly one person per idea: the one who got there first. Everyone after them is buying the shovel from someone who's already moved on.
The wrong door.
Here's one I'm trying to tattoo on my brain: consumers don't have a budget line for your product. They've got rent and groceries and a vague hope that the next subscription won't hit this month. Businesses have a budget line. They spend money to save time or to make money, and they'll pay real amounts if you do either. Aiming my early ideas at consumers means asking people with empty pockets to fund the dream. I want to build for the person who already pays for things.
Here's the cheapest market research I've found: pitch your idea to one person who is the entire target market, and listen when they say, almost apologetically, that it sounds cool but they wouldn't pay for it. That's the sentence most of us dodge for months, me included. It costs one conversation, and it's worth a year of building.
Five pitfalls. You'd think one would be enough. The graveyards are full of people like me who hit all five, in sequence, at full tuition.
I can read every field guide and still step in every hole, but it hurts less when someone points first.
Evergreen or Evergrave: 10 Genres That Keep Paying, and 10 That Will Eat You Alive
Aug 14, 202620 MIN READ3 CHAPTERSThirty Cents a Million: DeepSeek V4.1 Flash and the Collapsing Cost of Intelligence
Sep 12, 202620 MIN READ5 CHAPTERSLet It Be: My First Programming Lesson Came from a Beatles Song
Aug 11, 20269 MIN READ2 CHAPTERSAgent Time: Turning AI Idle Hours into the Ultimate Life-Lever
Sep 20, 20267 MIN READ1 CHAPTERThe Case of the Missing Thinking Toggle: ZCode, DeepSeek, and a Rename That Broke the Mapper
Sep 13, 20263 MIN READ1 CHAPTER