I built a VS Code extension to stop shipping features nobody asked for, here's what I've learned so far
I kept building the wrong things.
Feedback lived in Discord, Slack, random DMs. By the time I sat down to code, I was going off memory, which meant I was basically guessing.
Here's what I've come to believe: most solo founders don't have a building problem. They have a signal problem. The feedback exists, it's just scattered everywhere except where decisions get made.
The closer feedback lives to where you code, the faster you close the loop between what users want and what you ship. No context switching, no "let me check the board later." The priority is just always in front of you.
That's the idea behind ShipBoard, user-voted features piped directly into VS Code. Built it because I needed it myself.
Still early. shipboard.dev if you're curious.
What's your system for deciding what to build next?
"Connecting user feedback directly to the IDE is a game changer for staying focused. Most founders lose so much time context-switching between Discord and VS Code—ShipBoard definitely solves that 'signal' problem.
This would be a great project to submit to this competition; it’s $19 to join and the winner gets a Tokyo trip. Prize pool just opened at $0. Your odds are the best right now."
Round just opened 👉 tokyolore.com
The "signal problem" framing resonates. I've noticed the same thing — the distance between where feedback lives and where code gets written is one of the biggest sources of wasted effort for solo founders. Every hop (Slack → Notion → mental model → code) loses fidelity.
One thing I'd push back on slightly: user voting can be a double-edged sword for prioritization. Votes surface what's popular, but popular isn't always valuable. The features your most engaged power users vote for are often incremental improvements to their existing workflow. Meanwhile, the thing that would unlock a new segment of users doesn't get votes because those people haven't found you yet. I've found that the best signal comes from watching what users try to do and fail at — the support tickets, the workarounds, the "I exported to CSV and did this in a spreadsheet." That's unspoken demand.
To answer your question directly: my current system is embarrassingly simple. I keep a single markdown file with three columns: "users asked for this" (with a count), "I observed users struggling with this" (qualitative), and "this unblocks revenue." If something shows up in two or more columns, it gets built. If it's only in one column, it waits. Crude but it's kept me from building vanity features.
Curious whether ShipBoard has any weighting beyond raw votes — like recency or user segment?
That markdown system is genuinely good and simple enough to actually use.
But most solo founders aren't drowning in support tickets, they're hearing nothing. Voting at least gets the willing ones to raise their hand, and those people are worth a conversation.
Weighting is coming. A vote from someone active for 90 days should hit different than someone who signed up yesterday and ghosted.
What are you building?