Five votes a month each. Backing an idea has to cost something.

The problem

The best ideas about how a business actually runs come from the people doing the work. Those ideas were arriving in hallway conversations and one-off messages and going nowhere in particular — not because anyone dismissed them, but because there was nowhere for them to go.

Three things were broken. No list, so an idea existed only in the memory of whoever heard it. No way to tell whether an idea was popular or just loud, so priority went to whoever was most persistent or most senior. And no feedback loop: somebody who suggested something heard nothing back, so they stopped suggesting things.

Nobody asked for this to be fixed. The board, the argument for it and the mechanism underneath it were self-started; the alternative was deciding what to build next from whoever happened to be in the room.

Five votes a month, and that is the whole design

Anyone can post an idea, everyone can vote, and the board is one ranked list. The constraint is what makes it work.

Each person gets five votes a month. Five, reset on the first, and never for your own idea. Voting is one tap; pull a vote back and it returns to your budget to spend elsewhere. The budget sits permanently in the header, so you see what you have left at the moment you decide whether to spend one.

When votes are free, everything gets upvoted and the ranking measures how many people saw a thing, not how much anybody wanted it. With five, backing an idea means passing on another, so the top of the board states what the team wants most.

Every idea says where it got to

Ideas move through four stages — idea, planned, in progress, built — and the stage sits on the card as a coloured rail, permanently.

That is not a project management feature. It is the feedback loop: the person who suggested something watches it move without asking, and without anyone having to remember to tell them. Built ideas get their own card treatment, so completed work stays visible instead of disappearing into an archive. Nothing is deleted outright either — removed ideas can be restored.

The rest of it

• Writing it up is not the barrier. Appetite for writing was filtering out good thinking, so the compose form turns a rough sentence into a clean title and short description — keeping what you meant, inventing nothing you did not suggest. In the form itself, and optional, always.
• Taking part is rewarded, not just winning. Points accumulate for posting, commenting, having your idea backed and showing up at all. Standings run two columns, top ideas and top contributors, each switchable between this month and all time; the most-voted idea of the month wins a prize.
• The ideas come first. The leaderboard is collapsed by default, contributor email addresses are masked in the standings, and ideas and comments can both be posted anonymously.
• It comes to people. A short digest goes out every morning with the new ideas and new comments.
• Built for a phone. Cards reflow to a single column, category filters scroll sideways under a thumb, and text fields are sized so tapping into one does not zoom the screen.

What changed

• Ideas are written down in one place. [X] ideas posted, filed under seven categories, none dependent on somebody remembering a hallway conversation.
• Priority stopped being decided by volume. Five votes per person per month, non-transferable, no self-voting — so the top of the board is what the team most wants rather than what the most persistent person asked for.
• The feedback loop closed. Four visible stages on every card mean the person who suggested something sees it become planned, then built, without asking. [X] ideas have become built tools.
• Contributing got cheaper. A rough sentence becomes a properly written proposal, so the barrier stopped being willingness to write.
• The board comes to people. A morning digest and a phone-first design, so participation does not depend on anyone remembering the tool exists.

---