A kanban card looks like the simplest thing on your board – a little rectangle with a few words on it. But that little rectangle is doing a surprising amount of work. It’s a promise about a piece of work, a pointer to everything that piece needs, and a tiny status report all at once. When a card is well made, anyone can glance at it and know what it is, who’s carrying it, and whether it’s on track. When it’s badly made, the board turns into a pile of vague reminders that only make sense to the person who wrote them – and usually only for about a week.
Most teams get this wrong in one of two directions. Either their cards are too bare – a three-word title and nothing else, so every card triggers a “wait, what is this again?” conversation – or they’re stuffed so full of description, sub-tasks, and pasted email threads that the card becomes a document nobody reads. The goal is the narrow middle: enough on the card to act without asking, little enough that it can be taken in at a glance. Here’s what actually earns a place there.
Start with a title that names the outcome
The single most important thing on a card is its title, and the best titles describe a result, not an activity. “Website” tells you nothing. “Redesign the pricing page” is better, but “Pricing page shows the three new plans” is best of all, because you can tell when it’s finished. A good title answers “what will be true when this card is done?” – and if you can’t write that sentence, the card probably isn’t clear enough to work on yet. Aim for something a teammate could read cold and understand without you standing next to them.
Keep it short. The title is a headline, not the story. Everything else – the detail, the context, the how – can live in the card’s description or a linked doc, where it’s available if someone needs it but never in the way when they don’t.
The three anchors that make a card actionable
Beyond the title, three small things turn a note into a workable card. The first is an owner: a single name responsible for moving it, even if others help. Shared ownership is quiet non-ownership, and cards with no clear owner are the ones that sit untouched. The second is a needed-by date, but only a real one – a genuine deadline or commitment, not a wish. A board where every card is “urgent” or every card is dateless tells you nothing; a board with a few honest dates tells you exactly where to look.
The third anchor is a shared sense of what “done” means. For simple work this is just understood, but for anything with room to argue, a one-line acceptance note saves a round-trip later. This is your definition of done shrunk to card size – “approved by the client” or “deployed and the old page redirects” – so the person doing the work and the person receiving it agree in advance on where the finish line sits.
Context that saves a conversation
A few extra fields aren’t clutter – they’re the difference between a self-explanatory card and a chat thread waiting to happen. A short description gives the why and any constraints. A link to the relevant file, ticket, or email thread means the card points to its own evidence instead of forcing a hunt. And a note of who asked for the work, or which project it belongs to, tells you who to go back to when a question comes up.
Size matters too. A card should be small enough to actually move across the board in a reasonable time; if it’s a giant, that’s a signal to split it before it lands, not after. This is where good card content and right-sized cards meet – a card that’s clearly described is far easier to slice into pieces the board can carry, because you can see the seams in the work.
Signals that let you read the whole board at a glance
Some of the most useful information on a card isn’t words at all – it’s colour and markers your eye catches before you’ve read anything. A type tag (bug, client request, internal, admin) lets you see the mix of work at a distance. A blocked marker is the single highest-value signal you can add: the moment a card can’t move, flagging it turns a hidden stall into something the whole team can see and help clear. That visibility is the entire point of a board over a hidden to-do list – problems surface instead of festering.
Card age – how long it’s been sitting in its current column – is worth surfacing too, whether your tool shows it automatically or you add a small dot. A card that’s been “in progress” for two weeks is telling you something even if nobody’s said a word. These signals are cheap to add and they’re what let a manager glance at the board and understand it in seconds rather than reading every card.
What to leave off – and a template to steal
Restraint is half of good card design. Don’t paste whole email chains onto a card; link them. Don’t build a ten-item checklist for a two-hour task. Don’t add fields “just in case” – every field you add is a field someone has to fill in and everyone has to read past. If a piece of information doesn’t help move the card or read the board, it belongs somewhere else. The test is simple: could a teammate pick this card up cold and start, without messaging you? If yes, it’s complete. If no, something essential is missing – and if it takes them a minute to find it, something inessential is in the way.
Here’s a template worth standardising on: a title as an outcome, an owner, a needed-by date (only when real), a one-line description, a type tag, and a link to anything relevant – plus a blocked marker the moment it applies. That’s it. Agree on that shape as a team and every card gets filled in the same way, which is what turns a wall of rectangles into something anyone can read. If your board is digital, a simple kanban tool with consistent card fields does this quietly for you – so the next time a card lands on the board, it already carries everything the team needs to pick it up and run.
Sources & further reading: