Open almost any struggling board and you’ll find the same culprit sitting in the In Progress column, unmoved, for the better part of a month: a card that says something like “Redesign the website” or “Launch the new service.” It’s not a task. It’s a whole project wearing a task’s clothing, and no board on earth can make it flow. Every day it sits there, it tells you nothing about how the work is actually going.
The most useful skill in kanban isn’t picking the right tool or drawing the perfect columns. It’s learning to cut big work into pieces a board can actually move. Master that, and every other practice – limiting work in progress, spotting bottlenecks, giving clients a straight answer – suddenly works. Get it wrong, and the fanciest board in the world just becomes a place where big cards go to hide.
Why oversized cards quietly wreck a board
A card that takes three weeks is a problem for reasons that compound. It hides progress – it’s either “in progress” or “done,” with no signal in between, so you can’t tell a card that’s 90% finished from one that’s stuck. It can’t be delegated – it’s so tangled that only the person who opened it understands it, so it can’t be handed off. And it starves momentum – nothing crosses into Done for days, and a board with no movement is a board people stop looking at.
The fix isn’t discipline or a better weekend. It’s decomposition: breaking the big thing into small things that each mean something on their own.
Slice vertically, not horizontally
Here’s the rule that makes breakdown click. When you split a big project, cut it into thin vertical slices, not horizontal layers.
A horizontal split of “build the booking feature” looks like “do all the design,” “do all the backend,” “do all the front-end.” Each of those is huge, none delivers anything usable until they’re all done, and progress stays invisible for weeks. A vertical slice instead is “a customer can book a single appointment type and see a confirmation.” It’s small, it’s complete, and when it’s done you can actually see it work. Then you add the next slice: multiple appointment types. Then reminders. Each card delivers a real, finished outcome – not a layer that only matters once everything else lands.
The two-question right-size test
You don’t need story points or estimation poker to size a card well. Two questions get you most of the way:
1. Can it finish in a day or two? If a card will realistically take longer than a couple of days of focused work, it’s too big – split it. Short cards move, and moving cards keep the whole board honest.
2. Can you describe it in one sentence that starts with a verb? “Add email confirmations to the booking flow.” “Write the pricing page copy.” “Test checkout on mobile.” If you need a paragraph and three “and”s to explain a card, it’s really several cards holding hands. The single-verb sentence is your knife.
Too big, and too small
Most breakdown problems are cards that are too big, and the answer is to split until each piece passes the two-question test. But watch the other extreme too. If you find yourself making cards like “open the file” and “save the file,” you’ve over-diced it – now the board is buried in trivia and the real progress is impossible to read. Batch those tiny steps back into one meaningful card. Aim for cards that are roughly the same size as each other; when they’re consistent, your board becomes a speedometer instead of a guessing game.
Name the outcome, not the activity
One small wording habit sharpens everything: write cards as the outcome, not the activity. “Research competitors” is an activity that can run forever with no clear finish. “Pick three competitors and note their pricing” is an outcome – you know exactly when it’s done. Outcome-shaped cards are easier to size, easier to hand to someone else, and far more satisfying to drag into Done, because there’s no ambiguity about whether they belong there.
Why this makes the rest of kanban work
Right-sizing isn’t busywork – it’s the foundation the rest of your board stands on. Consistent, small cards are what make work-in-progress limits meaningful (limiting yourself to three cards only helps if the cards are comparable). They’re what let cycle time – how long a card takes to cross the board – actually tell you something. And they’re what make delegation possible, because a small, clearly-named card is one a teammate can pick up without a handoff meeting.
So next time a card has been parked in In Progress for a week, don’t push harder on it – break it. Slice it vertically, size each piece to a day or two, name the outcome, and watch the work start to move again. A big project never gets finished in one heroic go; it gets finished one board-sized card at a time. And a simple kanban board is exactly the place to watch that happen.
Sources & further reading:
- Breaking Down Large Projects Into Tasks in an Agile Process – AgileSherpas
- Break Down Any Project Into Manageable Tasks (Work Breakdown Structure) – Asrify