Topics
WIP limits, plainly: why constraining work-in-progress speeds the team up
The single setting most teams ignore on day one and miss on day ninety. What it does, why it hurts before it helps, and the numbers that actually work.
More work in progress doesn't mean more output - it means more switching cost, more half-finished items, and longer end-to-end times. WIP limits are the tool that names the trade-off and forces the choice.
What a WIP limit actually is
A WIP (work-in-progress) limit is a cap on how many cards may sit in a given phase at once. If the cap on "shoot" is three, you cannot start a fourth shoot until one of the three either finishes or moves backward to a previous phase. The number isn't a guideline; it's enforced by the board - the new card is blocked at intake until capacity opens.
The point is not to limit output. The point is to make hidden trade-offs visible: "we always have ten cards in shoot but only ever finish one a week" is the kind of pattern WIP limits expose. Without them the team feels busy; with them the team can see where the bottleneck actually is.
Why they hurt before they help
The first week of WIP limits feels punitive. Cards pile up at intake; phases sit at zero while their predecessor is full; people get told no, you can't start that yet. The reflex is to bump the limit "just for now". Resist for two weeks. The discomfort is the diagnostic - you're seeing the real shape of throughput for the first time.
After two weeks the team has adjusted in the right direction: people finish what's in flight before pulling new work, the bottleneck phase has been identified, and the conversation shifts from "are we busy?" to "are we shipping?". The limit is doing its job.
The two-week rule
Set a WIP limit, leave it for two full weeks before adjusting. The first week tests the team's discipline; the second week shows the real shape of throughput. Most adjustments after that are downward, not upward.
The numbers that actually work
There's no universal formula, but there's a strong default: WIP per phase ≈ number of people who can work that phase × 1.5, rounded up. Three editors? Cap edit at five. One person doing intake? Cap intake at one or two. The 1.5 multiplier leaves headroom for blocked items that are technically in the phase but not actively being worked.
The three worst mistakes
One: raising the limit the moment a phase fills. The whole point is to feel the limit; raising it pre-empts the diagnostic. If you must raise, raise after a calm conversation about why, not in the middle of the rush. Two: putting WIP limits on "done" or backlog columns. Those are buffers - capping them creates artificial back-pressure on completion. Three: WIP limits without owners. The cap means nothing if no one is assigned to either finish the in-flight work or formally pause it.
How Production Board enforces them
${PRODUCT} lets you set a WIP cap per phase on each board. New cards are blocked from entering a full phase; the dashboard surfaces phases at or near cap so you can see the queue forming before it tips into a real problem. Caps can be raised or removed at any time - they're not a contract, they're a measurement.
FAQ
Frequently asked
Try Production Board
Free plan, no credit card. We host in Germany. You can export and delete everything self-serve.