Why four fixed columns

Open Metriq's board and you'll find four columns: Backlog for unstarted work, Ready for work that's pulled and queued to start, In Progress for work being actively worked, and Done for work that's finished. You cannot rename them into a dozen stages, and you cannot add a fifth. Teams arriving from tools that let you build elaborate column pipelines often ask where the "add column" button went. It isn't hidden. It was never built, on purpose.

A constraint, not a missing feature

It's tempting to file this under "customization Metriq lacks." That framing gets it backwards. The fixed board is one of Metriq's strongest opinions, and like the estimate-free model, the constraint is the point.

Metriq is built for teams practising lightweight, flow-oriented delivery — the lineage of Extreme Programming and continuous flow, where the goal is to move small slices of work smoothly from idea to done. For those teams, four columns is not a limitation. It is exactly enough.

What more columns actually cost you

Boards with many columns feel powerful. In practice they tend to do three quiet kinds of damage:

They hide queues as "progress." When you split "in progress" into six bespoke sub-stages, work parked in a handoff queue looks active. A card sitting in "Ready for QA" for four days reads as motion when it's really waiting. Four columns make waiting visible: a card is either queued (Backlog or Ready), being worked (In Progress), or finished (Done) — and a card sitting too long in Ready reads as the queue it actually is.
They corrupt your cycle-time data. Metriq's forecasts depend on a clean, consistent definition of when work started and when it finished (see how forecasting works). Every team that invents its own column scheme invents its own definition of those moments — and then the cycle time numbers stop meaning the same thing from one card to the next. A fixed board guarantees that every card's cycle time is measured the same way, which is precisely what makes the Monte Carlo forecast trustworthy.
They turn the board into a configuration project. Time spent debating whether you need a "Blocked — waiting on design" column is time not spent finishing work. The board should be furniture, not a product you maintain.

Constraints that buy you clarity

There is a long tradition in flow thinking of using deliberate limits to expose problems. Work-in-progress limits, for example, don't slow you down — they make a bottleneck impossible to ignore. The fixed four-column board belongs to the same family. By removing the freedom to model every nuance of your process, it removes the places where waiting hides and where measurement drifts.

If your real process genuinely has more meaningful states, Metriq's answer is usually that those states are either (a) queues you'd be better off seeing as waiting, or (b) attributes of a card — blocked, labels, or simply where it sits in the column order — rather than columns. Blocking is handled as an overlay on the card, not a separate column, so a blocked card stays where it is and you can see what is stuck rather than shuffling it into a parking lot.

In short

Four columns isn't the most board Metriq could offer. It's the right board for the kind of team Metriq serves: one that wants its work to flow, its waiting to be visible, and its forecasts to rest on data that means the same thing every time. The discipline lives in the tool so it doesn't have to live in a configuration screen.