One board or many? The monday.com structure decisions that are expensive to undo

Some monday.com mistakes take an hour to fix. Others mean rebuilding and re-training.

These are the six decisions in the second category. None of them are difficult, but they all get made quickly and casually in the first week, usually by whoever happened to set up the account.

1. One board with groups, or many boards?

Default to one board. Split only when you have a specific reason from the list below.

Split into separate boards when:

• Different people need different access. Permissions live at board level. If sales must not see HR items, that is two boards. This is the only reason that is genuinely non-negotiable.
• The columns are genuinely different. A recruiting pipeline and a sales pipeline do not share fields. Forcing them together gives you a board where half the columns are blank half the time.
• You will cross 10,000 items or so. Very large boards get slow to load and unpleasant to work in.

Do not split because:

• Different regions or teams run the same process. Use a board with a Team column and filtered views. Five identical boards means every automation and every dashboard built five times, and they will drift.
• You want a clean view. That is what views are for.
• The stages are different. That is what groups are for.

Cost of getting it wrong: merging boards later means recreating items, which loses update history and reassigns item IDs. Anything referencing those items, including your integrations, breaks.

2. Groups as stages, or a status column?

Covered in more depth in the CRM setup guide, but the short version:

• Groups if the process is visual and under a few hundred active items. Better for humans.
• Status column if you need heavy automation or reporting, or you are at volume. Better for machines.

Cost of getting it wrong: moderate. You can migrate from one to the other with a bulk edit, but every automation and dashboard widget referencing the old model needs rebuilding.

3. How many workspaces?

Workspaces are the layer above boards, and people either use one for everything or make one per client and drown.

My rule: a workspace per audience, not per project. One for Sales, one for Delivery, one for Internal Operations. Everyone in a workspace should plausibly care about most of what is in it.

If you run client projects, resist a workspace per client until you have more than about ten simultaneous ones. Before that, one Delivery workspace with a board per client is easier to navigate and much easier to report across.

Cost of getting it wrong: low to moderate. Boards move between workspaces without data loss. The real cost is that people stop being able to find anything, and then they start creating duplicates, which is how accounts rot.

4. Where do permissions actually live?

Understand the three levels before you promise anyone anything:

• Workspace level: who can see the workspace and create boards in it
• Board level: who can see, edit, or only comment on a specific board
• Item level: possible via board permissions like “edit only your own items”, but coarse

The thing that catches people: there is no column-level permission. You cannot hide one column from one group of people on a shared board. If salary must be invisible to some viewers, it belongs on a separate board, mirrored or not at all.

Cost of getting it wrong: this is the one that can be a genuine incident rather than an inconvenience. Decide it before you put sensitive data anywhere.

5. Item names: what goes in the first column?

Boring, and it matters more than anything else on this list for daily usability.

The item name is what appears in notifications, in search, in linked-item pickers, in every dashboard widget. If your items are called Website project, and you have four of those, everyone wastes time every day.

Use a consistent pattern with the distinguishing detail first: Acme: Q3 website rebuild. Not Q3 website rebuild (Acme), because names truncate on the right in most of monday.com’s interface.

Cost of getting it wrong: low to fix, high to tolerate. Bulk renaming is easy, so fix it as soon as you notice.

6. Are you using the number columns as numbers?

Small and quietly expensive: if a value is a number, it must be a Numbers column. Not text.

I have inherited boards where deal values were text, because the person setting it up typed 45,000 EUR and monday.com accepted it. Nothing can total that column. Every dashboard built on it is impossible. Converting later works, but every value with a currency symbol or a comma has to be cleaned first.

Same for dates. A date in a text column is not a date, and no automation will ever fire on it.

Cost of getting it wrong: moderate and entirely avoidable. Set the column type before anyone types anything into it.

A useful test before you build

Ask: if this is still in use in two years with five times the data and half the current team, does it still work?

Most structural mistakes fail that question immediately. Five identical regional boards fail it. Deal values as text fail it. A workspace per client fails it at year two.

The structure that passes tends to be simpler than what people build on the first attempt, which is the whole point.