Design Empty, Zero, and Unknown Values
Content & UX Writing · empty states, tables, dashboards
Updated 2026-07-28
- Treat these as four separate states, because they lead to four different actions:
- Zero, which is a real measured value.
- Empty, where a value exists but is blank.
- Unknown, where nothing has been recorded.
- Not applicable, where a value could never exist here.
- Never display them all as a dash. A dashboard that shows a dash for both "no sales" and "could not load sales" will get a business decision made on the wrong one.
- Show zero as a number. Hiding it makes a working metric look broken and removes the reassurance that the system is reporting.
- Distinguish unknown from failed. "Not recorded" is a data gap the user may be able to fill, while "Could not load" is a system problem they should retry.
- Keep the distinction in exports. Blank cells, zeros, and nulls behave completely differently in a spreadsheet, and collapsing them corrupts the analysis.
- Say what the user can do about it. An empty field with an add action is useful, and an empty field with no explanation is just a hole.
- Do not compute totals across unknowns without saying so. An average that silently treats missing as zero is a wrong number presented with full confidence.
- Use consistent language everywhere. The same missing value should not be "None" in one table, "N/A" in another, and blank in a third.
The difference between nothing and nothing known is the difference between a fact and a gap, and interfaces lose it constantly.
Related guides
UX Handbook