Typography in data-dense UIs

Dense UIs invite a specific negotiation: how small can this text get before it stops being text? Twelve pixels sounds like a hard floor until you notice where it’s going. A pill, a tag, a column label. Twelve pixels of body copy is a strain; twelve pixels on a status tag the user recognizes by shape and color before reading it is fine. The size question is really a role question. Type that identifies can run smaller than type that informs.

Font choice does more of the work than people expect, and most of it is width. Too wide and the table runs out of room before it runs out of data. Too narrow and you save nothing, because users will zoom, or you’ll bump the size up yourself after the first round of feedback.

The sweet spot is a face that stays legible at small sizes without sprawling, which is a less exciting brief than “pick something with personality.” Display fonts are the enemy here. They were drawn to be looked at. A data table is for processing information quickly without admiring the font choice.

Which is where the identity question settles itself. If the product is data tables, the brand doesn’t come from a distinctive font. It comes from the tables working: scannable, quietly ranked, readable on the second monitor at the end of a long day. A typeface that disappears into that job is doing more for the identity than one that signs its name on every cell.