Density has to match how often and how expertly the screen is used
Aliases: display density · compact vs comfortable · expertise fit
What it is
The same CRM: an intern needs large cards and long labels to find “next follow-up”; a rep with eighty accounts a day needs twenty rows on one screen. Density is right when it tracks frequency and expertise. Sparse is not automatically kind, and dense is not automatically professional.
One density for everyone taxes one end: newcomers drown in texture, experts spend the day scrolling.
Why it happens
Experts chunk the interface and can fetch a field from a packed display. Newcomers still use labels and gaps as a map, so the same pixels are noise. Frequency decides how often that tax is paid: a yearly tax filing can be sparse; a dispatch desk that fires several times a minute treats sparse as added latency.
Density is a person–task fit, not an aesthetic grade. Clutter raises search time, but it hurts experts less because their search is no longer “hunt the screen,” it is “go to a known coordinate.” Frequent users grow those coordinates, so density can rise with skill.
Studying it
Clutter metrics (feature congestion, edge density) can describe the stimulus. Split participants by expertise, then run find-the-item and compare-across-fields tasks. Independent variables: items per screen, spacing step, expertise (novice / skilled) or within-session repetition as a frequency proxy. Dependent: search time, errors, subjective load, scrolls needed to compare two fields.
Lab “experts” are often ten minutes of practice; a sales rep’s chunks took months. Expertise effects are usually under-estimated. When field meanings are still being learned, sparse looks better — that is learning, not steady-state density.
Where it stops holding
Kiosks, one-shot errands, logged-out landing pages have no path to skill; bias sparse. Monitoring, trading, code: sparse blocks the job. The same person is often a novice density on a phone and an expert density on a desktop — the match is the situation, not a personality.
When the cost of error is extreme (a medical dose), even experts stay sparser and more labelled. Density yields to error-proofing.
Applying it
- For each task, write “how many times a day, what happens on error, is it the same people.” Assign a density grade per surface, not one grade for the product.
- Keep first-run paths (onboarding, first import) on a sparse version; do not paste the workbench’s packing.
- How to check: pair a weekly-only user with a daily user on the same comparison (find the three most overdue records). If the newcomer cannot find fields on the dense version, and the regular scrolls more than two screens on the sparse one, density missed the person. Note the screen each one stalled on, and change that screen’s grade — do not globally pad or un-pad.