Choosing a Density for an Analytics Dashboard

After a visual redesign, experienced users complained the dashboard had become slower to use, while new users and stakeholders praised its clarity.

An analytics product used daily by a small number of analysts and occasionally by a much larger number of managers. The redesign applied the company's marketing-site spacing scale to the product.

• The visual refresh was already shipped and had executive support.
• Analysts represented a minority of users and a majority of revenue-linked usage.
• Screen sizes ranged from a laptop to a wall display.
• Touch target minimums applied because a proportion of use was on tablets.

Keep the new visual language and add a density control with three levels, implemented as a multiplier on the spacing tokens, defaulting to comfortable and remembered per user.

Both groups were right. Occasional users benefited from the space; daily users lost the ability to see a full working set without scrolling, which cost them on every task. Density is a legitimate user preference rather than a single correct value, and implementing it as a token multiplier meant one set of components could serve all three levels.

• Task timing before and after the redesign, segmented by user tenure, which showed the two groups moving in opposite directions.
• Counting how many rows were visible without scrolling before and after.
• Interviews with analysts describing the scroll cost on a task they performed dozens of times a day.
• Verification that compact mode still met minimum target sizes on touch devices.

• Three densities to design, build and test rather than one.
• Screenshots in documentation and marketing no longer matched what every user saw.
• A settings decision added for users who did not need one, mitigated by a sensible default.

Analyst task times returned to their pre-redesign levels in compact mode, and occasional users kept the more spacious default. The density setting became one of the most-used preferences in the product.