The button that didn't exist
How three users tried to click a thing that was not there, and what I built instead.
Three users in a single week tried to click on a thing in my product that was not a button. It looked like text. It was text. But they had all read it as "click me." I had not seen it for two years and they saw it the first time.
What I had built#
A label, set in bold, that summarised the state of the page. A counter, basically. Twelve points to bold text, sat next to a sleeker secondary button. Visually, the bold text was the heaviest thing on the screen. Weight reads as importance. Importance reads as actionable.
The three users were not wrong. I had built a button by accident.
What I changed#
Two options. I could make the text less prominent, or I could make it a real button. I made it a real button. The button now opens a panel with the underlying detail. The panel is useful. The thing the users were trying to reach turned out to exist, but I had been hiding it behind a "more info" link three pixels tall.
What I learned#
Visual weight is a promise. If your strongest visual element does not respond to a click, you are breaking a promise. Either the element shouldn't be the strongest, or the click shouldn't be unanswered. Most of the time, the click should be answered.