You can test this directly by asking what piece of user feedback prompted small visual features of the design. If half or more of the design decisions have som

You can test this directly by asking what piece of user feedback prompted small visual features of the design. If half or more of the design decisions have some kind of backing in experimental data (including interview/qualitative data, that counts too!) you're good. If almost nothing about the design can be directly tied back to experimental data, the designer has gone wrong. I've worked at several places where the designers made a big show of interviewing/looking at data, but then could not incorporate any aspect of it into their designs. Unsurprisingly, their tools weren't used or appreciated.