According to Cubed, architects and designers have had a lot of idealistic approaches to office design but companies have pursued more ethnographic approaches in recent decades:
In other words, interior designers are struggling, hard, to be relevant and so are architects. And so are space planners and so are product designers. Both Steelcase and Herman Miller have intensified their use of anthropological techniques – participant observer, video ethnography, object testing – to understand office workers’ behavior and to design around behavior rather than attempt to influence or change it. (p.308)
In other words, some have moved from top-down designs borrowing from trendy ideas about office behavior and design to actually studying the people at work in offices to see what might or might not work. What you likely gain are the small but important pieces of information that might be lost with other methods of data collection. Imagine the all-important serendipitous short conversation between two employees who pass in the hallway. Designers and others often think these are really important as they generate social connections and innovation. But, if you asked on workers on a survey about such short encounters, they may not recall them or think much of them at the time.
Lots of businesses could benefit from ethnographic approaches as they would get the viewpoint of the workers instead of the interpretations of management.