I often wonder how this market research is done and if there are any external influences. Even the worst ideas seem to have the backing of "market research". Do the people they ask of these ideas just feel compelled to give a positive answer so they will get their $10usd? Or is there something else going on?

Take the late 1980s gimmick "matrix management", of which I was a victim (not as management). Management consultants simply blurted out that this was a great idea! Never having actually given it a test run in real life.

It means you have multiple managers so the employee (me) could help out different departments/projects with his/her skills. What actually happened was the managers could rarely meet to determine who's project had the highest priority, as every project was an emergency. Hence, very little work actually got finished as I couldn't do much until the managers decided which was the highest priority project.

In a deeper link we have a mention of GfK's Museum of Failures. I do surveys for GfK, I didn't know they had this museum. It would be fun to read through it.

As far as the speech-to-text software goes (brand Dragon Software), I begin seeing this in the mid-1990s as it was "maturing". Quite a few people bought it, but found out it failed, very hard. Why? Nearly every office had a noisy air conditioning system which would confuse the translator. So it worked fine in a quiet demo room, but use it in an actual office, and it failed miserably. I wonder if a unidirectional mike could have helped this.