This is a bizarre comment. Studies report average effects, not individual effects. Just because your experience doesn't match what they find on average, the study doesn't become bogus!
You should consider the high number of 'measures' and the negative objective outcomes (being unhelpful) coupled with not paying attention, pushing scripts over situation specific needs. Too much of coincidence over different organizations and broad type of areas to consider just one off experience, more like a trend. With rare exceptions. Of course calling bogus becuase of 'robots' provide the support where human interaction is expected part is a rhetoric here. : ) (not the unhelpfulness of pretty phrasing in place of efficiency though, that's fact)