I think this is a statement easier to conclude in hindsight, especially as you are primed with "this story is describing a scam, definitely". The author describ

I think this is a statement easier to conclude in hindsight, especially as you are primed with "this story is describing a scam, definitely". The author describes the thought process and what ended up nudging them toward believing the scammer about the workflow. A code sent like this in a legitimate workflow could be plausible. Maybe it's a requirement to ensure that the customer is indeed acknowledging the operation and the CSR isn't taking actions behind the customer's back, for instance. The author had a lot of signals pointing toward legitimacy to counteract their natural skepticism, it was a stressful situation and the nature of a phone call puts time pressure into the decision making, increasing the odds of a mistake. Your example points out that false positives on the "scam or ham" decision do have a cost to the contact recipient too, so "never respond to anything" comes with risks and costs too. It's hard to be perfect.