One main problem is shoppers who do not confirm their responses. When I edited, I could not approve a report unless each one was verified. (This is often complained about here, but after my stint at editing I totally get it.) OFTEN, the wrong yes/no answer was chosen. For example (in a list where all of the other questions had "correct" answers of "Yes"...), "Did you see a towel on the counter" Answer: "Yes." But it was not discussed in the narrative, and in this case, there was not supposed to be a towel out in sight. So back to the shopper for clarification: Q: Where did you see a towel? A: Oh, I didn't see any towels. So that answer had to be changed before sending it through.
Another question: "Was there any debris or insects in the dining area?" A: "Yes." Since there were no details in the narrative... back to the shopper to clarify. The resubmission said something like "There was any debris or insects in the dining area." Still unclear, so back to the shopper, who ultimately comprehended the meaning of the question and said: Actually, no, there was no debris or insects.
The editor has zero power over poorly worded questions and guidelines. If I sent a numbered list of five items to correct, maybe two would be addressed. I had to follow up until all of the inconsistencies were gone- hopefully without introducing any new ones. Each of these back and forths could easily add on 12 hours to the editing process, and if it leaked over onto a day I was scheduled off, it was still my responsibility. So many reports were just a hot mess, and the fastest solution was just to rewrite it from scratch using the shopper's information rather than trying to revise and move their sentences around to be coherent.
Why don't the MSCs just drop those bad shoppers? "We need them more than they need us." I'm more than happy to clean up someone's grammar problems, but the real problem with editing is the issue of reliable reporting. And I have to strongly question the observation skills of shoppers like in the examples above.