I worked at a PC shop as a teen. People would bring home their newly fixed system then phone it to complain that it was dead.
Inevitably this would be because the switch on the back was OFF but the user normally turns their computer on through some kind of power strip.
Any form of asking them to turn that switch ON wouldn't work-- either they'd angrily say they already turned it on, pretend to do so, or just refuse.
It was far more effective to tell them to reach behind and turn the switch OFF. "Oh! it's coming on!".
Much later I worked as a field engineer for Juniper and was impressed by the number of times the customer and I would grind on an issue only to try support and have them solve it with basic procedural steps that avoid assumptions and begin with the basics-- stuff that "obviously" doesn't apply. Ever since I've tried to remember that experience and apply it to my own troubleshooting and I think it's been helpful.