Told it before but:
Teacher in a school said that their computer was slower than everyone else's. Kept complaining and complaining and complaining.
We kept testing and timing it and it was exactly the same all the time. Identical to all the others. And there's a reason for that... it was identical hardware purchased at exactly the same time as everything else, with identical specs throughout the site, with identical imaging from a WDS installation image that was identical for all machines. It was reimaged at the same time as all the others with the same image as all the others. It even had an identical "maintenance" regime for when we cleaned it, etc.
But, honestly, no different to any other computer. Was happy to believe that maybe environment, or user, or random quirk, or component failure or whatever was making a difference... but we couldn't measure any difference.
At one point, we even went around every single computer on site (all identical) and benchmarked them one after the one... time to boot, time from login "Enter" key to loaded desktop with no disk activity, time to load a given (real) user's profile, as well as all the official benchmarks, etc. etc.
It came to a head after this teacher kept complaining, for probably a year or so, and raised it repeatedly to... the head. And obviously at that point I, the IT manager, "had to do something".
So I did. I walked over. I told them I'd handle this "personally." I disconnected the machine myself. I brought it to the IT department myself. I dismantled it myself.
And then I waited. Said teacher eventually visited to check on progress (as she had no machine) and saw what was clearly their machine (numbers, stickers, etc.) in pieces on our worktop. That's what I was waiting for.
Then I left it there for another day or so. Then recombobled it. The exact same components, connected in the exact same way in the exact same machine.
Gave it a bit of an external clean (nothing that would make any difference, just removed her stickers).
Benchmarked it (always have a paper trail!) and guess what... identical to before. Identical to last month. Identical to last year.
Then I made a big fuss of taking it back to the room and cabling it back up to the exact same cables in the exact same location in the exact same way.
They WOULD NOT SHUT UP about how great it was to have a "new" machine that "worked properly". How "fast it is". How much it had removed stress from their teaching. How wonderful it was. How my intervention had made such a difference. etc. etc. etc.
This went on for - I kid you not - years. She kept thanking me for her "new machine" every time I met her in the staffroom or the dining hall. Constantly.
It even became a problem because by then she had told other teachers who all wanted new machines for themselves too. Why should she get special treatment? Etc. The ones we could trust, we told them what we'd done. And they laughed and apologised and said "No worries, just thought we were being left out". The ones we couldn't trust we either made excuses or (if they were nice) gave them the EXACT SAME treatment as we had this teacher. Yes, we took their machine, applied the special "placebo" patch and, as if by magic, they "ran faster" (with the exact same benchmark numbers). Same login times (to the second!). Same performance metrics. Same everything. But apparently, by some mysterious homeopathic avenue, they were "better".
When that teacher left several years later, she still mentioned how great her machine was and how she was grateful for my intervention. As an IT department, we were PISSING ourselves trying not to laugh.
Placebo is strong in IT, don't let anyone tell you otherwise. Users absolutely CANNOT be trusted to gauge performance of a machine.