Her fourteenth ticket arrived on a Thursday afternoon. Subject line: "Document access issue again." I opened her account history and counted. Fourteen tickets in six weeks, all under the same category, all marked resolved, each one a slight variation on the one before it. The last agent's note was four words: "resent knowledge base article."
My first read was the obvious one. This person is not reading the documentation.
I pulled up the standard article, typed the link into the reply draft. Then I noticed something. Every follow-up ticket she opened started the same way: "Thank you for the help, I'm still having a bit of trouble with..." Not frustrated. Not escalating. Apologetic. That pattern nagged at me. I had seen frustrated users who came back repeatedly. They sounded different from this.
I closed the draft and called her instead.
The call started strangely. Before I had said much beyond my name, she apologized. Sorry for opening so many tickets. Sorry for not figuring it out herself. She said she was probably missing something obvious. I asked her to walk me through exactly what she was trying to do, and she did, clearly and in detail. She knew the interface. She knew the steps the documentation described. She could practically quote the articles back at me. She was not confused at all. She was precise.
Then she mentioned the USB drive.
Every Friday afternoon for the past six weeks, she had been copying her files to a USB drive. "Just in case," she said. Just in case the documents weren't actually saving. She wasn't sure the cloud storage was working, so she had built herself a backup. She did not frame this as a complaint. She framed it as a reasonable precaution she had taken to manage her own uncertainty.
I asked her to show me what she was doing in the system.
Three minutes in, I saw it. She was working in shared drives the way you would work with folders on a desktop: as containers you copy files into and out of, where location determines access. That is a sensible mental model. It is just the mental model for a different kind of file system. The platform she was on managed access through membership, not location. A file in a shared drive was accessible to the drive's members regardless of where you put it inside. Moving it did not change who could see it. Nothing in the documentation explained that distinction, because the documentation assumed you already understood the underlying logic. Every article said "add the file to the shared drive." Not one of them explained what a shared drive actually was, or why it behaved differently from a folder.
Her mental model was not wrong. It was just calibrated to something else. And we had spent six weeks sending her instructions that assumed she was already past the step she was stuck on.
The documentation had answered the question she wasn't asking. It told her where to click. It didn't tell her how the system thought.
I asked her to stop what she was doing and watch my screen. I shared my screen, pulled up one shared drive, and explained the membership model in plain language. Not the steps: the logic. I showed her why a file behaved the way it did, and how that was different from what a folder would do. I walked through three of her actual use cases, showed her what each outcome would look like, and let her try one herself. Thirty minutes. She thanked me, we hung up.
She never opened another ticket about document access.
The fourteen tickets before that call took longer to resolve than the call itself. Fourteen replies, fourteen article links, fourteen status updates. Probably five or six hours of cumulative team time, spread across six weeks, not counting the time she spent confused, or the USB backup she had been faithfully running every Friday afternoon.
Something shifted in the way I read my queue after that. I started looking for a specific pattern: multiple tickets from the same person, same general category, never angry, always polite, always followed by some version of "I tried what you said but I'm still..." These are almost never difficult users. They are almost always people who are embarrassed to say they don't understand something foundational, because they believe they should already understand it, and they don't want to look incompetent to IT.
They have read the documentation. They have followed the steps. The documentation has just been answering the wrong question: it explains the interface, and they needed someone to explain the concept underneath the interface.
The support answer for that gap is not another article. It is a conversation, the kind where you ask them to show you what they're doing and you watch until you find where their model of the system and the system's actual behavior diverge. That divergence is almost always small, and the fix is almost always faster than anything you've been sending.
There is a specific tell I watch for now: three or more tickets from the same person, same category, within four to six weeks, each one politely worded, each one opening with thanks. When I see that pattern, I don't queue up another article link. I send a calendar invite.
I think about that USB drive sometimes. Six weeks of work copied to a physical disk, every Friday, because no one had taken thirty minutes to explain how the system thought. She was not wrong to do it. She was managing uncertainty the best way she could. The certainty was sitting in our knowledge base the whole time, behind an assumption we had never thought to question.
If you're sitting on a queue with a few names that keep coming back, and you've started mentally filing them under "difficult," it may be worth pausing on that. The ones who stay polite are usually the ones who need the conversation most, and are least likely to ask for it. Happy to compare notes if this sounds familiar.
Comments