When a client said “the site won’t open,” my only move was to keep refreshing

Last Thursday afternoon, I was sitting in a Hangzhou café when a client messaged: “The admin panel keeps spinning after I submit. Is it broken?” I panicked for half a minute and even resent the screenshot three times. Later, I realized that checking what error the server actually returned before communicating usually worked better than telling people to refresh again and again.

Not new software—just friendlier names for status codes

This approach turns common web request results into Scottish-style expressions: 404 becomes “there’s nothing over there,” while 500 becomes “the server is not doing great today.” It works for customer-support self-checks and incident notes, and also helps nontechnical colleagues understand what happened. In Edinburgh, developer Marcus maintains client websites and translates raw admin-panel messages into plain language before sending them to clients. He says that explaining “the server hit an internal error” saves a round of back-and-forth compared with simply dropping “500 error” on the client.

Replicate it today: 0 yuan, 15 minutes, almost no barrier

No tools to buy and no software to install. Open the “status code guide” in the admin panel of a site I already use, or search for “HTTP status code 404 500,” then paste a comparison table into my support documentation. First step: click “Add bookmark” beside the browser bookmarks bar, create a bookmark called “web outage guide,” and copy in the meanings of the status codes I use most: 200, 301, 403, 404, and 500. The whole setup takes about 15 minutes. The only technical requirement is being able to tell “the page won’t load” from “the page loaded, but I don’t have permission.” I also blamed the client’s network before I understood the issue; it turned out their login had expired.

Whether it’s worth trying depends on my current client volume

If I’m just getting started and do not have many clients yet, I would organize one page with the most common status codes rather than building the whole thing at once. There is no urgency to try it now; saving screenshots of errors is enough for the moment.

If I have one or two stable clients, I would put the comparison table in a shared document and add three prompts: “What happened,” “What we’re doing,” and “When to expect the next update.” That would reduce repeated checks and confirmations.

If I’m scaling up, I would give support, development, and sales one shared set of terms, then regularly check whether any wording is unclear or whether the regional jokes cause misunderstandings. Not everyone needs this little system. If someone on my team already writes incident explanations well, adding another layer just for the sake of process would only add friction.