Journal
Fields we need before a ticketing extract will make a chart
A usable extract is usually a CSV. Fancy workbooks with merged header rows cost more time than they save. One header row, one ticket per row, UTF-8 if anyone has typed a name with an accent.
The columns we ask for first are: ticket identifier, opened timestamp, first-response timestamp if you store it, resolved timestamp, current status, queue or assignment group, priority, channel (phone, email, portal, chat), and whether the record is a child of a major incident. If you have a ‘clock paused’ flag, send that too.
We do not need the description field. In fact, please strip it. Descriptions are where people paste screenshots as text, card numbers, and arguments. They also make files too large to email. Identifiers plus timestamps are enough to draw volume and SLA charts.
Queue names that changed in March will appear as two queues unless you send a mapping. We will not guess that ‘IT Support – Staff’ is the same as ‘Staff IT’. Put the mapping in a second sheet if you must; do not silently rename history.
Time zones trip desks that take work from more than one country, and desks whose servers were set to UTC while the wall clock in Kingswood said BST. State the zone of the timestamps in the email that accompanies the file. If you cannot, we treat them as UK local and write that assumption on page one.
If your system will only export ‘this week’, say so early. A quarterly pack then needs four files and a promise that the identifier space did not reset.