A request that lands in the wrong queue rarely announces itself. It sits there looking routine until somebody calls back, or a supervisor finds a ticket that should have reached another shop a week ago. This is a troubleshooting guide for the office staff who correct these cases: five symptoms, each with a likely cause and a fix. Every scenario is illustrative and invented for this article. None comes from a real department or caller.
Where an agent fits is narrow on purpose. It reads the incoming request, suggests a category and a queue, and flags anything odd. A dispatcher or supervisor reads the suggestion and decides. Nothing is assigned to a crew until a person says so.
Before changing anything, find out how often each symptom actually happens. A few days of tallies on a clipboard will tell you whether you have a recurring wording problem, a form problem or simply a busy week. The fixes below are cheap, so try the one that matches your tally and watch whether the count drops.
Symptom 1: A water complaint sits in the streets queue
Likely cause. The caller said “road” and “flooded,” so the intake form guessed streets. A main break and a blocked storm drain use the same words but belong to different shops.
Fix. Add one clarifying question to the intake script: is water coming up from the ground, or pooling after rain? Have the agent flag any request that mentions both a road and running water. The dispatcher moves the ticket and writes a note on why, so the next similar case is easier to spot.
Symptom 2: The same missed pickup appears three times
Likely cause. The phone line, the web form and a shared mailbox each created a ticket, and the address was spelled three different ways.
Fix. Ask the agent to propose likely duplicates by address and date and to list them in groups. A coordinator looks at each group and merges what truly matches. Automatic merging is the step to avoid, because two neighbors on the same street may both have a real problem.
Symptom 3: A billing question is filed as a service request
Likely cause. A message such as “my bill jumped and I think there is a leak” mentions both a bill and a repair, so one label cannot hold it.
Fix. Split it into two items: a service request for the possible leak and a billing inquiry for the billing desk. The agent can draft an acknowledgment for each, and a staff member reads and sends them. Any decision about the account stays with the billing supervisor.
Symptom 4: An urgent request waits behind routine ones
Likely cause. The queue is sorted by arrival time, and the alarming detail sits in the third sentence of a long message.
Fix. Have your supervisors write the list of phrases that should raise a flag, and give the agent that list. It marks requests for a human to look at first; it does not rank crews or send anyone out. A supervisor checks the flagged items at set times through the shift.
Symptom 5: A request keeps bouncing between two shops
Likely cause. Each shop reads the description and decides the other one owns it. A sidewalk heaved by a tree root is a fair example: streets, parks and water can each make a case that it is not theirs.
Fix. Write down, in a short shared page, who takes the first look when ownership is unclear. Let the agent attach that rule to the ticket as a reminder, along with the earlier hand-offs, so the third person to open it sees the history. The supervisors named on the page settle the dispute, not the software.
One habit that prevents most of these
Once a week, pull ten tickets that ended up in the wrong place and ask two questions of each: what did the request actually say, and what did our form hear? Edit the script or the flag list to close that gap. Small changes made weekly beat a large cleanup made once a year, and the ten tickets double as plain material for training a new dispatcher on how requests really arrive.
If two of these symptoms sound familiar, bring us the queue that causes them. We will shape an agent around that queue, and your staff read every suggestion first.