Ticket Triage: Priority Matrix and a Daily Routine
Triage turns a pile of requests into an ordered queue with owners. It is quick when the rules are written down and slow when every request is a negotiation.
Triage is the first decision made about a new request. It answers four questions: what is this, how urgent is it, who should handle it and does it need escalating? The word comes from medicine, where the aim is to treat those in greatest need first. The same logic applies to a queue of tickets.
Step 1: Classify
- Request: something expected and standard, such as access or equipment.
- Incident: something that should work does not. If several people report the same thing, treat it as one incident.
- Question: information is needed. Answer it, and consider an article.
- Change: a modification is requested. Route it to the change process.
Step 2: Set the priority from impact and urgency
A common method sets priority from two factors. Impact is how many people or how much of the business is affected. Urgency is how soon the effect becomes serious. A matrix combines them, so that high impact with high urgency is critical, and low impact with low urgency is low. The method keeps priority from becoming a function of who shouts loudest.
- High impact and high urgency: critical.
- High impact and low urgency, or low impact and high urgency: high or medium, according to your rules.
- Low impact and low urgency: low.
Step 3: Assign
Every request needs one named owner. A request assigned to a team belongs to nobody in particular. Route by category to a person or a rota, and let the owner reassign with a note if wrong. Use automation to apply a label and notify the right group when a request arrives, so that routing is not manual.
Step 4: Decide the next date
Give each request a date: when it will next be looked at or when it is due. A request without a date is forgotten. In Atlas, the inbox triage view shows tasks with no assignee, no date or no priority one at a time, and a key press sets the due date to today or the next Friday, snoozes it for two weeks or archives it.
A daily routine
- Triage the new queue at fixed times, for example at the start of the day and after lunch.
- Work critical items first and keep requesters updated.
- Review anything past its due date before closing the day.
- Once a week, read the oldest open requests aloud and decide on each.
Mistakes
- Triaging in one person's head, so cover fails when they are away.
- Raising priority because a senior person asked, without a reason recorded.
- Leaving requests unclassified, so incidents sit in the same queue as questions.