Quick answer
A trouble ticket is a record that logs one reported problem, gives it a number and an owner, and tracks it from report to resolution. Its value is accountability: a problem stops being something anyone can mention and becomes something one named person must answer for by a set time.
Formula: SLA violation rate = tickets that missed at least one target ÷ tickets due in the period × 100%
What is a trouble ticket?
A trouble ticket is a record created to log one reported problem, track it and resolve it. It names who reported it, who owns it, how urgent it is, what state it is in and when it is due. Today the same thing is usually called a support ticket, an incident ticket or an issue ticket; "trouble ticket" is the older term from network operations and telecoms, and many help desks still use it.
The idea is older than help desk software. RFC 1297, a 1992 specification for a network operations centre's ticket system, compares a trouble ticket system to a hospital chart: one shared record that coordinates everyone who may need to work on a problem. The point of the comparison is that the record outlasts any one person's memory. A problem mentioned in a hallway belongs to nobody; a problem with a ticket belongs to one named person.
That is the core idea of this entry: a ticket turns "someone should look at this" into a number, an owner and a deadline. The number exists so people can ask about it later, not so you can count it. Two related terms are covered on their own pages: ticket deflection (solving the need before a ticket exists) and ticket escalation (moving a ticket to someone more senior or specialised).
What fields does a trouble ticket need?
A ticket with too many fields is not filled in, and a ticket with too few cannot be worked. RFC 1297 treats a small set as fixed (start time, who opened it, severity, a one-line description) and the rest as optional, including the reporter, the next action and the ticket owner, defined as one person responsible overall. A practical minimum for a small team looks like this:
| Field | Question it answers | Example | Rule |
|---|---|---|---|
| Ticket number | Which one are we talking about? | TT-004217 | Unique, never reused, quoted in every message |
| Reporter | Who has the problem and how do we reach them? | Customer email, visitor code | One contact, verified |
| Description | What is wrong? | "Checkout fails with error 502 on mobile" | One line a stranger understands, details below |
| Priority | What do we do first? | High, normal, low | Set by impact and urgency, not by who shouts |
| Owner | Who is responsible? | A named person | One person, not a team |
| Status | Where is it? | New, open, pending, resolved, closed | One of a short fixed list |
| Next action | What happens next, and who moves? | "Engineer to reproduce by 14:00" | Every open ticket has one |
| Due times | By when must we answer and fix it? | First response 1 hour, resolution 24 hours | Written as targets from the creation time |
Two distinctions hide in this table. Severity versus priority: RFC 1297 allows a customer's severity (how bad it is for them) and the operator's priority (what is worked first) to be separate fields, and keeping them apart stops a loud customer from outranking a worse outage. Owner versus team: "support team" as an owner means nobody knows who is on it.
How does a trouble ticket move from open to closed?
A ticket needs a short, fixed list of statuses and a rule for who may move it between them. A common five-status flow looks like this:
| Status | Meaning | Who moves it | Does the resolution clock run? (typical) |
|---|---|---|---|
| New | Logged, nobody has started | The system, on creation | Yes |
| Open | Someone owns it and is working | The owner | Yes |
| Pending | Waiting for the customer | The owner, after asking a question | Paused |
| Resolved | A fix or answer was given | The owner | Stopped |
| Closed | Confirmed, or no reply within the agreed period | The customer, or a timer | Stopped |
Zendesk's SLA documentation shows how statuses and clocks interact in one real system: requester wait time counts the New, Open and On-hold statuses, pauses in Pending and stops when the ticket is solved. Whatever tool you use, write down the same thing for yours.
Two rules prevent most disputes. A closing rule: for example, "closed when the customer confirms, or three days after Resolved with no reply". Without one, tickets sit indefinitely in statuses nobody reviews, and the open-ticket count only grows. A reopening rule: a customer reply after Resolved should reopen the same ticket for a short window (a day or a few), and after that start a new ticket linked to the old one, so history is not lost.
Why does a ticket have a number?
A ticket number is the one handle that lets a customer and an agent refer to the same problem without describing it again. It should be unique, never reused after closure, short enough to read out on a phone call and included in every message about the ticket, ideally in the email subject, so replies attach to the right record. Do not build meaning into it (customer name, category): meaning changes and the number must not.
What a ticket number is not is a measure. Counting numbers counts tickets, and tickets are not problems. Sample data: a team's tickets for one week run from TT-004101 to TT-004300, which is 200 tickets. During the week 24 were found to be duplicates of earlier ones, such as the same customer writing twice or two customers reporting one outage, and were merged. The problems actually reported were 200 − 24 = 176. Tickets per problem is 200 ÷ 176 = 1.14: 12% of the tickets (24 ÷ 200) were duplicates, so ticket volume overstates the number of problems by 200 ÷ 176 − 1 = 13.6%. That gap moves with how often customers write twice, not with how many things broke. Report tickets and distinct problems separately.
What counts as an SLA violation?
An SLA (service level agreement) target is a promised time for a stage of the ticket, and an SLA violation, also called a breach, is a ticket that misses it. Two targets are common: first response (time from creation to the first public reply from an agent) and resolution (time from creation to a solved status). Zendesk's documentation defines first reply time this way and lets teams choose calendar or business hours for each target.
The violation rate for one target is tickets that missed it ÷ tickets due in the period. To report one number for the desk, count tickets that missed at least one target once, not twice.
Sample data: the desk promises a first response within 1 hour and resolution within 24 hours. In a week it handles 200 tickets.
- 28 tickets missed the first-response target: 28 ÷ 200 = 14.0%, so 86.0% met it.
- 19 tickets missed the resolution target: 19 ÷ 200 = 9.5%, so 90.5% met it.
- 7 tickets missed both.
- Tickets that missed at least one: 28 + 19 − 7 = 40. 40 ÷ 200 = 20.0% SLA violation rate; 160 of 200, or 80.0%, were fully on target.
Adding the two single-target rates (14.0% + 9.5% = 23.5%) would overstate the violation rate by 3.5 points, because the 7 tickets that missed both would count twice.
Calendar hours versus business hours can change the answer completely. A ticket arrives on Friday at 16:30 with a 1-hour first-response target and a 24-hour resolution target, in a desk open 09:00 to 17:00 on weekdays:
| Clock | First response due | Resolution due |
|---|---|---|
| Calendar hours | Friday 17:30 | Saturday 16:30 |
| Business hours | Monday 09:30 | Wednesday 16:30 |
On a calendar clock the desk is breached on a Saturday when nobody works; on a business-hours clock the 24 hours are 24 working hours, three working days. Choose one on purpose and say which it is in the promise.
Trouble ticket or live chat conversation: which do you need?
A chat conversation is enough when the answer is immediate and the visitor is still there. A ticket is needed when the work outlasts the conversation. Use these tests:
- If it needs investigation or another team, open a ticket. The visitor will leave before the answer exists.
- If a deadline was promised, you need a due time and a record of it. A conversation has timestamps, not targets.
- If someone must be accountable by name, the ticket needs an owner field that persists across days.
- If the problem is answered from a written source in one exchange, chat is enough, and the cheaper outcome is to solve it before a conversation exists; see ticket deflection.
When a conversation has to become longer work, the handoff is an escalation, and what should travel with it is covered in ticket escalation. The two measures to watch on the chat side are first response and resolution rate. For how a store sets up the whole desk, see ecommerce help desk.
Common mistakes with trouble tickets
- Opening a ticket for every message. A customer who writes twice about one problem should add to one ticket. In the example, 24 duplicates turned 176 problems into 200 tickets.
- No closing rule. Resolved tickets that nobody closes inflate the open count, and the backlog hides the live problems.
- Making everything high priority. If most tickets are high, priority tells you nothing. Define the levels by impact and urgency.
- A team as owner. The owner is one named person; the team is where the ticket lives.
- Adding two rates to get one. Summing the first-response and resolution violation rates double counts tickets that missed both: 23.5% against the real 20.0%.
- Counting tickets as problems. Tickets measure contact volume; they overcount problems whenever people write twice.
How trouble tickets map onto TapCub live chat
A small team using TapCub asks: can chat replace tickets? For most of what a ticket does, yes; for the number and the deadline, no. TapCub has conversations, not tickets.
- Owner and status are there. In Live chat › Inbox, a conversation sits in Waiting until someone clicks Take and becomes its owner, then lives in Mine and All until Resolve moves it to Resolved. Notes go in Internal note, visible only to agents, and a transfer to a colleague is recorded on the conversation timeline and the audit log.
- Reopening works like the reopening rule. If the visitor writes again within the Reopen window (default 1,440 minutes) the same conversation reopens; after that a new one opens with the earlier conversation shown above it.
- After hours, messages are the nearest thing to a ticket. When nobody is online a visitor can leave an email and a question, and Live chat › Chat settings › Offline messages lists them with the statuses New, Replied and Closed. They are counted separately and are not counted as resolved.
- There is no ticket number, priority field or SLA timer. The inbox list is searched by visitor code or message text, not by a ticket number, and TapCub does not flag SLA breaches. The only built-in timer is the wait time in Waiting, which turns red after 120 seconds.
- Read the nearest numbers in Reports. Live chat › Reports shows resolution rate (resolved ÷ conversations), average first response with a median, satisfaction and a per-agent workload table. Average first response counts only an agent's first reply and leaves out conversations with no agent reply, so a conversation nobody answered does not appear in it; read the median and spot-check conversations that never got an agent reply when you test a first-response promise. TapCub does not compute an SLA violation rate for you; if you need ticket numbers and deadlines, use a ticketing tool and keep TapCub for measuring the chat side.
See also: What is troubleshooting? Definition, six steps and an example.
Sources
- D. Johnson, Merit Network. RFC 1297: NOC Internal Integrated Trouble Ticket System Functional Specification Wishlist, January 1992 — the hospital-chart comparison, fixed and optional ticket fields, severity versus priority, ticket owner. Verified 2026-11-17.
- Zendesk. Creating and managing service level agreements (SLAs) — definitions of first reply time and requester wait time, and the choice of calendar or business hours. Verified 2026-11-17.
Was this explanation clear?
We use your answers to fix and expand this page.
Couldn’t send. Please try again in a minute.
Thanks for the feedback. We’ll use it to improve this page.You’ve already sent feedback on this page. Thanks.

Growth Strategist
Explores acquisition channels, activation, and conversion experiments, turning data insights into actionable growth hypotheses and evaluation methods.