Support ticket

A support ticket is a structured record of a single customer issue or request, capturing who reported it, what the problem is, which channel it arrived on, its current status, who owns it, and every interaction that happens along the way. It is the atomic unit of every helpdesk and service desk platform.

It is also the primary record support teams use to measure workload, response times, and outcomes. Whether a case is closed by a human agent or an AI agent, it almost always exists as a ticket somewhere in the system.

The ticket format has held up remarkably well as support operations have modernized, in part because it doubles as both an operational record and an audit trail. As AI agents take over a growing share of first-line resolution, the ticket has not disappeared — it has become the object an AI agent reads, updates, and closes rather than a form a human fills out by hand.

What a ticket actually contains

Requester and description. Who reported the issue — customer ID, email, account tier — along with the original message or a summary of the request, which together establish who is affected and what they are asking for.

Channel and status. The channel a ticket came in on — email, chat, phone, SMS, in-app, social, or a messaging platform such as WhatsApp — is recorded alongside a status field that typically moves through new, open, pending, on-hold, solved, and closed.

Priority, ownership, and tags. Priority and severity are usually bucketed as low, normal, high, or urgent, sometimes tied directly to an SLA clock. An assignee or team owns resolution, and tags or categories capture the product area, issue type, or root cause for later analysis.

Interaction history and custom fields. Every reply, internal note, and status change is preserved on the ticket, along with whatever custom fields a given team needs to route or resolve cases efficiently — order number, subscription plan, device model, and similar account-specific details.

How a ticket moves from creation to close

A ticket begins the moment a customer submits a request through any channel, or an agent opens one on the customer's behalf after a phone call or in-person interaction. From there it moves into triage, where automated rules or an AI agent categorize it, assign a priority, and route it to the right owner.

Resolution is where the assignee — human or AI agent — actually works the case: adding notes, communicating with the customer, and taking whatever action closes the underlying issue. The ticket then moves to confirmation, either because the customer confirms the issue is resolved or because it auto-closes after a period of inactivity.

What happens after that is where the ticket earns its keep as a data asset: time-to-resolution, customer effort, CSAT, and root-cause tags all feed back into the feedback loop that shapes how the team and its automation improve over time.

Ticket, incident, and case — different vocabularies, same object

A support ticket is the term used broadly across both B2C and B2B support for a single customer issue tracked through resolution. An incident is the ITSM and ITIL term for an unplanned service interruption, which can affect many customers simultaneously and spawn a large number of individual tickets underneath it.

A case is Salesforce and enterprise CRM vocabulary for essentially the same concept as a ticket, usually wrapped in richer customer context and heavier SLA machinery.

The underlying object is the same in all three; the label just reflects which part of the industry is talking about it, and which system of record a given company has standardized on.

The metrics a ticket makes possible

Because every ticket carries timestamps for creation, first response, and resolution, it is what makes core support metrics measurable at all. First response time tracks how fast someone acknowledges a ticket, and time to resolution tracks how long it takes from creation to solved.

Reopen rate — tickets solved and then reopened — is a strong signal of low-quality first-touch resolution, and it is only visible because the ticket record persists after being marked solved rather than disappearing. Post-resolution survey scores like CSAT and customer effort attach to the same record, closing the loop between how a case was handled and how the customer actually felt about it.

Where support tickets sit in an AI-driven support stack

AI agents have changed how tickets get worked without changing the fact that they exist. A modern agent can triage a ticket by detecting intent and tagging it automatically, resolve it end-to-end by calling APIs to check order status, issue a refund, or update an account, hand off to a human through a warm handoff when its confidence is low, and follow up afterward to confirm the customer is actually satisfied rather than just marking the ticket solved.

Every one of those actions still gets logged against the ticket record, which is what keeps AI-resolved cases auditable in exactly the way human-resolved ones always have been. As support volume shifts toward AI-first resolution, the ticket remains the ledger both humans and agents are ultimately judged against.