Customer support glossary

Plain-language definitions of the support terms, metrics, and workflows small teams run into, written because clear answers were harder to find than they should be. Most are defined here on this page; the ones that need more room have a page of their own.

Metrics

CSAT (Customer satisfaction score)CSAT (customer satisfaction score) is the percentage of customers who rate a support interaction as satisfactory.
NPS (Net Promoter Score)NPS (Net Promoter Score) is a measure of customer loyalty based on a single question: how likely is a customer to recommend your product or service?
SLA (Service level agreement)An SLA (service level agreement) is a commitment to respond to or resolve customer issues within a defined timeframe.

Backlog

A support backlog is the accumulation of unresolved tickets that have not yet been answered or closed.

A backlog forms when more tickets arrive than the team can resolve. It represents deferred work: every ticket in the backlog is a customer still waiting. Some backlog is normal, but a growing backlog signals that demand is outpacing capacity.

Teams track backlog size as a snapshot metric ("how many open tickets do we have right now?") and alongside inflow vs. resolution rate to see whether the backlog is shrinking or growing. A consistently flat or growing backlog is an early warning sign.

Clearing a backlog typically requires either more capacity, faster resolution per ticket, or reducing inflow through better self-service. Short-term surges are often handled by temporarily shifting team priorities.

CES (Customer Effort Score)

CES (Customer Effort Score) measures how easy it was for a customer to get their issue resolved, usually as an agree/disagree rating where a higher score means less effort.

CES surveys ask customers to rate the effort required to resolve their problem, not whether they were satisfied but how hard they had to work. Research suggests that reducing customer effort is one of the strongest drivers of loyalty: low-effort experiences reduce churn even more reliably than delightful ones.

The common modern format (CES 2.0) sends a single statement after a resolved interaction, "The company made it easy to handle my issue," on an agree-to-disagree scale, where stronger agreement means less effort. Scores are averaged across responses.

CES is especially useful for identifying friction in support processes: confusing self-service, too many handoffs, or having to repeat information. It pairs well with FCR: a high FCR alongside a high effort score means customers resolved their issue easily, in one go.

First contact resolution (FCR)

First contact resolution (FCR) is the percentage of support requests resolved in a single interaction, without the customer needing to follow up.

FCR measures whether your team solves problems completely the first time. A high FCR means customers get a full answer in one exchange; a low FCR means multiple back-and-forth messages or re-opened tickets.

It is one of the most customer-friendly metrics: customers prefer getting a complete answer immediately over fast replies that require follow-ups. FCR and CSAT tend to move together: higher FCR usually produces higher satisfaction.

Improving FCR means investing in complete, thorough first replies. That includes giving agents access to full context, good knowledge resources, and enough time to write a proper response rather than a quick acknowledgment.

First response time (FRT)

First response time (FRT) is how long a customer waits between sending a message and getting the first human reply.

First response time measures the gap between when a customer writes in and when your team sends its first real reply, not an autoresponder. It is the metric customers feel most, because a fast first reply signals that someone is on it, even if the full resolution takes longer.

Teams usually report FRT as a median or average over a period, and often only count business hours so overnight gaps do not distort the number. Good targets vary by channel: live chat is expected in minutes, email in hours.

The two levers that move FRT are staffing and how quickly each reply gets written. AI-drafted replies cut the writing time, so the same team can answer faster without adding headcount.

Resolution rate

Resolution rate is the percentage of incoming support tickets that are fully resolved within a given period.

Resolution rate compares tickets resolved to tickets received. A team resolving as many tickets as it receives has a resolution rate at or above 100%. It is keeping pace with demand and clearing any backlog. A rate below 100% means the backlog is growing.

The metric is most useful when tracked alongside ticket volume and backlog size. High resolution rate with low volume is easier to achieve; the harder goal is maintaining a strong rate as volume grows.

Resolution rate can also be broken down by ticket type, agent, or time period to find bottlenecks. Some ticket categories may resolve quickly; others may consistently drag.

SLA breach

An SLA breach occurs when a support ticket is not responded to or resolved within the time limit set by the team's service level agreement.

An SLA breach means a ticket has exceeded the committed response or resolution window without being handled. Most support platforms highlight approaching or breached tickets with warnings so the team can prioritize them.

Breaches can result from unexpected volume spikes, understaffing, tickets falling through routing gaps, or no escalation process for tickets that stall. Tracking breach rate by ticket type or time period helps surface patterns.

Some teams distinguish a "first response SLA breach" (the first reply was too slow) from a "resolution SLA breach" (the full resolution took too long), since each calls for a different fix.

Ticket deflection

Ticket deflection is the practice of answering customer questions through self-service resources so that a support ticket never needs to be created.

Deflection is a form of demand reduction: instead of handling the same question repeatedly, you give customers a way to find the answer themselves through a knowledge base, FAQ page, chatbot, or contextual in-product help.

Deflection rate is typically estimated rather than measured directly, since you cannot count tickets that were never opened. Proxies include knowledge base views before contact, chatbot resolution rates, or comparing contact rates before and after publishing new help articles.

Deflection and support quality can conflict if taken too far. Customers who cannot reach a human when they genuinely need one will become frustrated. The goal is deflecting routine questions, not blocking access to support.

Ticket volume

Ticket volume is the total count of support requests received in a given period, used to track demand on the support team.

Ticket volume is the raw demand signal. Teams track it daily, weekly, and monthly to spot growth trends, seasonal patterns, and the impact of product changes. A spike in volume after a release often means something broke or changed unexpectedly.

Volume alone does not tell you how well the team is performing, but it sets the context for other metrics. A team with high CSAT scores handling low volume might struggle under a volume surge. Staffing decisions often start with volume projections.

Volume is often broken down by channel (email, chat, social), by issue type, or by product area to identify where the most tickets originate. High volume in one category is a signal to invest in self-service or product improvements there.

Related: Backlog, Ticket deflection, Queue

Time to resolution (TTR)

Time to resolution (TTR) is the total time from when a customer first contacts support to when their issue is fully resolved.

TTR covers the whole lifecycle of a support case, from the opening message to the final "your issue is closed" moment. Unlike first response time, it captures back-and-forth exchanges, waiting on the customer, and any internal escalation steps.

Teams typically report TTR as a median to reduce skew from a handful of complex, long-running tickets. Like FRT, it is often measured in business hours only so weekends and time-zone gaps do not inflate the number.

Reducing TTR usually requires both faster individual replies and fewer round trips. Clear first replies that anticipate follow-up questions cut round trips; AI drafts help write those replies faster.

Workflow

Auto-reply

An auto-reply is an automated message sent immediately when a ticket is received, typically to acknowledge receipt and set expectations about response time.

An auto-reply is not a resolution. It is a placeholder that tells the customer their message arrived and when to expect a real response. A typical auto-reply reads: "Thanks for contacting us! We'll get back to you within X business hours."

Auto-replies are almost universal in email support because the alternative, silence, creates anxiety. A well-written auto-reply sets accurate expectations, gives a reference number if the platform supports it, and may point to self-service resources for common questions.

Auto-replies should not be counted as first response time. FRT should measure the first substantive human reply, not the automated acknowledgment.

Business hours

Business hours are the scheduled times during which a support team is staffed and actively responding to tickets.

Business hours define when SLA clocks run. If a team supports 9am–5pm Monday to Friday, a ticket received Friday at 4pm has its SLA timer pause at 5pm and resume Monday at 9am. Without this adjustment, weekend and overnight gaps distort response time metrics.

Teams publish their business hours so customers know when to expect a reply. Many teams also post them in auto-replies and on contact pages to manage expectations proactively.

As teams grow, business hours can expand to cover more time zones or a follow-the-sun model, where different regional teams hand off at shift boundaries, without each individual team working nights.

Escalation

Escalation is the process of transferring a customer issue to a more senior agent, specialist, or different team when the original responder cannot fully resolve it.

Escalation happens when a ticket exceeds a frontline agent's authority, expertise, or the SLA window. A billing dispute might escalate to a billing specialist; a complex technical bug might go to engineering; a high-value customer complaint might go to a manager.

Well-defined escalation paths prevent tickets from stalling. Most support setups define at least two tiers: frontline (Tier 1) and specialist (Tier 2), with clear criteria for when to escalate rather than attempt further resolution.

Escalation adds handoff steps and usually increases time to resolution. That is acceptable when the issue genuinely requires deeper expertise, but excessive escalation can signal that frontline agents lack the resources or authority to resolve common issues themselves.

Related: Support tiers, SLA breach, Triage

Omnichannel support

Omnichannel support means handling customer contacts from multiple channels (email, chat, phone, social media) in a unified workspace so agents have full context regardless of how a customer reached out.

The "omni" distinguishes it from multichannel, where each channel is managed separately. In a true omnichannel setup, a customer who emails and then follows up via chat is recognized as the same person, and the agent sees the full history from both channels.

Omnichannel support reduces the frustration of customers having to repeat themselves and lets agents make more informed replies. It is more complex to set up (each channel needs an integration) and typically requires a dedicated help desk platform rather than a basic shared inbox.

For small teams starting out, handling one or two channels well is usually more important than spanning every channel poorly. Omnichannel becomes more valuable as the team grows and customers regularly switch channels.

Queue

A support queue is the ordered list of open tickets waiting for a response or action from the support team.

The queue is the work surface for support agents. It shows all open, unresolved conversations and is typically ordered by priority, SLA deadline, or arrival time. Agents work through the queue from top to bottom, picking up or being assigned tickets.

Queues can be shared (everyone works the same list) or split (agents have personal queues or queues by specialization). Most teams with more than two or three agents benefit from some queue segmentation to avoid agents working the same ticket simultaneously.

A growing queue is an early sign that volume is outpacing capacity. Queue management (triage, routing rules, and auto-assignment) helps ensure the right tickets reach the right agents without manual sorting.

Related: Triage, Backlog, Ticket volume

Self-service support

Self-service support lets customers find answers and resolve issues on their own through a knowledge base, FAQ, chatbot, or community forum, without contacting a support agent.

Self-service is a deflection strategy: the goal is to answer questions before they become tickets. Well-maintained self-service resources can handle the highest-volume, most-repetitive questions so agents can focus on complex or sensitive issues.

Common self-service formats include knowledge bases with searchable articles, FAQ pages, guided troubleshooters, community forums, and in-app contextual help. Each has different maintenance costs and suits different types of questions.

The challenge is keeping self-service resources current. Outdated or incomplete articles frustrate customers and may drive them to open a ticket anyway. Teams that maintain good self-service assets typically use ticket data (the most common questions) to decide what to document next.

Shared inbox

A shared inbox is a single mailbox several teammates work from together, so customer emails are handled as a team instead of in one person's inbox.

A shared inbox turns an address like support@ into a collaborative workspace: everyone sees the same conversations, can see who is replying, and can assign, comment, and track status without forwarding or BCC chains.

It is the lightest-weight step up from a personal inbox and the core of most small-team support setups, simpler than a full ticketing system but with the visibility a solo mailbox lacks.

Support ticket

A support ticket is a record of a single customer issue or request, tracked from the moment it is opened until it is resolved.

When a customer contacts support, a ticket is created, either automatically by the platform or manually by an agent. The ticket holds the conversation history, any internal notes, the assigned agent, priority level, and status.

Tickets create accountability: every request has an owner and a status, so nothing falls through the cracks in a busy inbox. They also generate the data behind support metrics like first response time, time to resolution, and ticket volume.

The word "ticket" comes from IT helpdesk culture where paper tickets were used to track requests. Today it is used broadly for any tracked support case, even if the channel is email or chat.

Support tiers

Support tiers are numbered levels of support, typically Tier 1, Tier 2, and Tier 3, that reflect increasing specialization and authority.

Tier 1 is frontline support: agents who handle the majority of incoming requests using standard knowledge and saved replies. Tier 2 is specialist support for issues requiring deeper product knowledge or elevated permissions. Tier 3 typically involves engineers or subject-matter experts for complex technical problems.

Not every team needs three tiers. Many small teams run a single tier where all agents handle everything, escalating only in extreme cases. Tiers become more useful as volume and product complexity grow.

Support tiers also sometimes refer to service levels offered to customers (for example, basic, professional, and enterprise support plans), each with different SLA commitments and access to higher-tier agents.

Related: Escalation, Triage, Help desk

Ticket tagging

Ticket tagging is the practice of adding labels or categories to support tickets so they can be filtered, routed, and reported on.

Tags are free-form labels applied to tickets: "billing," "bug-report," "refund," "urgent," "account-access." They let teams slice ticket volume by topic, see which issues are driving the most contacts, and route tagged tickets automatically to the right queue or agent.

Tagging can be done manually by agents or automatically using keyword rules. Consistent tagging is what makes support data useful for trend analysis and for identifying where self-service or product changes would have the most impact.

The value of a tagging system depends entirely on discipline in applying it. Inconsistent or overly granular tag libraries quickly become noisy and lose their analytical value.

Triage

Triage is the process of reviewing incoming support tickets to assign priority, route them to the right team, and decide what needs attention first.

The term comes from emergency medicine, where patients are sorted by urgency. In support, triage means looking at new tickets and making quick decisions: Is this urgent? Does it need a specialist? Is it a billing question or a technical one? Can it be answered with a saved reply?

Triage can be manual (an agent reviews the queue every hour and assigns tickets) or automated using rules based on keywords, customer tier, or channel. Most help desks support auto-assignment rules that do the routing automatically.

Good triage keeps the most important tickets visible and reduces the risk of high-priority issues sitting unnoticed in a busy queue. It is the decision-making step between a ticket arriving and a reply being written.

Related: Queue, Escalation, Ticket tagging

Channels & tools

AI-assisted support

AI-assisted support uses artificial intelligence to help support agents write replies, categorize tickets, or suggest answers, with a human reviewing and sending each response.

The defining feature of AI-assisted support is the human in the loop. Unlike fully automated chatbots, AI-assisted tools draft or suggest responses that a human agent reviews, edits, and sends. This preserves the accuracy and accountability of human replies while cutting the time it takes to write them.

Common applications include AI reply drafts, smart categorization and routing of incoming tickets, suggested knowledge base articles during reply composition, and automatic ticket summarization for handoffs.

AI-assisted support is distinct from AI-only or bot-first support. Teams with reputation-sensitive customer relationships (professional services, high-value accounts, complex products) often prefer the assisted model specifically because every sent reply has human sign-off.

Concepts

Customer support vs. customer serviceCustomer support and customer service are related but distinct: customer support focuses on solving specific product problems, while customer service covers the broader relationship and experience a company delivers to customers.