Resources · Glossary

The customer experience glossary

Plain-language definitions for the terms CX, support, product, brand, and marketing teams use every day. Updated weekly.

71 terms · Updated September 2026

Customer experience

How teams measure, diagnose, and improve what it feels like to be a customer.

Customer support

The operational vocabulary of contact centres — queues, handle time, resolution, and deflection.

Marketing

How growth and demand teams describe audiences, attribution, and pipeline.

Product management

The language product teams use to prioritise, scope, and measure what they build.

Brand

How marketing, brand and communications teams define, position, and measure a brand over time.

AI and automation

The terms behind AI agents, automation, and the metrics that prove they work.

Abandonment rate

Abandonment rate is the share of customers who leave a queue, chat, or callback line before ever reaching an agent. Contact centres track it alongside wait time because abandonment climbs sharply once queues pass a customer’s patience threshold, making it an early signal that staffing or deflection is not keeping pace with demand.

Agent assist

Agent assist is a copilot layer inside the agent desktop that drafts replies, retrieves relevant policy, and suggests the next best action while a human stays in control of the conversation. It shortens handle time and reduces variation between agents without removing the person from the interaction.

AI agent

An AI agent is software that resolves a customer request end to end, taking actions in connected systems rather than only answering questions. Unlike a scripted chatbot, an AI agent can look up an order, issue a refund, or update a record, then confirm the completed outcome back to the customer.

Automation rate

Automation rate is the percentage of incoming contacts handled entirely without a human agent. It measures reach rather than quality, so support teams read it alongside containment rate and satisfaction scores to confirm that automated volume is genuinely resolved rather than simply absorbed and later reopened by the customer.

Average handle time

Average handle time is the mean time an agent spends on a single contact, including talk or chat time and after-contact work. Because the definition varies between teams, comparisons are only meaningful when both sides confirm whether wrap-up time is included inside the calculation being reported.

Backlog

Backlog is the set of unresolved tickets carried past their target response window. A growing backlog signals that arrival rate has outpaced capacity, and because ageing tickets attract follow-up contacts, backlog tends to generate additional volume rather than simply waiting quietly to be cleared later.

Bot handoff

Bot handoff is the transfer of a live conversation from an automated agent to a human, ideally with the full transcript and any collected context attached. A clean handoff prevents the customer from repeating information, which is the most common complaint recorded about automated support experiences.

Business rules

Business rules are the conditions that decide how a contact is prioritised, routed, or escalated. Typical rules combine customer tier, channel, detected intent, and elapsed time to determine where a conversation goes and how quickly it must be answered under the applicable service level agreement.

Churn signal

A churn signal is a pattern in support contacts that predicts a customer is about to leave, such as repeated escalations, cancellation questions, or a sharp drop in usage alongside unresolved complaints. Identifying these patterns early lets retention teams intervene while the relationship is still recoverable.

Contact center

A contact centre is the team and tooling that handle inbound customer contacts across every channel, including voice, chat, email, and social. The term replaced call centre as support expanded beyond telephony, and it covers workforce management, quality assurance, and routing as well as the agents themselves.

Containment rate

Containment rate is the share of automated conversations that end without escalation to a human. It is the primary efficiency measure for AI support, but a high rate only indicates success when paired with satisfaction and repeat-contact data confirming customers actually got their issue resolved.

Conversational AI

Conversational AI describes systems that understand and respond in natural language across chat, email, and voice. The category spans intent-based bots and modern language-model agents, and the practical difference between them is whether the system can only reply or can also take action in business systems.

Copilot

A copilot is an assistive AI surface that works alongside a human agent instead of replacing them, suggesting replies, summarising history, and surfacing policy at the moment of need. Copilots are typically measured on handle time and answer consistency rather than on deflection or containment.

CSAT

CSAT, or customer satisfaction score, is a post-contact rating averaged across responses and usually collected with a single question immediately after an interaction closes. Because only a small fraction of customers respond, CSAT reflects the experience of people motivated to answer rather than the whole contact population.

Customer effort score

Customer effort score measures how hard customers say it was to get their issue resolved, usually on a seven-point scale asked immediately after contact. Effort predicts loyalty more reliably than satisfaction, because a customer can be satisfied with an outcome yet exhausted by the process of reaching it.

Data residency

Data residency is the requirement that customer data be stored and processed inside a specific country or region. Regulated industries and public-sector buyers frequently mandate it, which makes deployment location a procurement question rather than a technical detail for any support platform handling personal data.

Deflection

Deflection is steering a contact toward a self-service or automated resolution before it reaches an agent queue. Effective deflection resolves the underlying question; unsuccessful deflection merely delays the contact, so teams measure it by whether the customer returns rather than by how many contacts were diverted.

Detractor

A detractor is a Net Promoter Score respondent scoring between zero and six, treated as an active churn and word-of-mouth risk. Detractor comments are usually the most useful part of an NPS programme, because they name specific failures that aggregate scores cannot surface on their own.

Disposition code

A disposition code is the label an agent or AI applies to a closed contact to describe what actually happened. Disposition data drives volume reporting and root-cause analysis, so inconsistent or overly broad codes quietly destroy a support team’s ability to explain why contacts arrive at all.

Entity extraction

Entity extraction is the process of pulling structured values — an order number, an IBAN, a product name, a date — out of a customer’s free-text message. Reliable extraction is what allows an automated agent to act inside a business system rather than only classify the request.

Escalation

Escalation is moving a contact to a more senior agent, a specialist team, or a human when automation stalls. Escalation paths are defined by business rules, and the rate at which contacts escalate is a direct measure of whether the first line is equipped to resolve them.

Evaluation set

An evaluation set is a fixed sample of real conversations used to score an AI agent’s quality before and after changes. Holding the sample constant is what makes results comparable, so evaluation sets are versioned and refreshed only deliberately when the underlying contact mix genuinely shifts.

Fallback

A fallback is the safe response an AI agent gives when its confidence is too low to act, typically an offer to hand the conversation to a human. Well-designed fallbacks fail early rather than guessing, because a confidently wrong answer costs far more than an admitted gap.

First contact resolution

First contact resolution is the share of issues fully resolved during a customer’s first interaction, with no follow-up contact about the same problem. It correlates strongly with satisfaction and cost, but only when measured over a defined reopen window rather than at ticket close.

Full resolution time

Full resolution time is the total elapsed time from a customer’s first message to a confirmed resolution, including waiting, escalation, and overnight gaps. Unlike response time, it reflects the experience of waiting for an answer rather than the speed of the first reply alone.

Grounding

Grounding is the practice of constraining an AI agent’s answers to approved knowledge sources and live system data rather than to the model’s own training. Grounding is the main defence against hallucination, and it is what makes an AI answer auditable back to a specific document or record.

Guardrails

Guardrails are the policy, tone, and action limits an AI agent is not permitted to cross, such as refusing to issue refunds above a threshold or to discuss topics outside its remit. Guardrails are enforced in the system rather than merely requested in the prompt.

Hallucination

A hallucination is a fluent but unsupported answer produced when a model is not grounded in real data. In customer service the risk is specific: an invented policy, price, or delivery date reads as authoritative to the customer and creates a commitment the business never made.

Human in the loop

Human in the loop describes a review step where a person approves or corrects AI output before it reaches the customer. Teams typically apply it to high-risk intents such as refunds, cancellations, and complaints, then relax it selectively as measured accuracy on those intents proves out.

Hybrid support model

A hybrid support model has AI handling routine, high-volume contacts while human agents own complex, sensitive, or high-value interactions. The design question is not what the technology can handle but where the cost of an error is high enough to justify keeping a person in the conversation.

Intent

An intent is what a customer is actually trying to do, inferred from what they wrote or said rather than from the specific words used. Two very different messages can share an intent, which is why routing and automation are built on intent rather than on keyword matching.

Intent taxonomy

An intent taxonomy is the structured list of intents a support operation recognises and reports on. A good taxonomy is granular enough to drive different actions and stable enough to trend over time; taxonomies that grow uncontrolled become unusable for reporting well before routing.

Interaction analytics

Interaction analytics is the analysis of full conversation transcripts to surface demand drivers, friction, and risk across every contact rather than a sampled few. Because it reads what customers actually said, it identifies emerging issues that structured ticket fields and survey scores are too coarse to reveal.

Journey mapping

Journey mapping lays out every touchpoint a customer passes through, from first awareness to renewal, to find where support demand is created. The value for support teams is diagnostic: most contact volume originates in an upstream product or billing step rather than in the support experience itself.

Knowledge base

A knowledge base is the maintained source of truth that AI agents and human agents both answer from. Its accuracy sets a ceiling on the quality of every automated answer, which makes knowledge maintenance an operational requirement rather than a documentation project owned by nobody in particular.

Knowledge gap

A knowledge gap is a recurring customer question the knowledge base cannot answer, usually visible as repeat escalations or low-confidence fallbacks on the same intent. Systematically closing gaps is the highest-return maintenance activity available to any team running automated support at scale.

Large language model

A large language model is the model class behind modern conversational AI, summarisation, and reply drafting. Trained on broad text corpora, it predicts language fluently but holds no inherent knowledge of a specific business, which is why grounding and retrieval are required before answering customer questions.

Latency

Latency is the delay between a customer’s message and the first useful response. In live chat and voice it directly shapes perceived quality, and it is one of the few AI quality measures a customer notices immediately, well before judging whether the answer was correct.

Live chat

Live chat is real-time text support delivered in-product or on-site, fronted by a human agent, an AI agent, or both in sequence. Its defining constraint is synchrony: the customer is waiting, so response latency and handoff quality matter more than in asynchronous channels.

Macro

A macro is a saved reply or action set that agents apply to a common case, inserting standard wording and performing routine updates in one step. Macros reduce handle time and variation, but stale macros are a frequent source of incorrect policy reaching customers at scale.

Model drift

Model drift is the quiet degradation in AI performance that occurs as products, policies, and customer language change while the model and its knowledge stay fixed. Drift rarely announces itself; it appears as a slow rise in escalations and fallbacks on intents that previously performed well.

Multilingual support

Multilingual support means serving customers in their own language, including handling Arabic and English side by side within a single operation. Beyond translation it requires language-aware routing, quality scoring in each language, and knowledge content that is maintained rather than machine-translated once and forgotten.

Natural language understanding

Natural language understanding is the layer that maps free text to intents and entities so a system can act on what a customer wrote. It answers what the customer wants and which values are involved; generation is a separate step deciding what to say back.

Net promoter score

Net Promoter Score is a likelihood-to-recommend score used as a loyalty benchmark, calculated by subtracting the percentage of detractors from the percentage of promoters. It produces a single comparable number, but the free-text comment attached to each score carries most of the diagnostic value.

Occupancy rate

Occupancy rate is the portion of an agent’s logged-in time spent actively handling contacts rather than waiting for them. Sustained high occupancy predicts burnout and attrition, so workforce planners treat it as a health measure rather than a productivity target to be maximised.

Omnichannel

Omnichannel describes one continuous conversation and history across chat, email, voice, and social, so a customer who switches channel does not have to restart. The distinction from multichannel is continuity: offering many channels is not omnichannel unless context follows the customer between them.

Peak volume

Peak volume is the highest-contact period a staffing plan has to absorb, whether a daily rush hour, a seasonal spike, or an incident-driven surge. Because peaks drive headcount decisions, automation is usually justified on peak absorption rather than on average daily contact volume.

PII redaction

PII redaction is the automatic removal of personally identifiable information from transcripts, logs, and training samples. It lets support conversations be analysed, stored, and used to improve models without carrying customer identifiers into systems that were never scoped to hold them.

Prompt

A prompt is the instruction set that defines how an AI agent behaves in a given task, covering role, tone, permitted actions, and output format. Prompts express intent but do not enforce it, which is why hard limits belong in guardrails rather than in prompt wording.

Quality assurance

Quality assurance is the structured review of conversations against a scorecard, applied to human agents and AI agents alike. Traditional QA samples a handful of conversations per agent each month; automated QA scores every conversation, which changes the exercise from spot-checking into actual measurement.

Queue

A queue is the ordered set of contacts waiting to be handled, usually segmented by channel, language, skill, or customer tier. Queue design determines who waits and for how long, making it one of the most consequential and least visible decisions in a support operation.

Resolution rate

Resolution rate is the share of contacts closed with the customer’s issue actually solved, as distinct from contacts merely closed. The gap between the two is where reopened tickets and repeat contacts live, so the measure is only trustworthy when reopens are tracked against it.

Response time

Response time is how long a customer waits for a reply once they have made contact. Teams report it as a first-response figure and as an average across subsequent replies, because a fast opening reply followed by long silences produces a poor experience despite good headline numbers.

Retrieval-augmented generation

Retrieval-augmented generation grounds answers by retrieving an organisation’s own documents at the moment of answering and passing them to the model as context. It keeps responses current without retraining, and it makes each answer traceable to the specific source the system actually consulted.

Routing

Routing is deciding which agent, team, or AI agent receives a contact, based on intent, language, skill, customer tier, and availability. Accurate routing is the cheapest available reduction in handle time, because a misrouted contact pays the cost of being read and reassigned twice.

Self-service

Self-service is help a customer can complete without contacting support at all, such as a help centre article, an account action, or a guided flow. Its success is measured by task completion rather than page views, since a read article that fails to resolve still generates a contact.

Sentiment analysis

Sentiment analysis scores the emotional tone of a conversation to flag risk and measure experience without waiting for a survey response. Applied across every contact rather than a sampled few, it covers the large majority of customers who never complete a satisfaction survey at all.

Service level agreement

A service level agreement sets the committed response and resolution targets for a customer, tier, or contract, often with financial penalties attached. Support operations design queues, routing, and escalation around SLA thresholds, which makes the agreement a structural constraint rather than a reporting line.

Shrinkage

Shrinkage is the share of paid agent time unavailable for handling contacts, covering breaks, training, meetings, absence, and system downtime. Workforce plans that ignore shrinkage systematically understaff, because scheduled headcount and available headcount differ by a predictable and often substantial margin.

Ticket

A ticket is the record of a single customer issue and everything done about it, including messages, internal notes, actions taken, and final disposition. The ticket is the unit most support reporting counts, which is why inconsistent ticket creation quietly distorts every downstream volume metric.

Time to first response

Time to first response is how long a customer waits before any human or AI reply arrives. It is the most visible service measure to the customer and the easiest to improve with automation, which is why it should always be read alongside resolution quality.

Topic clustering

Topic clustering groups conversations by theme to size the drivers behind contact volume, without requiring an analyst to define categories in advance. Because clusters emerge from what customers actually said, the method surfaces new issues that a fixed intent taxonomy would file under a generic other.

Triage

Triage is sorting incoming contacts by intent, urgency, and value before they are worked, so the most consequential conversations reach the right place first. Automated triage now performs the classification step, leaving prioritisation rules and exception handling as the main human contribution.

Unified inbox

A unified inbox is one workspace where conversations from every channel land, so agents work a single prioritised list instead of switching between tools. It reduces context-switching cost and makes cross-channel history visible at the moment an agent is drafting a reply.

Utterance

An utterance is a single thing a customer said, exactly as the model receives it, before any cleaning or interpretation. Utterances are the raw material for intent classification and evaluation sets, and preserving them verbatim is what allows AI errors to be reproduced and diagnosed later.

Verbatim

A verbatim is a customer’s exact words, kept unedited as evidence in voice-of-customer reporting. Verbatims carry the specificity that aggregate scores strip out, which is why they persuade product and executive audiences when a percentage change on a dashboard cannot do so on its own.

Voice AI

Voice AI is automation that handles spoken conversations using speech recognition, language understanding, and speech synthesis. It carries constraints text automation does not: recognition errors on names and numbers, sensitivity to accents and dialects, and a much lower customer tolerance for response latency.

Voice of customer

Voice of customer is the discipline of turning reviews, support tickets, surveys, and conversations into operational decisions. The distinguishing feature of a working programme is the closed loop: findings reach the team that can fix the cause, and the resulting change is measured against contact volume.

Workflow automation

Workflow automation chains system actions — issue a refund, reissue a document, update a record — so a resolution completes itself without an agent performing each step. It separates an AI agent that resolves requests from a chatbot that can only describe how a request would be resolved.

Wrap-up time

Wrap-up time is the after-contact work an agent performs before taking the next conversation, including notes, disposition coding, and system updates. It is frequently excluded from handle time reporting, which is the most common reason two teams’ average handle time figures cannot be compared.

Zero-touch resolution

Zero-touch resolution is an issue resolved end to end with no human involvement and no follow-up contact from the customer. It is a stricter measure than containment, because it requires the customer not to return, and it is the cleanest available proof that automation genuinely worked.

Support teams are no longer a cost centre.

See how AI Support Agent, Agent Assist, and VoC Analyst work on your own ticket volume.

Talk to an expert

System normal

MSA

Cookie Notice

System normal

MSA

Cookie Notice

System normal

MSA

Cookie Notice