A
After-Call Work
After-call work is the admin an agent completes immediately after an interaction, such as writing notes, tagging the ticket, and setting up any follow-ups, before taking the next one.
Agentic AI
Agentic AI refers to AI systems that can pursue a goal by planning, making decisions, and taking actions on their own, rather than only responding to a single prompt.
AI Agent vs Chatbot
A chatbot answers questions or guides a conversation, often from a script or trained content. An AI agent goes further: it can reason over a request and take actions to resolve it end to end. Modern AI chatbots increasingly blur the line.
AI Evaluation
AI evaluation is the practice of measuring how accurate, safe, and useful an AI system's outputs are, using a mix of test sets, benchmarks, human review, and live production metrics.
AI Governance
AI governance is the framework of policies, roles, and processes an organisation uses to manage its AI responsibly, keep it accountable, and reduce the risks it can create.
AI Grounding
AI grounding is the practice of tying an AI model's output to a trusted, verifiable source or context, so its answers are based on real information rather than invented.
AI Guardrails
AI guardrails are the constraints, on what an AI system reads, says, and does, that keep it safe, accurate, and on-topic, and route it to a human when it reaches the edge of what it should handle.
AI Hallucination
An AI hallucination is when an AI model produces information that sounds confident and plausible but is factually wrong or entirely made up.
AI Observability
AI observability is the practice of monitoring an AI system in production, through logs, metrics, and traces, so you can understand how it behaves and catch issues like drift or hallucination early.
Auto QA
Auto QA is the use of AI to automatically evaluate and score support interactions for quality, letting teams review every conversation rather than a small manual sample.
Auto-Triage
Auto-triage is the automated reading, categorising, and prioritising of incoming support requests, so each one is understood and labelled before, or as, it is routed to an agent.
Average Handle Time
Average handle time is the average total time an agent spends on a single interaction from start to finish, including talk time, hold time, and the after-call work that follows.
Average Resolution Time
Average resolution time (ART) is the mean time it takes to fully resolve a support request, measured from when it is first received to when it is marked as solved.
Average Response Time
Average response time is the average time it takes to reply across every message in a support conversation, not just the opening one, capturing the pace of the whole back-and-forth.
Average Speed of Answer
Average speed of answer is the average time a contact waits in the queue before an agent picks up, measured only across the contacts that were answered.
Average Wait Time
Average wait time is the average length of time customers wait before being helped in a support queue, measured across the contacts who joined the line.
C
Chatbot vs Conversational AI
A chatbot is any software that chats with users, including simple rule-based ones. Conversational AI is the technology, using natural language processing and machine learning, that lets a chatbot understand and respond in natural language.
Confidence Score
A confidence score is a number, often between 0 and 1, that represents how certain an AI model is that its output, such as a predicted intent or answer, is correct.
Consent Management
Consent management is the process of capturing, recording, and honouring the permissions people give for how their personal data is collected and used, and letting them withdraw it.
Containment Rate
Containment rate is the percentage of conversations that a chatbot or self-service channel resolves on its own, without transferring the customer to a human agent.
Conversational AI
Conversational AI is the set of technologies, including natural language processing and machine learning, that let software understand and respond to human language in a natural, back-and-forth way.
Customer Centricity
Customer centricity is a way of running a business that puts the customer's needs at the heart of every decision, from product and pricing to support, rather than treating it as one team's job.
Customer Delight
Customer delight is the act of exceeding a customer's expectations to create a positive emotional response, going beyond merely satisfying them to leave them pleasantly surprised.
Customer Effort Score
Customer effort score (CES) is a survey metric that measures how easy, or how hard, it was for a customer to get their issue resolved, usually captured by a single question rated on an agreement scale just after an interaction.
Customer Feedback Loop
A customer feedback loop is the ongoing cycle of collecting customer feedback, analysing it, acting on it, and closing the loop by telling customers what changed as a result.
Customer Friction
Customer friction is anything that makes it harder for a customer to get what they need, from confusing navigation to repeated identity checks, adding effort to an interaction that should feel simple.
Customer Self-Service
Customer self-service lets people find answers and resolve issues on their own, through resources like a help centre, FAQs, or an AI chatbot, without contacting a support agent.
D
Data Anonymisation
Data anonymisation is the process of irreversibly removing or altering identifying details from data, so that it can no longer be traced back to a specific individual.
Data Residency
Data residency is the physical or geographic location where an organisation's data is stored and processed, which determines the laws and regulations that apply to it.
Data Retention Policy
A data retention policy is a documented set of rules governing how long an organisation keeps each type of data, and when and how that data is securely deleted.
Data Subject Access Request
A data subject access request (DSAR) is an individual's right, under laws such as the GDPR, to ask an organisation what personal data it holds about them and to receive a copy.
Deflection Rate
Deflection rate is the percentage of support queries resolved by self-service, such as help articles, FAQs, or an AI chatbot, without reaching a human agent.
Dialogue Management
Dialogue management is the part of a conversational AI system that keeps track of the conversation's state and decides what the assistant should do or say next at each turn.
Digital Customer Service
Digital customer service is support delivered through digital channels, such as live chat, messaging apps, email, social media, and self-service, rather than over the phone.
E
Entity Extraction
Entity extraction is the process of identifying and pulling out specific pieces of structured information, such as names, dates, products, or order numbers, from unstructured, free-form text.
Escalation Management
Escalation management is the process of moving a support issue to someone better placed to resolve it, such as a specialist team or a more senior agent, when the first responder cannot.
Escalation Matrix
An escalation matrix is a documented map of who an issue should go to, when it should be escalated, and how far up it should travel, based on its priority and type.
Escalation Rate
Escalation rate is the percentage of support interactions passed up to a higher tier of support, such as a senior agent, a specialist team, or a manager, because the first responder could not resolve them.
F
First Contact Resolution
First contact resolution is the share of support issues fully resolved during the first interaction, so the customer gets a complete answer without any follow-up or repeat contact.
First Response Time
First response time is how long a customer waits between submitting a support request and receiving the first reply, whether from a human agent or an automated system.
I
Intelligent Routing
Intelligent routing is a form of ticket routing that uses AI to read the intent, priority, sentiment, and context of each request and assign it, rather than relying on fixed keyword rules alone.
Intent Recognition
Intent recognition is how an AI system works out what a user is trying to achieve from their message, such as tracking an order or cancelling a plan, so it can respond appropriately.
K
Knowledge-Centered Service
Knowledge-Centered Service (KCS) is a support methodology where teams capture and reuse knowledge as a natural by-product of solving customer issues, so answers are written down once and improved through use.
L
Large Language Model
A large language model (LLM) is an AI system trained on vast amounts of text to understand and generate human language by predicting likely sequences of words.
M
Model Drift
Model drift is the gradual decline in an AI model's accuracy over time, as the real-world data it meets in production drifts away from the patterns it originally learned.
Multi-Turn Conversation
A multi-turn conversation is a back-and-forth exchange in which an AI keeps track of context from earlier turns, so each reply builds on what has already been said rather than treating every message in isolation.
N
Natural Language Processing
Natural language processing (NLP) is the field of artificial intelligence concerned with enabling computers to understand, interpret, and generate human language.
O
Omnichannel Routing
Omnichannel routing is a form of ticket routing that channels support requests from every source, such as chat, email, phone, and social, into one unified queue governed by a single set of rules.
P
PII Redaction
PII redaction is the practice of finding and masking or removing personally identifiable information, such as names, emails, or card numbers, from a document or dataset.
Proactive Outreach
Proactive outreach is the practice of contacting customers first, before they come to you, to share updates, check in, or head off a problem they haven't raised yet.
Proactive Support
Proactive support is anticipating customer needs and resolving issues before the customer has to reach out, rather than waiting for them to report a problem.
R
Resolution Rate
Resolution rate is the percentage of support cases that end in a resolution, at any point in the conversation, as opposed to being abandoned, left open, or closed without the customer's issue being solved.
Retrieval-Augmented Generation
Retrieval-augmented generation (RAG) is a technique that first retrieves relevant documents from a trusted source, then has an AI model generate an answer grounded in that retrieved content.
Rule-Based Chatbot
A rule-based chatbot answers using predefined rules, keywords, and decision trees, following a script set by its designers rather than interpreting language on its own.
S
Schedule Adherence
Schedule adherence is a workforce-management metric that measures how closely support agents stick to their rostered schedule, the share of their scheduled time spent on the activities they were rostered to do.
Sentiment Analysis
Sentiment analysis is the automated use of natural language processing to detect the emotional tone, positive, negative, or neutral, expressed in a piece of text.
Service Recovery
Service recovery is how a business responds to a service failure, the steps it takes to fix the problem and rebuild the customer's confidence after something has gone wrong.
Service Recovery Paradox
The service recovery paradox is the finding that a customer whose problem is resolved exceptionally well can end up more loyal than one who never had a problem at all.
Service Request
A service request is a formal ask from a customer or user for a standard, pre-approved service or item, such as gaining access, resetting a password, or getting information.
Skill-Based Routing
Skill-based routing is a form of ticket routing that matches each request to an agent whose specific skills, product knowledge, or language make them best suited to resolve it.
SLA Breach
An SLA breach is when a support team fails to meet a target set out in its service level agreement, such as replying to or resolving an issue within an agreed time.
Slot Filling
Slot filling is the process an AI assistant uses to gather the specific pieces of information (slots) it needs to complete a task, prompting the customer for any that are still missing.
Suggested Replies
Suggested replies are AI-generated response options presented to a support agent, who can send one as is, edit it first, or write their own, speeding up replies while keeping a human in control.
Support Queue
A support queue is the ordered list of customer requests waiting to be picked up and resolved by a support team, usually managed so the most urgent tickets are handled first.
T
Ticket Deflection
Ticket deflection is the practice of resolving customer questions through self-service, such as help articles or an AI chatbot, so they are answered before they ever become a support ticket.
Ticket Priority
Ticket priority is the severity or urgency level assigned to a support request, often on a scale like P1 to P4, that determines the order it is worked in and the SLA target it must meet.
Ticket Routing
Ticket routing is the process of assigning each incoming support request to the agent or team best placed to resolve it, using either manual rules or automated logic.
Tiered Support
Tiered support organises a support team into levels, typically tier 1, 2, and 3, so that each issue is handled by the level with the right expertise, with harder cases moving up the tiers.
Transfer Rate
Transfer rate is the percentage of support interactions passed from the first agent to another agent or team, whether sideways to a peer or up to a higher tier, before the case is resolved.
V
Voice of the Customer
Voice of the Customer (VoC) is the practice of systematically capturing what customers need, expect, and feel about a business, then using that insight to guide decisions.