[ Glossary ]
The vocabulary of support, defined without padding
57 terms that come up daily in any support team and that almost nobody stops to define. Each one with its definition, what it gets confused with, and why it matters when what you sell is software.
14 términos
Support concepts
El vocabulario de andar por casa de cualquier equipo de soporte: lo que hay detrás de las palabras que se usan a diario sin definirlas nunca.
Ticket
A ticket is the record of a customer question inside a support tool. It groups every message in that conversation under one identifier and adds what is needed to manage it: status, owner, priority, tags and the history of what has happened.
SLA — Service level agreement
A service level agreement is the written commitment to respond within set deadlines. It fixes maximum times — usually first response and resolution — defines the hours in which they count, and normally states what happens when they are missed.
Escalation
Escalation is handing a conversation to somebody with more knowledge or more permissions to resolve it. It can go from one support tier to another, from an automated agent to a person, or from support to engineering — and in every case whoever receives it has not seen the case until that moment.
Knowledge base
A knowledge base is the organised set of articles documenting how a product works and how to resolve what gets asked often. It serves two audiences at once: customers looking up the answer themselves, and the team using it as the source for their replies.
Help centre
A help centre is the public site where a company publishes its support documentation so customers can look things up themselves. It combines the articles with a search box, category navigation and, normally, the route to get in touch when the answer is not there.
Self-service
Self-service is everything that lets a customer meet their need without talking to anybody: searching the documentation, checking the state of something in their dashboard, changing a setting or downloading an invoice. It is the cheapest form of support and, when it works, the one customers prefer.
Omnichannel
Omnichannel describes support where the customer can write through whichever channel they prefer — chat, email, WhatsApp, phone — and the conversation stays a single thread: the history travels with them and they repeat nothing when switching channel.
Macro
A macro is a saved reply inserted into a conversation with one click, usually with fields that fill themselves in — the customer name, the order number. It exists so nobody rewrites from scratch what has already been answered many times.
Shared inbox
A shared inbox is a mailbox the whole team can access, where every conversation has an assigned owner, a status and a history visible to everyone. It replaces forwarding emails between colleagues and the generic address one person quietly runs.
Support queue
A queue is a set of conversations waiting to be handled by one specific group of people. Splitting the inbox into several queues — by channel, by language, by product or by customer type — is the usual way of dividing the work once a single list stops working.
Routing
Routing is the set of rules deciding which queue or which person each incoming conversation goes to: by detected language, by customer plan, by words in the message or by team workload. It replaces somebody distributing by hand.
Front line
The front line is whoever receives conversations before anybody else: they resolve what is documented, gather the missing information and pass on what they cannot close. It is the filter that decides what reaches the rest of the team and in what state.
Overflow
Overflow is what happens when more conversations come in than the team can handle within its usual window. It is not a one-off failure: it is a state, and it usually lasts days, because what goes unanswered today piles on top of tomorrow.
Satisfaction per agent
Satisfaction per agent is CSAT broken down by the person who handled each conversation. It is used to see who needs support and what is working, and it is one of the metrics that does the most damage when used for anything else.
9 términos
Metrics
Lo que se mide y lo que significa cada número. Casi todas estas métricas se pueden calcular mal de una forma concreta, y aquí está explicado cuál.
FRT — First response time
First response time is how long passes between a customer sending their question and receiving the first reply from a person. It measures the initial wait, not the resolution, and it is the metric that weighs most heavily on whether support feels good or not.
Resolution time
Resolution time is how long passes between a question arriving and it being resolved and closed. Unlike first response time, it measures the whole process, including waits on the customer and escalations to other teams.
CSAT — Customer satisfaction
CSAT is the metric that measures how satisfied a customer was with one specific interaction. It is asked right after resolving — usually on a one-to-five scale or a thumbs up and down — and expressed as the percentage of positive answers over total answers.
NPS — Net Promoter Score
NPS measures how willing a customer is to recommend a company. It is asked on a zero-to-ten scale and calculated by subtracting the percentage of detractors — zero to six — from the percentage of promoters — nine and ten. The result runs from −100 to 100, and the sevens and eights do not count.
CES — Customer effort
CES measures how much effort it took the customer to get their problem solved. It is asked with a statement along the lines of "the company made it easy for me" and an agreement scale, and it starts from a specific idea: reducing friction retains people better than trying to delight them.
FCR — First contact resolution
First contact resolution is the percentage of questions resolved in the first interaction, without the customer having to write again and without the conversation passing to another person. It is the metric that best summarises whether the team has what it needs to answer within reach.
Deflection rate
Deflection rate is the percentage of questions resolved without a person getting involved, either because the customer found the answer themselves or because an automated agent closed it. It is the metric used to justify any investment in self-service or in AI.
AHT — Average handle time
Average handle time is the time an agent actively spends on a question: reading it, finding the context, answering it and logging whatever needs logging. It excludes waits, which is why it is the metric that best approximates what a given support volume costs in hours.
Ticket backlog
The backlog is the set of questions open and still unresolved at a given moment. More than its size, what gets watched is its trend: if more questions come in than get closed, the backlog grows, and it grows cumulatively even when the daily difference is small.
6 términos
Customer success
El vocabulario de quien se ocupa de que un cliente renueve: lo que se mira, cuándo se mira y qué se hace con ello.
TTV — Time to value
The time between a customer signing up and getting something out of the product that genuinely matters to them. It is not setup time and it is not the first session: it is when the product has solved, for the first time, the problem they paid for.
Health score
A health score is a rating that summarises, in one number, how likely an account is to stay a customer. It combines usage, adoption and relationship signals — how many people log in, which features they use, how many tickets they open, how they respond — into something that can be sorted highest to lowest.
NRR — Net revenue retention
Net revenue retention measures what a group of customers bills today against what they billed a year ago, counting what was lost to cancellations and contractions and what was gained from expansion by those same customers. It excludes new customers: it measures what the existing base does.
Expansion
Expansion is the increase in what an existing customer pays: more seats, more volume, a higher plan or a new module. It is the cheapest source of growth a SaaS has, because nobody has to be convinced to trust you for the first time.
QBR — Quarterly business review
A QBR is a periodic meeting between vendor and customer to review what has been achieved with the product, what has not, and what will happen over the coming months. It leans on usage data and outcomes, not on a deck of new features.
Retention playbook
A retention playbook is the written set of actions triggered by a specific account signal: what happens when usage drops two weeks running, when the sponsor stops replying, or when a renewal approaches. It turns into process what otherwise depends on who happens to be around that day.
13 términos
AI in support
Los términos que aparecen en cualquier demostración de soporte con IA, explicados sin dar por hecho que ya sabes lo que significan.
AI agent
An AI agent is a system that, beyond generating text, reads data and takes actions to finish a task. In support that means it checks the actual state of the customer account before answering, and when the answer involves doing something, it does it instead of explaining how.
Chatbot
A chatbot is a program that holds a written conversation with a user. In support, current ones index the company documentation and answer from it; earlier ones followed decision trees written by hand.
RAG — Retrieval-augmented generation
Retrieval-augmented generation, or RAG, is the technique of finding relevant information in your own sources and handing it to a language model so it writes the answer with that. The model does not memorise the content: it looks it up at answer time.
LLM — Language model
A language model is a system trained on enormous amounts of text that, given a fragment, predicts how it continues. Writing, summarising, translating and holding a conversation all come out of that one simple capability: they are all ways of continuing a text plausibly.
Embedding
An embedding is a text represented as a list of numbers that encodes its meaning. Two texts that say the same thing with different words produce similar lists, and that numeric closeness is what makes it possible to search by meaning rather than by exact match.
Hallucination
A hallucination is an answer from a language model that sounds right and is not: an invented figure, a feature that does not exist, a plausible number nobody checked. It is not a bug, it is a direct consequence of how a model works — it produces the most probable continuation, not the verified one.
Tool calling
Tool calling is the mechanism by which a language model can run functions you define instead of only writing. You describe the available tools, the model decides which one to use and with what arguments, and it receives the result to fold into its answer.
MCP — Model Context Protocol
The Model Context Protocol is an open standard for connecting AI systems to external data sources and tools. It defines a common way to expose what can be read and what can be run, so one connection serves different assistants instead of writing a bespoke integration for each.
Context window
The context window is everything a language model can have in front of it at once to produce an answer: the system instruction, the conversation history, the passages retrieved from documentation and the data looked up. It is measured in tokens and it has a ceiling.
System prompt
The system prompt is the fixed set of instructions a model receives before every conversation: who it is, what it can and cannot do, what tone it answers in and what to do when it does not know something. The user never sees it and it is sent on every request.
Guardrails
Guardrails are the restrictions that stop an AI system doing or saying certain things: which sources it may take data from, which actions it may run, which topics it does not touch and what requires confirmation from a person. They are applied outside the model rather than by trusting it to obey.
Evals — Evaluation
An evaluation is a set of test cases used to check, repeatably, whether an AI system answers as it should. Each case carries an input and a success criterion, and the whole set is re-run every time something changes: the prompt, the model or the documentation.
Latency
Latency is how long an AI system takes to return an answer from the moment it receives the question. It includes the information retrieval, the lookups against other systems and the text generation, and it is perceived whole even though the model is only one part of it.
5 términos
Roles and teams
Quién hace qué en un equipo de soporte, y por qué los nombres de los puestos se solapan tanto entre empresas.
CSM — Customer Success Manager
A Customer Success Manager (CSM) is the person responsible for making sure a customer gets the outcome they signed up for. They work proactively across a book of accounts, own adoption, retention and expansion, and their success is measured in renewals, not in closed tickets.
Support agent
A support agent is the person who handles incoming customer questions: reads them, finds the context needed to answer, resolves them or escalates them to whoever should take over. It is a reactive job by definition, and the work is measured in response time, resolution and satisfaction.
Service desk
A service desk is the single point of contact between an organisation and the people who use its technology services. Beyond resolving incidents, it handles service requests, changes and access, and works through formal processes: a service catalogue, service level agreements and approval flows.
Help desk — Help desk
A help desk is the team and the tool that receive user questions and resolve them. It centralises in one place whatever arrives by email, chat or form, turns each question into a ticket with a status and an owner, and keeps the history of what was answered.
Support tiers
Support tiers are the usual way of splitting questions by what it takes to resolve them. Tier 1 handles the known and documented, tier 2 what requires technical knowledge of the product, and tier 3 what ends up with whoever wrote the code.
10 términos
Business and startup
Las cifras con las que se decide en una empresa de suscripción, y cómo se tocan con lo que pasa en el soporte.
Churn
Churn is the percentage of customers — or of revenue — lost in a period. It is the metric that decides whether a subscription business genuinely grows: with high churn, every new customer only fills the hole left by another.
MRR — Monthly recurring revenue
MRR is the sum of recurring revenue a subscription business bills each month, normalising annual contracts to their monthly equivalent. It excludes anything that does not repeat — implementations, consulting, one-off charges — because its job is to measure the stable base.
Onboarding
Onboarding is the process that takes a customer from signing up to getting their first real result with the product. It covers the initial setup, the data import, training the team and everything else needed for usage to become routine.
Contact rate
Contact rate is how many support questions a business generates relative to its size: normally tickets per customer per month, or per hundred active customers. It is the metric that says whether support will scale with the business or suffocate it.
Runway
Runway is how many months a company can keep operating on the money it has, at the current rate of spend. It is calculated by dividing available cash by what gets burned each month, and it is the figure that orders every other decision.
Burn rate — Burn rate
Burn rate is the net money a company consumes each month: everything it spends minus everything it brings in. It is the denominator of runway and, at a company that bills, it is considerably lower than total spend.
PMF — Product-market fit
Product-market fit is the point at which a product solves a problem well enough, for enough people, that growth stops depending on pushing. You recognise it because demand pulls the team along rather than the other way round.
ICP — Ideal customer profile
The ideal customer profile describes the kind of company the product works best for: its size, its sector, how it works and what problem it has. It is not who you sell to, it is who you should be selling to if you could choose.
CAC — Customer acquisition cost
CAC is what it costs to win a new customer: all marketing and sales spend in a period divided by the customers won in that same period. It includes salaries, tools and advertising, not only the ad budget.
LTV — Lifetime value
LTV estimates how much revenue a customer brings across their whole relationship with the company. At a SaaS it is approximated by dividing average monthly revenue per customer by the monthly churn rate: the less people leave, the more each customer is worth.
