Customer service glossary
Short, plain definitions of the terms that come up when you run support for an online shop or a small SaaS product.
A
- AI deflection
- AI deflection is the share of customer questions that an AI chatbot or agent closes without a person on your team getting involved.
- Auto-send
- Auto-send is a setting in which an AI helpdesk sends its drafted replies to customers without a person approving each one first.
C
- Chargeback
- A chargeback is a reversal of a card payment, made by the customer's card issuer after the customer disputes the charge with their bank.
- Conversation
- A conversation is one customer's thread with your team about a single matter, however many messages go back and forth in it.
- CSAT
- CSAT (customer satisfaction score) is a measure of how satisfied customers are with a specific support interaction, usually collected with a one-question rating after the conversation closes.
E
- Escalation
- Escalation is the act of passing a support conversation to a person or team with more authority, expertise or access than whoever is currently handling it.
F
- First response time
- First response time is the time between a customer sending their first message and your team sending its first reply, not counting automatic acknowledgements.
H
- Helpdesk
- A helpdesk is software that gathers customer messages from different channels into one place, so a team can assign, answer and track them.
K
- Knowledge base
- A knowledge base is a collection of your policies, product details and answers to common questions that your team, and any AI drafting replies, rely on to answer customers correctly.
M
- Macro
- A macro is a saved reply, often with placeholders for details such as the customer's name or order number, that a support team reuses for common questions.
O
- Omnichannel support
- Omnichannel support is customer service that handles every channel a customer uses, such as email, chat and social messages, in one place with one shared history per customer.
P
- Per-resolution pricing
- Per-resolution pricing is a billing model in which a helpdesk or AI vendor charges a fee each time its AI resolves a customer conversation, on top of or instead of a seat or plan price.
- Prompt injection
- Prompt injection is an attack in which someone hides instructions inside text an AI will read, such as a customer email, to make the AI ignore its rules or take actions it should not.
R
- Resolution time
- Resolution time is the time between a customer first getting in touch and their issue being fully solved and the conversation closed.
- RMA
- RMA (return merchandise authorisation) is the approval and reference number a seller gives a customer before they send an item back, so the return can be tracked and matched to the order.
S
- Shadow mode
- Shadow mode is a way of running an AI helpdesk on your real customer conversations in which it drafts replies and makes decisions but sends nothing to customers.
- SLA
- SLA (service level agreement) is a target, sometimes a formal promise, for how quickly a support team will reply to or resolve customer conversations.
T
- Ticket
- A ticket is a single customer request tracked in a helpdesk, with its own status, owner, tags and message history from first contact to close.
- Ticket triage
- Ticket triage is the sorting of incoming customer messages by topic, urgency and who should handle them, so the most important ones are answered first.
W
- WISMO
- WISMO (where is my order) is the name for customer messages asking about the status or location of an order they have placed.