New agent onboarding

A checklist for your first week on the support team.

Day one

1. Get your accounts and access (see Tools and access).
2. Read The support workflow and Triage and prioritisation.
3. Shadow a senior agent on live tickets.

Week one

• Handle tickets with a lead reviewing your replies before they send.
• Learn the macro library and when to use each macro.
• Read the escalation paths so you know who owns what.

When you are signed off

You handle tickets independently, escalate by the matrix, and flag anything missing from this KB so we can add it.

Tools and access

The systems the support team uses, and how to get into them.

Core tools

• Helpdesk: where tickets, macros and views live.
• Status page: where we post incident updates.
• Internal chat: the escalation and outage channels.

Getting access

Access is granted by your team lead on day one. If something is missing, raise it in the team channel rather than working around it.

The support workflow

The path every ticket follows, from arrival to resolution.

1. Triage: set priority and category (see Triage and prioritisation).
2. Acknowledge: reply within the first-response target, even if only to say you are on it.
3. Resolve or escalate: solve it, or hand it off by the escalation matrix.
4. Close: confirm the fix, then close with the right macro and tags.

A closed ticket reopens automatically if the customer replies, so closing is safe.

Triage and prioritisation

Every ticket gets a priority, because it drives the target time and the queue order.

Priorities

• Urgent: the customer is fully blocked, or it affects many users.
• High: a core feature is broken for one customer.
• Normal: a question, or a minor issue with a workaround.
• Low: feedback, or a request with no time pressure.

How to set it

Read the ticket, judge the impact, and set the priority before you reply. When in doubt, pick the higher priority and let a lead adjust it.

Writing a good reply

Consistent, clear replies are the point of this team.

The shape of a reply

1. Acknowledge the problem in one line.
2. Give the answer or the next step, plainly.
3. Say what happens next, and by when if relevant.

Tone

• Be direct and warm. Short sentences.
• Never blame the customer, and never guess. If you are unsure, say you will check.
• Start from a macro where one fits, then tailor it. Do not send a raw macro.

Using macros and canned replies

Macros keep answers consistent and fast. They are a starting point, not a script.

When to use one

Use a macro when the situation matches a known, repeatable case: a password reset, a known bug, a refund acknowledgement.

How to use one well

1. Insert the macro.
2. Edit it for the specific customer and ticket.
3. Check any placeholders are filled before sending.

When to escalate

Escalate early rather than sitting on something you cannot resolve.

Escalate when

• The fix needs access or permissions you do not have.
• It is a bug that needs engineering.
• It is outside policy (a refund beyond your limit, a legal or security matter).
• The customer is at risk of leaving and needs a faster path.

How to escalate

Summarise what you know, what you tried, and what you need, then hand off by the matrix. Do not simply reassign without context.

Escalation paths and owners

Who owns what, so an escalation reaches the right place first time.

• Billing and refunds: the billing owner, within policy limits.
• Bugs and outages: on-call engineering, via the incident channel.
• Security or data requests: the security contact, immediately, no exceptions.
• Anything unclear: your team lead.

Handling an outage

What the support team does when something is down for many customers.

1. Confirm it is a real, widespread incident, not a single-customer issue.
2. Post to the incident channel so engineering and leads are aware.
3. Hold the line: acknowledge, point customers to the status page, do not speculate on cause or timing.
4. Update as the status page updates. One source of truth.
5. Follow up with affected customers once it is resolved.

Macro library

The canonical list of macros and when to use each. This page is the index; each macro's full text lives in the helpdesk.

Common macros

• Acknowledgement: first reply while you investigate.
• Password reset: the steps to reset, with the self-serve link.
• Known bug: acknowledge a tracked bug and set expectations.
• Refund acknowledgement: confirm a refund is being processed.
• Closing: a friendly close once the issue is resolved.

Macros are owned by the team. Propose changes to a lead; do not edit shared macros ad hoc.

Priority and SLA matrix

The response targets for each priority. These are internal targets that drive the queue.

• Urgent: first response within 1 hour, updates hourly.
• High: first response within 4 hours.
• Normal: first response within 1 business day.
• Low: first response within 2 business days.

Resolution time depends on the issue, but the customer should always get an acknowledgement within the first-response target.

Glossary and internal terms

Shorthand the team uses, so a new agent is not lost in a thread.

• KB: this internal knowledge base.
• Macro: a saved, reusable reply.
• SLA: the response target for a priority.
• Triage: setting a ticket's priority and category.
• Escalation: handing a ticket to the owner who can resolve it.
• Urgent / P1: a widespread or fully blocking issue.

Refunds and exceptions

What an agent can approve, and where approval is required.

Within your authority

• Refunds up to the standard limit for a clear service failure.
• Extending a trial by a short, standard period.

Needs approval

• Refunds above the limit.
• Any goodwill credit outside the standard cases.
• Anything that sets a precedent, or that a customer could cite later.

Data and privacy for agents

Handling customer data responsibly is part of every ticket.

• Verify identity before making account changes or sharing account details.
• Share the minimum: only the information needed to resolve the ticket.
• Never ask for a full password, a card number, or any other secret in a ticket.
• Escalate any data-access or deletion request to the security contact; do not action it yourself.