Macros & canned responses

Most of what a support team types, it has typed before. Canned responses and macros turn that repetition into one keystroke — and, more importantly, make the answer consistent no matter who is on shift.

The difference

Use a canned response when the only repeated thing is the words. Use a macro when the same words always come with the same bookkeeping.

Canned responses

Create them under Settings → Canned responses. Each has a short code and the message body.

Agents insert one by typing / in the composer followed by the short code. The text drops in, ready to edit.

Short codes are worth thinking about for a minute. Group them by prefix so the list stays navigable as it grows: ship-delay, ship-tracking, ship-lost, bill-refund, bill-invoice. An agent who types /ship then sees every shipping response.

Writing good ones

The failure mode of canned responses is not that they save no time — it is that customers can tell. Some things that keep them from reading as boilerplate:

• Leave a gap for the specific. Start with a line the agent must personalise. A response that can be sent unedited will be sent unedited, to someone whose problem was slightly different.
• Write like a person. "We're on it — give me twenty minutes to check with the warehouse" beats "Your request has been received and is being processed."
• Answer the question, then explain. People stop reading after the first paragraph.
• Keep them short. A long canned response is one nobody reads before sending.
• Say what happens next, with a time. Most follow-up messages exist because the first message did not.

Macros

Create them under Settings → Macros. A macro is an ordered list of actions, and it can do most of what a workflow can do — the difference is that a macro runs when an agent chooses it, not automatically.

Available actions include sending a message, adding or removing labels, assigning to an agent or team, changing priority, changing status, sending an attachment, and adding a private note.

Macros can be personal or shared with the whole account. Let people build their own and promote the good ones to shared — that is how you find out what your team actually does, rather than what you think it does.

Macros worth building

Refund approved. Sends the confirmation message, labels the conversation refund, sets the status to pending, and adds a private note for finance. Four actions, one click, and nothing gets forgotten.

Escalate to engineering. Adds the label bug, sets priority to high, assigns to the Engineering team, and posts a private note with the template your engineers want — steps to reproduce, account ID, when it started.

Close politely. Sends the "anything else?" message and sets the conversation to pending. The most-used macro on most accounts.

Ask for the missing detail. Sends the message asking for an order number, sets pending, and labels awaiting-customer. Now you can build a folder of everyone you are waiting on.

Out of scope. Sends the explanation, applies the label, resolves. Kind, consistent, and the label tells you later how much of this you are getting.

Keeping the library healthy

A canned response library grows until nobody can find anything, and then people stop using it and type from scratch again.

Two habits prevent that. Review quarterly and delete anything unused — an untouched response is either wrong or unfindable, and both are worth fixing. And watch what agents type by hand: anything typed repeatedly that is not in the library is a gap, and anything in the library nobody uses is noise.

If you find agents heavily editing the same response every time, the response is wrong. Rewrite it to match what they actually send.