Canned responses, macros, and automation solve three different slices of the same problem — repetitive support work — and the trick is matching each to the right job. Use a canned response when you only need to insert the same text; a macro when that text always comes with the same housekeeping actions; and automation when no human judgment is needed at all. Template the repetitive, automate the deterministic, and let a person type whenever the answer depends on context.
Three tools, one spectrum
It helps to see these three as points on a single line measured by how much human judgment is involved. A canned response assumes a human is doing everything except retyping. A macro takes a few mechanical steps off that human's plate. Automation removes the human entirely. The skill is not knowing the definitions — it is knowing where on that line a given task belongs, because putting a task one notch too far toward automation is how customers get the wrong answer with full confidence.
Canned responses: reusable text
A canned response is a saved reply template, often with placeholders and a / shortcut. Use them for the exact same answer you type ten times a day — “here's how to reset your password,” “we've issued the refund, here's the timeline.” Good canned responses have two properties: placeholders for the bits that change (name, order number) so they never feel robotic, and a short, memorable shortcut so agents reach for them instead of retyping. Keep them in a shared library, not in each agent's personal snippets, so the whole team answers consistently and you can update one place when the answer changes.
Macros: text plus actions
A macro is a canned response that also does things — set status, assign, add a tag, snooze — in one click. Reach for a macro when the reply always comes with the same housekeeping. The classic case: a refund reply that should also tag the ticket refund, set it to pending, and assign it to billing. Doing that by hand four times is fine; doing it forty times a day is where steps get skipped. A macro makes the whole sequence atomic, so the side effects happen every time the reply does.
Macros keep humans in the loop
The important property of a macro is that a person still chooses to run it. That keeps judgment in the loop — the agent decided this reply fits this ticket — while removing the mechanical tax. That is exactly the line you want to hold for anything customer-facing.
Automation: no human at all
Automation runs without an agent: auto-tag inbound mail by keyword, auto-route by topic, notify or escalate when an SLA is at risk, auto-close pending tickets after a quiet period. Automate the decisions that are always the same and carry low downside if they are occasionally wrong. The danger is silent failure: a rule that mis-routes for weeks before anyone notices is worse than no rule. Build automations incrementally, test them on real traffic, and prefer platforms with a dry-run mode so you can see what a rule would have done before you let it act. The broader picture of how rules tie statuses, owners, and views together is in ticketing 101.
A quick decision table
When you are unsure which tool fits, this is the short version:
| If the task is… | Reach for… | Because… |
|---|---|---|
| The same text, typed by a human | Canned response | Only the words repeat; judgment stays human |
| The same text plus the same housekeeping | Macro | A person decides, but the side effects are mechanical |
| A deterministic decision, no judgment | Automation | The rule is right every time and low-risk if it isn't |
| An answer that depends on context | A human typing | No template captures the nuance |
Drawing the line
Template the repetitive, automate the deterministic, and let a human type whenever the answer depends on context. The mistake is automating empathy — customers notice when a reply is templated past the point where it fits their situation. A practical test: if you would be embarrassed to have a customer see exactly how the reply was produced, it has moved too far toward automation. Speed matters — a fast first response is worth a lot — but not at the cost of sending the wrong answer quickly.
Where AI assist fits
AI sits alongside these tools rather than replacing them. Opt-in, admin-gated AI-assisted replies can draft a response from the thread and your knowledge base, but the draft is a starting point an agent reviews and edits — not something that sends on its own. The same line applies as with automation: AI is a faster way to propose, while a human still decides. Used that way, it shortens the time to a good reply without putting an unsupervised answer in front of a customer. The admin controls govern who can turn it on and what it can see.
Keeping templates from rotting
Templates and rules are not set-and-forget. Treat them like code:
- Give each category an owner who is responsible for keeping its templates accurate.
- Review on a schedule and retire anything nobody has used in months.
- Watch for heavy edits. If agents rewrite a template every time before sending, it no longer matches reality.
- Prune the tag and rule list so the automation surface stays small enough to reason about.
Stale templates are worse than none, because they send outdated information confidently. A little maintenance keeps the whole library trustworthy.
Frequently asked questions
What is the difference between a canned response and a macro?
A canned response only inserts text — a saved reply template you drop into the editor. A macro inserts text and also performs actions in the same click, such as setting a status, assigning the ticket, or adding a tag. Use a canned response when all you need is the words; reach for a macro when the reply always comes with the same housekeeping.
When should I use automation instead of a macro?
Use a macro when a human is in the loop making a judgment call but the follow-up steps are predictable. Use automation when no judgment is needed at all — the decision is deterministic and the same every time, like tagging inbound mail by keyword or escalating a ticket that is about to breach its SLA. The dividing question is whether a person needs to look at it first.
Should I automate replies to customers?
Be careful. Automating acknowledgements and routing is safe, but automating substantive replies risks sending the wrong answer with full confidence, and customers can tell when empathy is templated. Keep a human on anything where the answer depends on context. If you use AI assist to draft replies, treat the draft as a starting point an agent reviews and edits, not as something that sends on its own.
How do I keep templates from going stale?
Treat templates like code: review them on a schedule, retire the ones nobody uses, and assign an owner for each category. Stale canned responses are worse than none because they send outdated information confidently. Watch which templates get edited heavily before sending — that usually means the template no longer matches reality and needs rewriting.
Cherryrise has macros, canned replies and a full automation engine with dry-run. See it.