Two words used as one, for two different things
“Macro” and “canned response” get thrown around interchangeably, and the blur causes real confusion when a team tries to organize and share theirs. They overlap, but they aren't the same idea. Getting the distinction straight helps you build a set of shortcuts that actually speeds agents up instead of becoming a tangle nobody trusts.
What a canned response is
A canned response is a saved piece of text an agent inserts into a reply — a pre-written answer to a common question, dropped into the composer and usually personalized before sending. Its job is to save typing on things you say often. Think of it as a well-written paragraph you keep on hand, not a workflow.
What a macro is
A macro is broader: it can insert text, but it can also perform actions — applying a tag, setting a status, triggering a next step. Where a canned response is words, a macro is words plus a small bundle of the routine actions that usually go with them. A macro is closer to “do the whole standard move for this situation,” not just “paste the standard wording.”
Words, versus words plus the actions that go with them
|
Canned response |
Macro |
| What it is |
Saved text an agent inserts into a reply |
Text, plus a bundle of the routine actions that usually follow |
| What it saves |
Typing on things you say often |
The whole standard move for a situation |
| Can it change a record? |
No |
Yes — tags, status, a next step, a webhook |
| Reach for it when |
The repeated thing is what you say |
The repeated thing is what you do |
| Care required |
Low — review the wording and personalize before sending |
Higher — anything touching money or entitlements deserves a deliberate step |
When to reach for which
Reach for a canned response when the repeated thing is purely what you say — a clear answer to an FAQ. Reach for a macro when the repeated thing is what you do — a sequence of an answer plus the tagging or status changes that always accompany it. Using the heavier tool for a simple paste is overkill; using a bare paste where actions should follow leaves work undone.
Combine them without a mess
The failure mode for both is sprawl — hundreds of half-remembered shortcuts nobody can find. Keep the set small, name each by the situation it answers, personalize before sending, and prune on a schedule. A tight, well-labeled library of either beats a bloated one, and the two coexist happily when each is used for the job it's actually good at.
How MyLiveChat fits
MyLiveChat supports canned/predefined responses agents insert into the composer, with a shortcut that expands a short trigger into the full reply while typing — so the wording side pays off as speed at the keyboard. Keep the library named, personalized, and pruned, and those shortcuts stay an accelerant rather than clutter.
Where automation should stop
The moment a macro does something other than insert text — tagging, assigning, closing,
firing a webhook, changing a record — it stops being a writing shortcut and becomes an action
with consequences. That deserves a different level of care than a snippet of friendly wording.
The rule worth adopting is that anything reversible can be automated freely, and anything that
touches money, entitlements, or another team's queue should require a deliberate step. An agent who
can close a ticket with one key will occasionally close the wrong one; an agent who can issue a
refund with one key will eventually issue the wrong refund, usually while handling three chats.
Side effects also make macros harder to reason about later. When something goes wrong, a text
snippet leaves an obvious trail and an automated chain of actions does not. If a macro performs
more than one consequential action, write down what it does somewhere a person can read without
opening the configuration.
Test one before you trust it
Both canned responses and macros fail quietly. The text stays live long after the policy it
describes has changed, and nobody notices because it reads fluently and nobody re-reads what they
paste. The failure surfaces as a customer quoting your own outdated wording back at you.
Before a snippet goes into the shared library, read it cold in the context where it will be
pasted. Does it make sense as the second message in a conversation, without the sentence you
happened to be thinking when you wrote it? Does it still make sense if the visitor has already
explained their situation, or does it ask them to repeat themselves?
For anything with an action attached, run it once on a real but harmless case and check what
actually happened rather than what you expected. Verify the tag applied, the assignment landed, and
the visitor-facing text is what you intended. The gap between intention and configuration is where
most macro problems live.
What to measure
Track which snippets are actually used. Libraries grow monotonically unless someone measures, and
the ones nobody uses are not harmless — they make the useful ones harder to find.
Watch satisfaction on conversations that used a canned response versus those that did not. A
significant gap usually means the snippet is being pasted where it does not fit, which is a training
and naming problem rather than a writing one.
Keep a note of when each snippet was last reviewed. Age is the best available proxy for
inaccuracy, and a date beside each entry makes the pruning conversation short.
Whichever of the two you reach for, the console remembers which ones you actually use and puts them in front of the rest. Why your canned replies reorder themselves covers what that memory holds and what it silently loses.