Quick replies: stop retyping the same answer

A quick reply is an answer you save once and insert with a shortcut, instead of retyping it several times a day. It is the least glamorous feature in a shared inbox and one of the highest-return ones, because it applies to work your team is already doing.

It also improves consistency, which matters more than the time saved: six agents explaining your returns policy in six ways is a support problem that a saved answer quietly removes.

Setting them up

Open the Inbox and pick any conversation.
In the message bar, click the lightning-bolt icon, then Manage Quick Replies.
Give each one a short shortcut — hi, price, returns — and the message text.
Save.
While chatting, type / followed by the shortcut, or tap the icon, to insert it.

The shortcut is the part worth thinking about for ten seconds. Short, lowercase, and guessable by somebody who did not create it — returns rather than r2.

What to save

Good quick reply	Why
Delivery times by region	Asked constantly, and the answer is factual
Returns and exchange policy	Must be consistent, word for word
Payment methods you accept	Short, fixed, tedious to retype
Bank or payment details	Retyping these is where mistakes happen
"We're closed now, here's when we reply"	Used dozens of times a week
Sizing or compatibility guidance	Long to type, easy to get subtly wrong

The fourth row is the one worth adopting first even if you adopt nothing else. An account number typed by hand at speed is a mistake waiting to be made, and it is the most expensive kind of typo in a customer conversation.

What not to save

Anything that changes often. A stale saved answer is worse than none, because it will be sent confidently.
Anything that should be automatic. If every customer asks it, build a flow instead — see below.
Long blocks of several answers at once. Customers read short messages; a wall of text gets skimmed.
Anything personal to one agent. These are shared, and they should read like your business rather than like a person.

Quick reply or automation?

Both answer a repeated question, and the line between them is who decides.

An automation answers without a person, which is right when the question is unambiguous and the answer never depends on context: opening hours, a tracking link, a menu.

A quick reply is right when a person needs to judge whether the answer applies. "Can I return this?" usually has a standard answer and occasionally does not, and the agent deciding which is the value they add.

Most teams end up using both for the same topic — the bot handles the plain version and the agent has the nuanced version a keystroke away. When a question is common enough to automate, the route is in how to build your first automation.

They are not templates

An important distinction, because the words get used interchangeably and they are not the same thing.

	Quick reply	Approved template
Approved by Meta	No	Yes, before use
Works outside the 24-hour window	No	Yes
Costs money	No	Yes
Edit it whenever	Yes, instantly	Resubmit for review

A quick reply is an ordinary message with a shortcut, so it is subject to the same window as anything else you type — free and unlimited inside it under Meta's pricing documentation, and impossible outside it. That is covered in the 24-hour countdown.

Keeping the list useful

Start with six. A list of forty is a list nobody browses.
Ask the team which they actually use after a month, and delete the rest.
Review them when anything changes — prices, delivery times, policy.
Check them before peak season, when delivery answers stop being true.

Ten minutes a quarter. The failure mode here is never too few quick replies; it is a long list where three are wrong.

Frequently asked

Can each agent have their own?

They are shared, which is the point — consistency across the team is most of the benefit.

Can a quick reply include the customer's name?

Personalising by hand is fine. If you want it filled in automatically for every customer, that is an automation rather than a saved answer.

Do they work on Instagram and Messenger?

They are ordinary messages, so they go wherever you can type.

Should the bot's answers and my quick replies match?

Yes, word for word where they cover the same question. A customer who gets one answer from the bot and a different one from a person notices.

Six to start with

If you want a list you can create in ten minutes and improve later:

hi — your greeting, in your voice.
delivery — times and cost by region.
returns — the policy, in full, word for word.
pay — accepted payment methods and any details.
closed — you are outside hours, and when you will reply.
thanks — a proper sign-off, so nobody ends a chat abruptly.

Those six cover a surprising share of what a small team types. Add the seventh when you notice yourself typing it twice in a day.

The consistency argument

Time saved is the obvious benefit and the smaller one. The real gain is that your returns policy is explained the same way by everyone, which matters when a customer disputes what they were told.

It also makes onboarding much faster. A new agent with six saved answers can handle a real conversation on day one without having memorised anything, which is not true of a new agent with a policy document.

Where a quick reply is the wrong tool

One case worth naming clearly, because it is the mistake that costs money. A quick reply cannot reach a customer whose window has closed — it is an ordinary message, so it needs the same open conversation everything else does under Meta's sending messages documentation.

If you find your team wanting to send the same saved answer into cold chats, what you actually need is an approved template under Meta's template documentation. Those are a different object with a different cost, and the distinction is covered in utility versus marketing templates.

A rough sense of what they save

Take a common answer — a returns policy, say — at roughly 40 words. Typed carefully that is perhaps 30 seconds. Sent 10 times a day across a small team, that is a few minutes daily and somewhere near an hour a month, from one saved answer.

None of which is the main point, but it does explain why teams that adopt them never go back. The consistency is the benefit; the time is the thing that makes it obvious.

Rolling them out

Ask each agent for the 3 answers they type most.
Deduplicate — there will be a lot of overlap.
Write the best version of each, once, together.
Save them with obvious shortcuts.
Review after a month and delete whatever nobody used.

Step three is worth doing as a group rather than delegating. Agreeing the wording is the moment a team discovers it has been telling customers three different things.

The short version

Save the 6 answers your team types most, give each a short lowercase shortcut, and insert them with / or the lightning icon. Keep the list short, review it when anything changes, and remember they are ordinary messages — so they need an open window, and anything that must reach a cold chat is a template instead.

The consistency is worth more than the time. Six agents explaining your returns policy identically is a quieter inbox than six agents explaining it six ways, and it takes ten minutes to arrange.

Keeping them from going stale

The failure mode is not neglect, it is success — a list that works gets trusted, and a trusted list stops being read. Then delivery times change and your team keeps sending last season's answer with total confidence.

Two safeguards, neither of which needs a process. Tie a review to anything that changes prices, delivery or policy, and read the whole list before peak season, when almost every operational answer stops being true for a few weeks.

It takes ten minutes and it prevents the specific complaint that is hardest to recover from: a customer who was told something in writing that was not true.

Naming the shortcuts

Short, lowercase, and describing the answer rather than abbreviating it. A new agent should be able to guess half of them without being told, which is the whole test — returns passes it and ret2 does not.

Avoid shortcuts that are prefixes of each other if you can, since typing the shorter one is then ambiguous, and keep them to one word so nobody has to remember whether it had a hyphen in it.

One more small thing: tell new agents the list exists. It is discoverable only if somebody points at the lightning icon, and a saved answer nobody knows about saves nobody any time.