When a Post Goes Wrong: A Social Media Crisis Plan for Small Brands

Most content advice assumes everything works. You fill the calendar, schedule a few weeks ahead, and the posts go out while you run the business. That's the good case, and it's most days.

Then there's the other morning. A price in a carousel is wrong and a few hundred people have already seen it. A line that read as funny in the draft reads as something else entirely at 8am. Or nothing is wrong with your post at all — something happened in the world overnight, and the cheerful promo sitting in your queue for 9:15 is about to publish straight into it.

Small brands rarely plan for this, because it feels like a big-company problem with a big-company answer. It isn't. You need three things to happen fast — notice, decide, act — usually done by one person, on a phone, before coffee. A plan is just making those steps shorter.

Two kinds of wrong, two opposite responses

Before you do anything, sort which one you have. They look similar in the moment and need almost opposite handling.

You got it wrong. The post itself is the problem: a factual error, a broken promise, a tone that landed badly, a claim you can't back up. The content is the thing to fix.

The timing got it wrong. Your post is fine. A disaster, a tragedy, or a major news event has made the room wrong for it. Nothing needs correcting — things need to stop.

Confusing the two wastes the only resource you have, which is the first twenty minutes. Brands apologize for posts that needed no apology, and quietly delete posts that needed a correction in public.

The first move is the same either way: stop the queue

This is the step small brands skip, and the one that does the most damage. A scheduled week is an asset right up until the moment it isn't. If you've done the work of planning ahead, you have content that will publish itself with no human in the loop — exactly the point, and exactly the risk. A promo auto-posting ninety minutes after bad news reads as though you weren't paying attention. Nobody checks the timestamp to see it was scheduled three weeks ago.

So: pause everything first, triage second. Pausing the whole queue is one action; deciding post by post under pressure is twenty. You can always resume tomorrow.

The practical part is knowing where that pause is before you need it. If your posts are spread across each platform's native scheduler plus a spreadsheet, there is no pause — there's a scavenger hunt. Seeing the whole week in one calendar and stopping it in one place is the difference between a ten-second fix and a bad afternoon.

Delete, edit, or leave it

Once the queue is stopped, deal with the post that's already live. There are three options, and the right one depends on how much engagement it already has.

Delete it when the content is genuinely harmful, offensive, legally risky, or exposes someone's private information — or when the error is factual, consequential, and still spreading. Then post the correction separately, so people who saw the original can find it.

Edit it when the platform allows it, the mistake is small, and nobody has engaged with the error yet. A typo, a broken link, a missing credit. This is the narrow case.

Leave it up when people have already replied, quoted, or screenshotted it. This feels counterintuitive and it's usually right. Deleting a post a hundred people are discussing doesn't remove it — it removes your side of it, and reads as hiding. Correct it in public instead.

The one thing to never do is silently edit something people are already talking about. Someone will notice, and "they changed it quietly" becomes a bigger story than the original mistake.

Write the correction before you write the explanation

The instinct is to explain how it happened. Resist it. Nobody outside your company is interested in your internal process, and the explanation always sounds like a defense.

A correction that works is about four sentences:

1. What was wrong. Plainly, without "it has come to our attention."
2. What's actually true. The corrected price, the real date, the accurate claim.
3. What you're doing about it. Only if there's something concrete — refunds, a fix, a changed process.
4. Where to go next. A link, a DM invitation, a contact.

Put it where the original was. If the mistake was in a feed post, the correction is a feed post — not a story that disappears in a day, and not a comment buried under forty replies.

If the mistake affected someone, say so in the first sentence rather than the last. The brands that recover from a genuine misstep are the ones that name it early. Handling the replies that follow is its own skill, and it gets much easier once the correction itself is honest.

For external events, silence is a real option

When the problem is timing rather than content, the plan is mostly restraint.

You are not obliged to comment on every news event, and a small brand posting a solemn statement about something unrelated to its work reads as opportunism more often than solidarity. Three questions before you say anything:

• Do we have standing? Does this touch our customers, our team, or our actual work?
• Do we have something to add? Beyond restating what everyone already said this morning.
• Can we handle the replies? If you post it and go into a meeting, don't post it.

If any answer is no, stay paused and say nothing. Then unpause when the moment has genuinely passed — not at an arbitrary 24 hours. This is the inverse of trend-jacking: same speed, same judgment, opposite direction.

Prepare the four things you can prepare

None of this needs a document. It needs four decisions made on a calm day.

Know how to pause. Not theoretically — open your scheduler now and find the control. If stopping everything across every channel takes more than a few seconds, that's the real problem to fix.

Name one decider. One person decides delete-edit-leave — not a committee of two, each waiting for the other. If you've handed social off to someone else, agree on this before you hand over the password.

Draft the holding line. One sentence you can post in sixty seconds while you work out the real answer: "We're aware of an error in this morning's post — we're checking it now and will update shortly." Having it written removes the worst twenty minutes.

Have a few saved replies. When something goes wrong, the same question arrives forty times. Answering it forty times by hand, at speed, is how a small problem becomes a thread. A single inbox across platforms absorbs the volume while you deal with the substance.

Afterwards: fix the step, not the person

A day or two later, do a short post-mortem. Where did it slip? The brief was vague. The draft was written at 11pm. It went live without a second read. Each has a different fix, and most are one added step in your approval workflow: a price check, a second pair of eyes on anything with a number in it, a Monday scan of what's queued. The lesson is almost never "be more careful" — it's "that step didn't exist."

And don't draw the wrong conclusion from a bad morning. Abandoning the calendar to post manually, so you always have control, trades one rare failure for a daily one: inconsistency, rushed captions, and the slow drift away from posting at all. Scheduling ahead is still the right call. The requirement is only that the queue stays visible and stoppable — every channel's next two weeks in one view, pausable in one action, resumable when the room is right again. That's less a crisis plan than a property of the setup, and it's the difference between a scheduled calendar being a safety net and a liability.