Vazante

Metrics and analysis

How to pause an ad in Meta Ads

How to pause an ad, ad set or campaign in Meta Ads, what each level affects, and why pausing and resuming costs more than most people assume.

Also available in: Português · Español

Pausing an ad is two clicks. This article exists for what comes before and after that: picking the right level, and knowing what the pause costs — because pausing and resuming is not a neutral operation.

Where you pause, and what each level affects

Meta Ads has three levels, and pausing works the same in all three: a toggle in the first column of the table.

LevelWhat stopsWhen to use it
AdOnly that pieceTired or rejected creative
Ad setEvery piece in that audience and budgetSaturated audience, misallocated budget
CampaignEverything inside itEnd of a promotion, strategy change

The rule: pause at the lowest level that solves your problem. Pausing the campaign because one creative got tired switches off what was working alongside what was not.

What happens to the budget

It depends on where the budget sits, and this is the part that catches people out.

Budget at the ad set. Pausing an ad set frees that money — it simply is not spent. Other ad sets keep what they had.

Budget at the campaign. Pausing an ad set redistributes its budget among the remaining ones. Total daily spend does not drop; it concentrates.

The second case surprises anyone who pauses an ad set expecting to spend less and sees the same total. It is not a bug: it is the campaign doing what it was configured to do. The difference between the two structures is in CBO vs ABO.

The cost of pausing and resuming

Here is the real reason for this article. Pausing feels free and is not.

When an ad set stops for several days and returns, it loses the stability it had. The system has to find who responds again, and in the first days cost per result rises — sometimes well above what it was before the pause.

The pattern that costs the most money in an account: pause Monday because Sunday was bad, resume Wednesday because budget is piling up, pause again Friday. The ad set never leaves the unstable state, and cost per result sits permanently above where it would be.

The explanation is in Meta Ads learning phase. The practical summary: pausing is a decision, not a fine adjustment.

When pausing is the right call

A rejected ad or a policy warning. Pause the piece, fix it, upload again.

Clear and sustained creative decline. Not one bad day: two weeks of falling CTR with rising frequency. The signal is in creative fatigue.

End of a promotion. The offer ended, the ad has to stop.

A problem on the site or with stock. Paying for clicks to a page that cannot sell is the only situation where pausing is urgent.

Abnormal spend. Delivery concentrated somewhere you did not configure.

When pausing is the wrong call

One bad day. Daily variation is normal, and the day that looks terrible at 10am usually closes inside the average.

Cost per result above target on a small sample. Three conversions is not a sample. Pausing on them is pausing on noise.

During the learning phase. You are judging an ad set that has not stabilized, and the pause restarts the process.

Because the client called upset. It happens, and the conversation is worth more than the pause.

Pausing is not deleting

Worth separating, because the confusion erases history that does not come back.

Pausing keeps everything: history, learning, the option to resume, data in reports.

Deleting removes the object from the panel. Historical spend stays accounted for on the account, and the object itself disappears — you can no longer open that ad to see what it was.

In practice there is almost never a reason to delete. An old paused campaign bothers nobody and one day serves as a reference.

Pausing in bulk, without damage

When a lot has to stop — a holiday, a campaign ending, a site problem — the order that costs least:

  1. Pause at the campaign level, not ad by ad.
  2. Write down what was active before pausing. On resuming, the list stops you leaving a good campaign behind.
  3. Check the next day's spend. If there is still spend, something stayed on.
  4. Resume everything at once, not gradually, if the intent is returning to the prior state.

Step 2 is the one nobody does and the one that pays off most. Two weeks later, nobody remembers which of forty ad sets were running.

The alternative that is usually better

Before pausing for performance, consider reducing. Lowering an ad set's budget to a minimum keeps its history alive and cuts spend — and resuming later does not go through the full restart a pause causes.

It does not fit every case: a rejected ad has to stop, a site that is down has to stop. But for "this is too expensive", reducing usually costs less than pausing.

The mirrored decision, when to increase, is in when to scale a campaign.

What to do if you paused and should not have

It happens, and the correction has an order.

Resume the same day if you can. A pause of hours is usually absorbed with no restart.

Do not duplicate to "start clean". The duplicate is born with no history and takes longer to stabilize than the original would take to recover.

Do not raise the budget on resuming. Returning with more budget than before stacks two changes.

Wait 48 hours before judging. Cost per result in the first days after a resume represents nothing.

Schedule it instead of pausing by hand

For predictable stops — a holiday, a stock closure, the end of a promotion — scheduling prevents forgetting in both directions.

Setting an end date on the campaign makes it stop on its own. More importantly, it prevents the most expensive case, which is forgetting to resume. A campaign paused by hand on a Friday for a holiday and resumed the following Wednesday lost five days of delivery nobody planned to lose.

Two warnings about scheduling:

The time zone is the account's, not yours. An account created with another country's time zone ends the campaign at a different hour than you expect.

Ending is not pausing. A campaign that reached its end date shows as completed, drops out of the active filter, and resuming it requires changing the date — the toggle alone will not do it.

For unpredictable stops, scheduling does not help and the toggle is what you have. There, the written list from the section above is what prevents the forgetting.

Before pausing anything

Look at the closed week, not the day. Check how many conversions support the number you are judging. And if the answer is under twenty, the decision you are making is about noise.

The reporting format that prevents the reflex pause is on the paid media reporting page.

Frequently asked questions

Does pausing cost anything?

No. A paused ad does not deliver and does not spend. What can happen is a residual charge for impressions from the last hours, still in processing when you paused.

What is the difference between pausing and deleting?

Pausing is reversible and keeps history, learning and the option to resume. Deleting is permanent and removes the object from the panel. In practice there is almost never a reason to delete.

Does pausing for a day restart learning?

Short pauses are usually absorbed. A multi-day pause, or repeatedly pausing and resuming, throws the ad set back into the unstable phase and cost per result rises before settling.

Can I pause from my phone?

You can, through the Ads Manager app. It is useful in an emergency, and it is where the worst decisions get made: pausing from a phone on a bad morning, without seeing the whole week, is the most expensive easy decision there is.

Read next