Is Meta Ads down? How to confirm and what to do

"Meta Ads is down" is the first hypothesis when the panel will not open, and it is the least frequent. In most cases the problem is local, and spending ten minutes confirming it avoids an hour of waiting for something that will not fix itself.

Confirm before concluding

Three tests, in order, in under two minutes:

1. Another browser. Open Ads Manager in a private window or another browser. If it works, the cause is an extension or local cache.
2. Another network. Switch from Wi-Fi to mobile data. If it works, it is your network or its DNS.
3. Another device. Your phone, if the computer fails.

If all three fail, then it is worth looking for external information. If any works, you already found the cause and it is yours.

Where to check status

The platform has a business status page inside its help centre, which logs acknowledged incidents. It is the official source and tends to be conservative: small incidents take time to appear.

Useful complements: collective outage report sites and a search for recent mentions on social. When many people report the same thing in the same window, it is a real outage.

The signal worth more than all of them: other accounts, from other clients, on other networks, with the same symptom. If you manage several accounts, that is the fastest check there is.

A down panel is not stopped delivery

A confusion that produces wrong decisions: Ads Manager and the delivery system are separate.

The most common incident is being unable to see numbers while ads keep running, spending and converting normally. There is nothing to do but wait.

The opposite also exists, and is rarer: a working panel with stalled delivery. There the numbers show zero impressions with an active status — which is the case in Meta Ads not delivering, not an outage.

Knowing which of the two you are in determines whether there is anything to do or only to wait.

What not to do during an outage

Do not pause everything by reflex. Pausing and resuming restarts ad set learning, and that costs more than a few hours without visibility. The explanation is in Meta Ads learning phase.

Do not duplicate campaigns. With an unstable panel, duplicating can create partial objects you will have to clean up.

Do not conclude anything about performance. Metrics collected during instability are usually reprocessed in the following hours. The number you saw at 2pm can change by 6pm.

Do not resubmit the same request five times. True for publishing ads and for review requests: repeated attempts during instability create duplicates.

When pausing does make sense

There is one case, and it is specific: evidence of abnormal spend.

If the panel returns partially and shows spend far above the configured daily budget, or delivery concentrated somewhere you did not configure, pausing is correct. Losing the learning beats continuing to spend without control.

Outside that, inertia is the right policy. Most outages last less time than it would take you to reconfigure everything.

What to do with budget while you wait

If the outage runs past a few hours and you have delivery to protect:

Google Ads absorbs existing demand without depending on the same infrastructure. It does not replace Meta, and it covers the people already searching. The difference is in Meta and Google Ads in one dashboard.

TikTok Ads is a feed alternative, with the caveat that it needs its own creative — reusing the Meta piece performs worse, as covered in TikTok Ads vs Meta Ads.

For a few hours, none of that is worth the effort. For a full-day incident, it is.

How to tell the client

A platform outage is one of the few situations where telling someone before they ask genuinely builds trust.

A short message covers it: what happened, whether delivery was affected or only visibility, what you did, and when you will report again. No jargon, and no promising a timeline that is not yours.

What destroys trust is silence followed by a report with an odd day and no explanation. The structure for that conversation is in presenting results to a client.

After it comes back

Two checks are worth doing, and nobody does them:

The period's spend. Compare spend during the unstable hours against the configured budget. A large discrepancy deserves investigation, and the record needs making while the data is fresh.

The period's conversions. If the pixel was also unstable, you may have lost events. The number will rise in the following hours with reprocessing, and some may not return.

And do not judge the outage day's performance. It represents nothing, and including it in a weekly comparison contaminates the whole week.

When the problem is only yours

If the initial tests showed the problem is local, the most frequent causes, in order:

1. A browser extension, especially ad and tracking blockers. Ads Manager uses resources they block.
2. Old cache and cookies after an interface update.
3. An expired session on an account with multiple business profiles.
4. A VPN or corporate network filtering domains.
5. Removed permission in the portfolio, which appears as an empty screen rather than an error message.

Item 5 is the most deceptive: the interface loads and shows nothing, which looks like an outage.

The two-minute routine worth having beforehand

Anyone managing several accounts gains from a routine that answers "is it me or is it them?" without depending on an external site.

What it needs:

Two different clients' accounts open. If both fail, it is the platform. If one fails, it is permission or configuration on that account.

One reading through the API. If the interface will not open and the API responds, it is a front-end problem and delivery continues. That is the most useful distinction and the least available to anyone using only Ads Manager.

Yesterday's spend written down. Having the reference number to hand lets you compare the moment the panel returns, instead of reconstructing it later.

None of that is a tool: it is a habit. And the return shows up the first time you answer the client in two minutes instead of half an hour.

The summary

Test another browser and another network before concluding it is down. Confirm on the status page and in collective reports. Assume delivery continues until you have evidence otherwise. Do not pause by reflex, do not decide anything on data from the window, and tell the client before they ask. The reporting routine is on the paid media reporting page.