Why are my Meta Ads stuck in Preparing?

Quick answer: Preparing happens before an ad is ready to deliver

When a Meta ad shows Preparing, the platform is still processing a new publication or a material edit into a version that can move through review and delivery. A short stay in this state can be normal, especially after uploading video, changing creative, duplicating an ad, or publishing several items at once. It does not yet prove that the ad was rejected, that the campaign lacks budget, or that the auction cannot spend.

Start at the ad level and record when the last publish or edit completed. Confirm that the campaign, ad set, and ad are switched on, but do not repeatedly edit or duplicate the ad while it is processing. If the status persists beyond the account's normal publishing time, isolate whether the blocker is the asset, the publish operation, an account dependency, or a wider platform incident.

• User symptom: a newly created, duplicated, or edited ad remains in Preparing and receives no delivery.
• Control level: the individual ad, its creative assets, publish operation, identity, destination, and account dependencies.
• Time grain: minutes or hours since the most recent successful publish, not a multi-day performance trend.
• Likely mechanism: media processing, incomplete publishing, asset validation, permission sync, or platform-side delay.
• Decision produced: wait, repair one dependency, republish once, replace one asset, or escalate with evidence.

Keep Preparing separate from review and delivery problems

Preparing is a pre-delivery processing state. It is different from In review, where Meta is evaluating the ad against policy; Active but not spending, where an eligible campaign is failing to win or enter enough auctions; and Learning Limited, where an ad set is delivering but lacks enough optimization events. Treating these as one generic no-delivery problem leads to unnecessary budget, bid, and targeting changes.

The practical boundary is the last confirmed stage. If the ad has not completed preparation, investigate publication and creative processing first. If it reaches review, switch to policy and review diagnostics. If it becomes active but receives no impressions, move to account, audience, schedule, bid, budget, and auction constraints.

• Preparing: Meta is processing the published ad and its dependencies.
• In review: the prepared ad is undergoing policy review.
• Rejected: a policy decision prevents delivery until resolved.
• Active but not spending: preparation and review completed, but delivery is blocked or constrained.
• Learning or Learning Limited: delivery exists, but optimization is still stabilizing or underfed.

Preserve the timeline before touching the ad again

Record the ad ID, campaign and ad set IDs, current status at every level, account time zone, last edit time, publish time, and the person or tool that made the change. Capture the exact Delivery-column message and any warning in the ad editor. If an API, bulk tool, automated rule, or agency workflow published the ad, preserve its job or request result too.

This timeline matters because every new edit can create another version that must prepare. A team that keeps changing copy, replacing media, toggling status, or pressing publish may continually restart the process and erase evidence about the original failure. Freeze the affected ad while you compare it with a healthy control.

• Save the immutable ad ID rather than relying on the ad name.
• Note the exact timestamp and actor for the last material edit.
• Capture whether Ads Manager reported the publish as complete, partial, or failed.
• Check campaign, ad set, and ad status without changing them.
• Identify whether one ad or many unrelated ads entered Preparing at the same time.

Test whether the problem follows the creative asset

If only one ad is affected, compare its creative package with a recently prepared ad in the same account. Inspect image or video format, file size, aspect ratio, duration, thumbnail, destination URL, call-to-action configuration, tracking parameters, catalog references, and the Page or Instagram identity. A damaged upload, incomplete video processing, unavailable post, or mismatched asset dependency can hold one ad while the rest of the campaign is healthy.

For video, confirm that the uploaded media can preview fully rather than showing a placeholder or processing error. For an existing-post ad, verify that the source post still exists and that the ad account retains access. For catalog or dynamic formats, confirm that the referenced catalog, product set, and required fields are available. Change only the dependency that fails this comparison.

• Preview every placement and confirm the media resolves completely.
• Check the destination URL and any required domain or app association.
• Verify Page, Instagram account, catalog, product set, and post permissions.
• Compare file type, dimensions, duration, and thumbnail with a known-good ad.
• Look for one missing or stale asset reference before replacing the whole ad.

Distinguish an incomplete publish from a platform delay

A visible ad row does not always mean every draft change reached Meta successfully. Reopen the ad and check for unpublished changes, failed items, red validation messages, or a publish summary that shows only part of the batch succeeded. When a large bulk publish contains mixed outcomes, isolate the affected ad instead of re-publishing the entire campaign.

If several unrelated ads across campaigns become stuck at the same time, the likelihood shifts toward an account-wide or platform-side delay. Check account notifications and official platform status information, compare with another ad account if you legitimately manage one, and avoid destructive troubleshooting while a broad incident is plausible. Preserve screenshots and timestamps so support can distinguish a persistent record from a transient outage.

• One affected ad usually points to its publish record, media, identity, or destination.
• One affected batch can point to a partial bulk-publish failure.
• Many unrelated ads in one account can point to permissions, account state, or queueing.
• Multiple accounts affected at once can indicate a broader platform incident.
• A completed publish receipt is stronger evidence than the presence of a draft row.

Use a controlled recovery sequence

After a reasonable observation window, choose the smallest reversible action supported by the evidence. First resolve any explicit validation or permission warning. If the original publish was incomplete, publish the existing ad once. If one media asset will not finish processing, replace that asset with a validated file while keeping the rest of the setup stable. Duplicate the ad only when the original record remains stuck after its dependencies are proven healthy and the duplicate will serve as a controlled test.

Do not duplicate repeatedly, rebuild the campaign, raise budget, broaden targeting, or change optimization merely to force a Preparing ad to move. Those changes do not repair media processing and can create duplicate delivery, fragment social proof, restart learning, or make the eventual root cause impossible to identify. Pause or archive abandoned test copies so only one intended version can deliver.

• Fix an explicit asset, URL, identity, catalog, or permission error first.
• Republish once when the recorded publish operation was incomplete.
• Replace only the failed media component when the issue follows the asset.
• Create one controlled duplicate only after the original setup is documented.
• Escalate with IDs, timestamps, screenshots, and tested dependencies when the state persists.

Verify recovery at each stage

Recovery is not complete merely because Preparing disappears. Confirm that the ad advances into review, receives a policy outcome, becomes eligible, and then records impressions. Keep campaign and ad set settings stable while this happens so the test answers one question: whether the ad itself can complete the publication-to-delivery path.

Once impressions begin, verify that the intended creative, identity, destination, tracking parameters, Pixel or app event, and placement previews match the approved version. If a duplicate became the live ad, prevent both versions from serving accidentally and document which ID should remain in reporting.

• Preparing completed and the ad moved to the next explicit state.
• Review completed without an unresolved policy block.
• The intended ad ID received impressions and spend.
• Creative, destination, identity, and tracking match the approved setup.
• Temporary duplicates or drafts cannot create duplicate delivery.

How an AdSpecIt-style audit helps diagnose a stuck ad

An AdSpecIt-style audit can place the stuck ad inside the wider account state: campaign and ad set status, recent edits, creative formats, destination consistency, identity dependencies, review outcomes, delivery diagnostics, spend, impressions, and tracking configuration. That context prevents a single processing issue from being mistaken for a budget, targeting, or account-wide performance problem.

The useful output is a stage-by-stage decision path: preserve the publish timeline, prove where progress stopped, compare the affected ad with a healthy control, test its dependencies, apply one recovery action, and verify impressions on the intended ID. Related diagnostics for stuck review, rejected ads, active campaigns that do not spend, and learning limitations should begin only after the ad leaves Preparing.