When an ad stops delivering, do not rush to duplicate and relaunch it. Check the account, ad, creative, audience and bidding, then the landing page and pixel; only change campaign structure after those layers are ruled out.
When ads stop performing or delivering, the most common reaction is to duplicate one and relaunch it. Doing that too early can be costly: existing learning data, review history, and audience feedback are reset while the original problem remains. Troubleshooting should follow a sequence because clearing each layer narrows the scope of the next one.
First identify which layer the problem is in
Use three quick descriptions to classify the symptom.
An account-level problem usually means all ads under the account are struggling in the same way, possibly alongside restrictions, unpaid balances, or failed payments.
An ad-level problem means the account itself looks normal and can spend budget, but one ad stays in review for a long time or appears active without delivering.
A creative-level problem shows up when an ad is rejected, rated as low quality, or delivers but produces clearly abnormal click performance.
The key is scope: if every ad is affected at once, go back to the account level. If only one ad is affected, inspect that ad itself.
Clear the account-level issues first
- Account status and permissions: Is the account restricted, or is there an appeal that still needs to be submitted?
- Payment method: Is there an outstanding balance, a failed charge, or a spending limit that has been reached?
- Asset relationships: Are the Page, pixel, or domain restricted, or do they still require verification?
Until these issues are resolved, changing creatives or audiences will not make a difference because the ad does not have a normal chance to enter delivery.
At the ad level, check review and settings
- Review status: Has the ad been stuck in review, or was it rejected and never resubmitted?
- Delivery status: Can the budget actually spend, or are scheduling settings or overly narrow targeting preventing delivery?
- Level settings: Are the campaign objective, conversion location, or optimization event wrong? If the optimization event does not match the real conversion, the system will look for the wrong people. You may see volume, but not from the audience you actually want to convert.
At the creative level, check rejection and clicks
If an ad is rejected, start with the stated reason. Determine whether the issue is a content violation or a category or regional restriction, then change it according to the guidance instead of repeatedly submitting the same version.
When an ad is rated as low quality, the creative itself may not show an obvious problem. The trigger is often in the copy or the landing-page experience.
Click performance is also useful at this layer. A low click-through rate suggests the opening is not compelling enough or the value proposition is unclear. If click-through rate is normal but conversion rate is low, the creative has already done its job. Continuing to change it is unlikely to help, so move down the funnel.
Audience and bidding
- Reach is too small: If delivery is weak, first consider broadening the audience.
- Audience overlap: Multiple ad sets competing for the same people effectively push up your own costs.
- Frequency is too high: If the same people keep seeing the ad, the problem is often a pool that is too narrow rather than a weak creative.
- Bid and budget: A bid that is too low can prevent delivery, while an overly concentrated budget can force spend into a small segment of the audience. Change only one setting at a time. If you change the audience and bid together, you will not know which change caused the result.
Landing page and pixel
- Load speed: Slow loading directly reduces conversions, especially on mobile.
- Promise consistency: Does the page say the same thing the ad promises?
- Conversion path: Form fields, sign-up steps, and redirects all create leakage; every extra step can lose people.
- Pixel and conversion events: Are events firing correctly, are any counted twice, and does the optimization event match the real conversion? If the pixel is configured incorrectly, the conversion numbers in the report are already wrong, and every later optimization decision follows bad data.
Leave structural changes until last
Only after the earlier layers have been ruled out should you change the account structure by splitting ad sets, reallocating budget, or rebuilding campaigns. Rebuilding restarts the learning phase. It is a last resort, not the first move. Using it first usually wipes out useful clues and then reproduces the same result.
Add one more check when running multiple accounts
If you manage several advertising accounts at the same time, also check whether creatives are being reused unchanged across accounts. Repeating the same creatives, similar time windows, and identical audiences across multiple accounts can look like bulk activity from the same operator and may affect account health.
Give each advertising account its own stable login environment, and differentiate creatives across accounts by changing copy structure, visual style, or posting times. These steps can reduce that risk. In parallel multi-account scenarios like this, PurpleMark provides environment isolation and permission separation.
