Meta Ads Audit: 10 points you can check yourself in your ad account

Why most ad accounts end up in the yellow zone

We see external ad accounts every week. Companies that contact us get an audit before every call, and most of these accounts end up in the yellow or red zone. Not because the teams do a bad job, but because an ad account grows over years: new campaigns are added, old ones are left lying around, tracking is set up once and never touched again. At some point, the algorithm optimizes for an account that no one completely understands anymore.

The good news: you can check the ten points that stand out most frequently in audits yourself. All you need is the Ads Manager, the Events Manager, and one hour of your time.

1. Is your budget really running where you think it is?

The "Active" status says nothing. What matters is which campaigns, ad sets, and ads have actually spent budget in the last 30 days. Filter by spend instead of status and count the units with real spend. If your budget is spread across twelve ad sets, eight of which are under 50 euros, none of them will learn. The algorithm needs volume per unit, not variety in the structure.

2. Are your ad sets bidding against each other?

Two ad sets with similar target audiences and the same goal compete in the same auction. You then pay for your own competition. Check how many ad sets share the target audience and optimization event, and consolidate where there is no content-related reason for the split.

3. How much budget goes to goals without revenue?

Traffic and reach campaigns drive up the account's click-through rate and look good in reporting. They rarely contribute to revenue. Calculate the share of your budget in the last 30 days that went to campaigns that did not deliver tracked purchases or leads. In many accounts, this share is around one third. This is the first lever for blended ROAS.

4. Are there paused winners?

Sort all campaigns from the last 90 days by ROAS or cost per lead, including the paused ones. Almost always you will find a campaign with excellent metrics that someone paused for a long-forgotten reason. Every day it stands still costs profitable revenue. Reactivate, and then scale in a controlled manner, no more than about 20 percent budget increase per day so that the learning phase does not break.

5. How diverse are your creatives really?

Meta recognizes similar ads and treats them as one. A new headline on the same image is not a new ad, but a duplicate. Count the creatives per ad set that differ conceptually: different hook, different angle, different format. As a guideline, we work with 8 to 20 creatives per ad set, at least four visual styles, and at least two formats. Anyone below this gives the system too little to learn. Why this has become so important is explained in our article on Andromeda.

6. Frequency and fatigue

A frequency above 3.5 combined with a falling click-through rate is the classic picture of creative fatigue; from 5 onwards, it becomes critical. Typically seen in retargeting. The solution is not more budget, but fresh creatives and a frequency cap.

7. Tracking: Pixel, Conversions API, Event Quality

Open the Events Manager and check three things. First: Is the Conversions API running parallel to the pixel, or does everything rely on the browser? Second: How high is the Event Match Quality of your purchase event? Values below 6 mean that a large portion of your purchases are not being attributed. Third: Are events being counted twice because the pixel and server are not deduplicated? Without clean events, the algorithm optimizes for the wrong thing, no matter how well the rest is set up.

8. What does your ROAS consist of?

A ROAS of 4 can be excellent or worthless, depending on how it is made up. In your reporting, toggle the attribution windows next to each other: 1-day click, 7-day click, 1-day view. If the majority of your revenue comes from view attribution, your account is telling a reach story, not a performance story. Establish one window as a binding valuation basis and stick to it.

9. Where does the funnel break after the click?

Build the chain completely once: impressions, link clicks, landing page views, cart, checkout, purchase. The biggest drop-off shows you where the problem lies. If it drops between the landing page and the cart, the issue is the offer, product page, or traffic quality, not the creative. Many teams optimize ads even though the product page is the problem.

10. The Breakdown Trap

The most common mistake in audits we see in retrospect: A placement, an age group, or a device shows higher average cost per conversion and is paused. This sounds logical and is usually wrong. Meta optimizes for the cost of the next conversion, not the average. A segment with a low average can have steeply rising marginal costs, and that is exactly why the system pushes budget to the supposedly more expensive segment. Pausing it makes the overall result more expensive. Rule: Never pause based on breakdowns alone. Every change is a hypothesis that you measure against the overall performance of the account.

What you do with the result

If you stumble on more than three points, that is no reason to worry, but rather the normal case. Prioritize in this order: first tracking, then structure and goal mix, then creatives. Everything else builds on that. And if you want a second opinion, we do exactly this audit before every initial consultation, using numbers from your account instead of slides.