Most people audit a Meta account backwards. They open Ads Manager, sort by cost per result, find the worst advert and start rewriting it. An hour later they have a new advert and the same account, because the thing holding the account back was two levels above the creative.
I run Growth Bully, a Malta performance marketing agency, and the order below is the one I actually work in when I open an account I have never seen. It is not the order that feels productive. It is the order that stops you spending the hour on the wrong layer. Nothing here is held back, because the people who do this properly themselves are not the people who were going to hire anybody.
Can you really audit a Meta account in an hour?
You can reach a verdict in an hour. You cannot fix it in an hour, and you should not try. The hour produces one output: a short list of findings ranked by whether they break the account or merely slow it down. Repairs are a separate session, and mixing the two is why most self-audits never finish.
Budget it as fifteen minutes on structure, fifteen on tracking, twenty on creative and ten on audiences. If a step overruns, write the finding down and move on. The sequence matters more than the depth, because a shallow pass in the right order beats a deep pass in the wrong one.
What is the difference between a broken account and an underperforming one?
A broken account is producing numbers that are not true, or is not doing the thing you believe it is doing. An underperforming account is working correctly and losing. The first needs a repair and the second needs a better argument, and no amount of creative work fixes the first.
That distinction is the entire point of auditing in this order. Structure and tracking tell you whether the account is broken. Creative and audience tell you why a working account is losing. Running those checks the other way round means you judge creative on numbers you have no reason to believe.
What do you check first, and why is it the structure?
Because the structure tells you what the account is actually doing, which is often not what its owner thinks. Fifteen minutes: list every active campaign, its objective, where the budget sits, and when each one last changed. You are building a picture of what is live, not grading it yet.
What I am looking for in that fifteen minutes:
- Objective mismatch. A campaign optimising for traffic or engagement while the owner judges it on enquiries. This single error explains more disappointing accounts than anything else.
- Too many live campaigns for the budget. Spend sliced so thin that nothing accumulates enough data to be readable.
- Overlapping campaigns chasing the same people. You are bidding against yourself and paying the difference.
- Last edited dates. An account touched daily never finishes learning. An account untouched for months has stopped being managed.
- Anything paused that nobody can explain. Often the thing that used to work.
Write down what is live in one line per campaign. If you cannot do that, the structure is the finding.
How do you tell whether the numbers can be trusted?
Fifteen minutes on the plumbing, before you form any opinion about performance. Check that the pixel and the server side connection both fire, that a conversion event registers when you complete the form yourself, that the event you are optimising for is the one you care about, and that there is only one of it.
The specific things that mean broken rather than weak:
- No conversion event registers on a real submission. Test it yourself, end to end, on a phone. If nothing lands, every number above it is decoration.
- Two events counting the same action. Conversions roughly double and your cost per result looks excellent for no reason.
- The optimisation event is a page view or a click. The platform will dutifully buy you the cheapest version of the wrong thing.
- Lead events firing on page load. Common after a site rebuild, and it quietly destroys the audience signal as well as the reporting.
- No consistent campaign tagging. If your analytics and your CRM cannot agree on where an enquiry came from, nobody can settle an argument later.
My team keeps the implementation detail at thegrowthbully.com/bytes/conversion-tracking-that-survives, and the reason your platform number will never match your analytics number, even when both are correct, is at thegrowthbully.com/bytes/ga4-vs-meta-numbers. Expect a gap. Worry about a gap that moves every month.
What are you looking for in the creative?
Twenty minutes, and you are counting arguments rather than judging taste. Sort by spend, look at what has consumed most of the budget, and ask how many genuinely different reasons to act are present. Three crops of one image is one argument, not three, and an account with one argument has nowhere to go.
Then check two things that are easy to see and hard to admit. First, whether the best performer is also the oldest, which tells you nothing new has been made in a while. Second, whether the first three seconds of your video say anything specific, because that is where most of the budget is lost. Hooks decay on a schedule, and my team has the pattern at thegrowthbully.com/bytes/hooks-decay.
A broken signal here is narrower than it looks: almost all spend sitting on one advert while everything else was switched off within days. That is not a winner, it is an account with nothing to compare against. For the testing structure that prevents it, see my creative testing framework.
What is worth checking in the audiences?
Ten minutes, and far less than people expect, because audience construction is where self-auditors want to spend the hour and it is rarely the cause. Check your exclusions, check that retargeting is not being counted as prospecting, and check nothing is so narrow that delivery is strangled.
In a small market the audience dial is nearly flat anyway. What still matters is that you are not paying prospecting prices to reach people who already know you, and that your existing customers are excluded from campaigns written for strangers. Both are quick to see and both are worth fixing before anything clever.
What do you do with the list at the end?
Sort it into broken and underperforming, then do all the broken items first, even the boring ones. Tracking repairs before creative rewrites, always, because until the measurement is honest you cannot tell whether a change helped. Then change one thing at a time and give it long enough to report.
If the audit produced nothing in the broken column, the account is working and losing, which is a better problem. That is a message and offer problem, and twelve specific things to check when ads are not converting is the companion piece to this one.
The structured version of everything above, as thirty items you can tick off rather than remember, is the 30-Point Meta Ads Health Check. It is free and it is the same sequence. And if you would rather have the account run properly than audited twice a year, that is what my team does at Growth Bully.
Frequently Asked Questions
How do I audit my own Facebook ads account?
Work in one order and give yourself an hour: fifteen minutes on campaign structure, fifteen on tracking, twenty on creative, ten on audiences. The goal is a ranked list of findings, not repairs. Checking creative before you have confirmed the conversion data is real wastes most of the hour.
How long should a Meta ads audit take?
An hour to reach a verdict on a single account, and a separate session to fix what you found. Trying to repair as you go is why most self-audits stop halfway. A shallow pass in the right order tells you more than a deep pass that never reaches the tracking.
What does a broken ad account look like?
No conversion event registering on a real form submission, two events counting the same action, optimisation set to clicks or page views instead of enquiries, lead events firing on page load, or campaigns optimising for a different outcome than the one you judge them on. All of these make the reporting untrue.
Why check tracking before creative?
Because creative decisions are made from numbers, and broken tracking produces numbers that point the wrong way. Duplicate events make a weak advert look efficient. A missing event makes a working advert look dead. Fix the measurement first and the creative verdict changes on its own.
Do I need a tool to audit a Meta account?
No. Ads Manager, your website on a phone to test the form, and your CRM or enquiry inbox to check that the leads are real. The value is in the sequence and in writing findings down, not in software. A checklist is useful because it stops you skipping the dull steps.



