Most paid media audits find the same short list of structural problems: campaigns bidding against each other, too many campaigns splitting conversion data too thin, budget locked in the wrong container, settings inherited from a build nobody remembers, and naming that makes clean reporting impossible. None of them show up as an error. The account runs, the reports render, and the money goes out the door anyway.
Structure is the layer that decides what everything else is allowed to do. Automated bidding can only optimise inside the containers you built for it. Reporting can only group what your names let it group. When the structure is wrong, every fix downstream of it is a workaround.
Here is what turns up most often, what each one costs, and how to look for it in your own account.
Campaigns that compete with each other
The most common finding, and the one people are most surprised by. Two campaigns targeting the same keywords, the same audience, or the same geography are not two shots on goal. They are one budget split in half, with both halves learning from half the data.
It happens for ordinary reasons. Someone launches a test campaign and never turns off the original. A new region gets its own campaign that duplicates terms already live in a national one. A prospecting audience quietly overlaps with a retargeting pool.
What to look for: sort campaigns by spend and read the targeting side by side. In Google Ads, run the search terms report across the whole account rather than one campaign at a time, and watch for the same term drawing spend in two places. In Meta, check audience definitions for overlap before you check anything else.
Too many campaigns, not enough data in any of them
Automated bidding is a learning system, and learning needs volume. A campaign producing six conversions a month never gets out of the guessing stage. Ten starved campaigns reliably perform worse than three well fed ones.
The instinct that creates this is a good one. Marketers separate campaigns to see performance clearly. The cost is that the separation you built for reporting starves the bidding. You can get that visibility back from labels, naming, and segments in reports, without paying for it in performance.
What to look for: conversions per campaign per month. Anything consistently in single digits is a consolidation candidate.
Budget sitting in the wrong container
Budget is set at a level above the thing that actually performs. The campaign controls the money; the ad group or ad set does the work. So a campaign holding one strong ad group and four weak ones spends most of its budget on the weak ones and reports an average that hides both.
This is also how spend gets stranded. Budget capped on a campaign that could take more, sitting next to a campaign that cannot spend what it has. The account looks fully invested at the top line while the best performer is throttled all month.
What to look for: impression share lost to budget on your strongest campaigns, next to underspend on the weakest. If one is capped while another is coasting, the containers are wrong, not the bids.
Settings nobody chose on purpose
Every account carries settings from whoever built it. Search Partners and Display expansion left on in a search campaign. Location targeting set to presence or interest, so you pay for people who merely searched about your city. Audience expansion quietly widening a carefully built list.
None of these are wrong by default. They are wrong when nobody decided them. An audit separates the settings someone chose from the settings that came with the template.
What to look for: open one campaign, write down every non-default setting, and ask who chose it and why. Repeat for the next highest spender. Two campaigns in, the pattern is usually obvious.
Names that make reporting impossible
Structure problems and reporting problems are the same problem viewed from two ends. When campaigns are named four different ways, no filter finds all of them, and no blended report across Google, Meta, and TikTok lines up without a spreadsheet and an hour of hand mapping. Our campaign naming convention guide covers the tokens and the migration path.
When you run paid media across many locations
For multi-location brands the structural questions get sharper. Do location campaigns overlap on geography or terms, so your own locations bid each other up? Does the rollup report hide the fact that three locations carry the whole result? Is budget allocated by location size or by what each market can actually absorb?
The reporting side of this is covered in multi-location reporting both corporate and locations trust. The structural side is a paid media audit question, because it takes a read across every account at once rather than one at a time.
What these problems actually cost
Structure problems are expensive because they compound quietly. Overlapping campaigns inflate your own costs. Thin campaigns keep bidding from ever learning. Stranded budget buys nothing. And every one of them corrupts the data you use to make the next decision, so the wrong structure produces the wrong conclusions, which produce the next wrong structure.
That is also why structure gets fixed before tactics. Rewriting ad copy inside a broken container is effort spent where it cannot compound. If you want the full sequence, the paid media audit checklist walks through it step by step, and how we run paid media explains where structure fits in a build.
Related questions
Can you fix account structure without starting over? Usually. Consolidation and settings cleanup happen in place. A full rebuild is for accounts where the naming and the containers are both unusable, and it is planned so history is preserved.
Will consolidating campaigns reset the learning phase? Changes to budget and structure can reset learning, which is why consolidation is done deliberately and in a sequence rather than all at once. Short-term volatility for a structure that can actually learn is a trade worth making.
Is structure or tracking the bigger problem? Tracking, when both are broken. Bidding toward a wrong conversion signal makes good structure efficient at buying the wrong outcome. Start with the signs your conversion tracking is wrong.
Start with a diagnostic
Bring what you found. Overlapping campaigns, thin conversion volume, settings nobody chose. We will tell you which of them is costing the most and what order to fix them in.