An amazon ads audit is a diagnosis, not a repair. The temptation when you open an unfamiliar account is to start fixing: pause the obvious waste, raise the bids on the winners, restructure the campaigns that look untidy. We do not, for a simple reason. Until you understand why the account looks the way it does, you cannot tell the difference between a mistake and a deliberate decision made for a reason nobody wrote down. Accounts are full of both, and they look identical from the outside.
So our first week is reading. We want to arrive at a short written list of findings, each one with evidence attached and an estimate of what fixing it is worth. That document is the deliverable. Whoever acts on it, us or the in-house team, then works from a shared picture rather than from whoever has the strongest opinion in the meeting. Most of the value in an audit comes from that agreement, not from the individual findings.
Step one: does the account know what it is trying to do
We ask what the advertising is for before looking at a single number. Defending share on established products, launching new ones, clearing stock and pushing into a new category are four different jobs with four different acceptable cost structures. Accounts get into trouble when all four run under one target because the target was set for one of them years ago. Half the findings in a typical audit trace back to this rather than to any technical fault.
The second framing question is which products actually matter. Catalogues are usually top heavy, and advertising attention rarely matches. We rank products by contribution, not revenue, and check how spend is distributed against that ranking. Finding meaningful budget on products that cannot be profitable at any volume is common, and it is usually the largest single item the audit produces.
Step two: structure, and what it prevents
Campaign structure is a means of control. The only real question is whether the current structure lets you change one thing without disturbing another. When products with different margins share a campaign, you cannot set a sensible target for either. When discovery and proven terms share a budget, discovery gets starved on good days and overfed on quiet ones. We map what is grouped with what, then note every place where a change to one product would force a change to another.
We also check for the duplicates. Accounts accumulate campaigns built by different people for the same products, and they compete for the same placements. This rarely shows up as a dramatic problem in the reporting. It shows up as a gradual drift in cost that nobody can attribute, which is exactly the kind of thing an audit exists to catch.
Step three: the read-only amazon ads audit checklist
The middle of the amazon ads audit is mechanical. We work through a fixed list so that nothing depends on what happened to catch our eye, and so two audits of the same account produce the same findings.
- Search term reports for the last full quarter, sorted by spend, with anything that has cost money and returned nothing flagged rather than paused.
- Negative keyword lists, checked for terms that were negated too broadly and are now blocking relevant traffic.
- Budget caps, and which campaigns are hitting them consistently enough that the data is being shaped by the cap.
- Out of stock periods lined up against performance drops, because a large share of unexplained decline is inventory, not advertising.
- Placement and targeting settings that were changed once and never revisited.
That last point about inventory earns its place. We have opened accounts where the entire reported collapse in performance matched a stock gap that nobody connected to the numbers, and where three months of increasingly aggressive bidding had been spent trying to fix a problem that had already resolved itself. Checking supply before diagnosing demand is cheap and saves a great deal of misdirected effort.
Step four: the pages behind the ads
An audit that stops at the campaign level will produce a list of bidding recommendations and miss the constraint. We take the products carrying the most spend and look at them as a buyer would: how the main image reads at thumbnail size, whether the first bullet answers the obvious question, where the price sits against what surrounds it in search results, and what the recent reviews say. If traffic is right and conversion is weak, the finding belongs in the audit even though it is not an advertising fix.
What the audit produces
The output is ranked by expected value, not by how easy each item is. Three or four things usually account for most of the available improvement, and a long list of small tidying items accounts for very little. Presenting them as equals is how audits turn into busywork. We separate them explicitly so the first month goes to the items that matter and the rest waits.
We also record what we deliberately left alone. If a campaign looks inefficient but is the only thing defending a product against a competitor, that context belongs in the document, otherwise someone will optimise it away in six months and nobody will remember why it existed. An amazon ads audit is most useful when it survives the people who commissioned it.
Keep reading: Amazon Ads · Amazon