An Amazon PPC audit checks nine account-health metrics against the benchmark range for your strategic phase, then hunts the structural leak behind every metric that fails: wasted spend, search term bleed, placement premiums that outrun their return, auto campaigns holding budget they should have handed off, and branded sales propping up the average. This is the exact checklist I run when a seller sends me a bulk file, in the order I run it, written so you can run it yourself. I built my free tools as the automated version of this same procedure, so nothing here is held back. Work through it manually and you will find your leaks. Or upload the same file to the free Account Health Snapshot and get the nine-metric scorecard in about 60 seconds. Either way, the checklist is identical.
Before you start: pull one 60-day bulk file
Every check on this list runs off a single file: the bulk operations spreadsheet from Seller Central, covering the last 60 days. You need the Sponsored Products Targeting and Keywords tabs (Enabled and Paused), Placement Data, and the Search Term Reports for Sponsored Products and Sponsored Brands. Add Sponsored Display data if you run it.
Sixty days is the window I use because it is long enough to judge conversion behavior on most targets and short enough that the account it describes still exists. Twelve-month exports blend four seasons and two strategies into one misleading average. If you have never pulled this file, my Amazon bulk file guide walks through the download click by click; it takes about five minutes. One more input before you grade anything: know your strategic phase. Awareness, Market Share, Growth, and Profitability carry different passing scores for every metric below, and grading a launch-phase account against Profitability numbers produces panic instead of information.
The checklist: six groups, nine metrics
The audit covers six groups: spend efficiency, targeting hygiene, structure, placement, discovery, and brand. Between them they produce the nine metrics I grade on every account: account ACOS, Top-of-Search ACOS, wasted spend %, conversion rate, branded vs non-branded split, auto-campaign share, budget concentration, revenue concentration, and spend/revenue alignment. Run the groups in order; each one narrows the next.
| Metric | What it grades | A failing score usually means |
|---|---|---|
| Account ACOS | Overall ad efficiency vs your phase target | Any of the leaks below; this is the symptom, not the cause |
| Wasted Spend % | Share of spend on zero-order targets above target CPA | Missing negatives, stale targets left running |
| Conversion Rate | Clicks that become orders | Irrelevant traffic, or a listing problem PPC cannot fix |
| Top-of-Search ACOS | What the premium placement costs vs returns | Placement modifiers set on hope instead of math |
| Auto-Campaign Share | Spend still in discovery vs your phase ceiling | A broken harvest loop; auto campaigns hoarding proven terms |
| Branded vs Non-Branded Split | How much of ad sales your own brand name drives | Branded spend flattering the average, unseparated structure |
| Budget Concentration | Spend share held by your top 3 ASINs | Budget piled on favorites while the catalog starves, or spread too thin |
| Revenue Concentration | Ad-sales share from your top 3 ASINs | The account's income riding on too few products |
| Spend/Revenue Alignment | Whether spend distribution tracks sales distribution across ASINs | Budget pointed at ASINs that are not producing the matching sales |
1. Spend efficiency: ACOS, wasted spend, conversion rate
Start with three numbers: account ACOS against your phase target, the percentage of spend on targets with zero orders and spend above your target CPA, and account conversion rate. These three tell you how big the problem is before you know what the problem is, and they decide how urgent the rest of the audit becomes.
- Account ACOS vs phase. Total spend divided by total ad sales, graded against your phase range (the full table with all four phases is in my Amazon PPC benchmarks guide). If ACOS fails, do not touch bids yet; the rest of the checklist finds out why. The bid-last logic is the core of how I lower ACOS.
- Wasted spend %. Sort targeting rows by spend, filter to zero orders, and total the spend above your target CPA. The free Wasted Spend Finder does this sort for you and ranks the offenders.
- Conversion rate. Orders divided by clicks, account-wide. A failing conversion rate with clean targeting is usually a listing problem, and I would rather tell you that now than after a structural rebuild.
2. Targeting hygiene: search terms, negatives, n-grams
Targeting hygiene asks whether the account is paying for the right searches. Read the search term report for terms that spend without converting, confirm negatives exist where those terms should have been blocked, and run n-gram analysis to catch the losing words that no single search term makes visible.
- Search term review. Work through the report looking for the four actions in my search term report guide: harvest, negate, bid down, and bleed. Terms with meaningful spend and no orders are your negation list.
- Negative coverage. Check that past negation decisions actually made it into the account at the right level and match type. The free Negative Keyword Finder builds the candidate list; my negative keywords guide covers exact vs phrase.
- N-gram patterns. A word like "cheap" can lose money across 80 different search terms while every individual term looks too small to act on. N-gram analysis totals performance by word; the free N-Gram Analyzer runs it on a pasted search term report.
3. Structure: architecture and search term bleed
Structure asks who owns each search term. Every term that matters should have exactly one campaign responsible for winning it; when broad and phrase targets in several campaigns surface the same term, the conversion data splits, every copy looks weaker than the truth, and bid decisions degrade. That is search term bleed, and it is the most expensive item on this list because it corrupts the data every other decision runs on.
- Architecture read. Can you tell, from campaign names and contents alone, what each campaign is for? A deliberate build follows a discovery-to-performance flow like the one in my campaign structure guide. If you cannot reconstruct the intent, neither can the person bidding.
- One owner per term. Pivot the search term report by term and count distinct campaigns converting on each. Your top terms appearing in three or four campaigns at once is bleed, not coverage.
- Ownership fix queued. For each bleeding term, note the intended owner (usually a manual exact campaign) and the negatives needed everywhere else. You are building the fix list, not executing it mid-audit.
4. Placement: is the Top-of-Search premium earning its keep?
Placement grading is one comparison: Top-of-Search ACOS against your phase ceiling. Top of search converts better and costs more, so the question is never whether to pay the premium but whether the conversion lift still covers it. A placement modifier set months ago and never re-checked is one of the quietest leaks in an account.
Pull the placement rows from the bulk file and compute ACOS by placement. If Top-of-Search ACOS blows past its ceiling while rest-of-search holds, the modifier is buying position the economics do not support. My Top of Search guide covers the modifier math, including when a high premium is genuinely correct: usually a ranking push in an early phase, priced deliberately and dated for review.
5. Discovery: auto share and the harvest loop
Auto campaigns are for finding terms, not for keeping them. Grade the share of total spend sitting in auto campaigns against your phase ceiling, then check the loop: converting search terms should move from auto into manual exact campaigns you control, and get negated in the auto so the same term is not paid for twice.
A high auto share in a mature account almost always means the harvest loop broke: the discovery engine kept the winners instead of handing them off, so proven terms sit at auto bids in auto placements with no owner. Cross-reference your auto search terms against your manual exact targets; every converting term in the first list missing from the second is an unharvested win. The full loop, including when auto share should fall as the account matures, is in my auto campaigns guide.
6. Brand and concentration: separation, split, alignment
Branded clicks convert people who already chose you, so they are cheap and they flatter every average they touch. The audit checks that branded and non-branded targets live in separate campaigns, that the branded share of ad sales sits under your phase ceiling, and that spend and revenue are not dangerously concentrated or misaligned across ASINs.
- Brand separation. If branded and non-branded terms share campaigns, your blended ACOS is unreadable: the account can look healthy while non-branded spend loses money behind the branded subsidy. Separate them before trusting any efficiency number.
- Branded share vs phase. A branded split above your phase ceiling means growth spend is quietly defending instead of acquiring.
- Concentration and alignment. Check the spend share and ad-sales share held by your top 3 ASINs, and whether each ASIN's spend share tracks its sales share. Budget flowing to products that are not producing the matching revenue is a leak no keyword-level fix will find.
How to score it: Healthy, Investigate, Critical, always against phase
Score each of the nine metrics one of three ways: Healthy (inside the range for your phase), Investigate (outside the range with a plausible explanation you should verify), or Critical (outside the range with dollars actively leaking). The one rule that makes the scoring honest: grade against your phase, never against an absolute number.
Here is what phase-grading looks like on real numbers. The sample dataset behind my audit demo runs $36,303 in spend against $111,058 in ad sales, which is a 32.7% ACOS, with 4,140 orders on 45,672 clicks, a 9.1% conversion rate. Graded as an Awareness-phase account (ACOS ceiling under 40%, conversion floor above 3%), both metrics pass. Graded as a Profitability-phase account (under 15% and above 10%), both fail. Same file, same numbers, opposite verdicts; the phase is not a footnote, it is the rubric. Then triage by dollars, not by count of red flags: one Critical wasted-spend score with real spend behind it outranks three Investigate flags on thin budgets.
The automated version: the same checklist in about 60 seconds
Everything above is exactly what my free tools automate. The Account Health Snapshot takes the same 60-day bulk file, parses it in your browser, and grades all nine metrics against your phase with a Healthy / Investigate / Critical flag on each, in about 60 seconds, with no email and no account. It is this checklist, mechanized.
For the row-level work, the free Audit Dashboard runs the deeper pass: it classifies every keyword into one of four health buckets using the Stratum methodology, so the negation candidates and the starving winners fall out of the data instead of out of your patience. If you want to see it before uploading anything, the demo account shows the whole thing running on real anonymized rows. I publish the manual checklist anyway, and keep the tools ungated, for the same reason: a seller who can verify my method is a seller who does not have to take my word for anything.
When the audit says get help
The audit hands you a decision rule, not just a diagnosis: if the recoverable leak is bigger than a management fee, paying for management is arithmetic; if it is smaller, keep self-managing with the free tools and I will say so on the call. The full framework for that decision is in my guide to whether Amazon PPC management is worth it.
Two published examples of what a failed audit turned into, both on my results page. One account graded fine on effort (campaigns built, bids adjusted) and failed on structure: search term bleed and spend on non-converting keywords. Rebuilt, its ACOS went from 39.66% to 27.02% in 90 days, worth $4,735 a month in profit with no budget increase. A $45K/mo skincare brand failed hardest on wasted spend, 41% at audit; concentrating spend on its three converting variants and rebuilding the negative structure cut that to 16%, worth $2,890 a month net of my $1,499 flat fee. Both leaks were sitting in the bulk file the whole time, findable with exactly the checklist you just read. What the fee buys, and how flat-fee pricing compares to a percentage of spend, is in my management cost guide.
Frequently asked questions
How do I audit my Amazon PPC account?
Download a 60-day bulk file from Seller Central, then grade nine metrics against your strategic phase: account ACOS, Top-of-Search ACOS, wasted spend percentage, conversion rate, branded vs non-branded split, auto-campaign share, budget concentration, revenue concentration, and spend/revenue alignment. For every metric that fails its phase range, trace the structural cause: missing negatives, search term bleed, placement premiums, or a broken harvest loop. Fix causes, not symptoms.
How often should I audit my Amazon PPC account?
Run the full nine-metric audit quarterly, and additionally whenever the account changes strategic phase, changes operator, or ACOS drifts several points without an obvious cause like seasonality or a price change. The lighter weekly work, search term review and negatives, is maintenance, not an audit. An audit re-checks the structure those weekly routines run inside, which is where the expensive problems hide.
What reports do I need for an Amazon PPC audit?
One bulk operations file covering the last 60 days does almost all of it: Sponsored Products Targeting and Keywords (Enabled and Paused), Placement Data, and the Search Term Reports for Sponsored Products and Sponsored Brands, plus Sponsored Display data if you run it. The bulk file beats dashboard exports because it shows every target with its bid, placement, and results in one dataset you can sort and total.
Can I audit my Amazon PPC account for free?
Yes. This checklist is the manual version and costs nothing but time. The automated version is also free: the Account Health Snapshot grades all nine metrics from your bulk file in about 60 seconds, parsed in your browser with no email and no account required, and the Audit Dashboard runs the row-level version. I keep both ungated because sellers who see their own numbers make better decisions, including the decision to keep self-managing.
How long does a manual Amazon PPC audit take?
Plan for two to three hours the first time you run this checklist on a 60-day bulk file: most of it goes to the targeting hygiene and structure steps, because those are row-level reads. Repeat audits go faster once you know the account. The free snapshot compresses the nine-metric scoring pass to about 60 seconds, which leaves your manual time for the part that actually needs judgment: tracing causes.
What is the biggest red flag in an Amazon PPC audit?
A large share of spend sitting on targets that have never produced an order. It is the most common failure I find and the most direct to fix, because negating a zero-order term costs no sales by definition. One $45K/mo skincare account on my results page was at 41% wasted spend at audit; cutting it to 16% added $2,890 a month in profit with no budget increase.
Run the checklist, then run the machine
My suggestion for the next hour: pull the 60-day bulk file, run groups one and two by hand so you see your own leaks with your own eyes, then drop the same file into the free Account Health Snapshot and let it grade all nine metrics in about 60 seconds. If the scorecard comes back mostly Healthy, you did not need me, and the tools are yours to keep using. If it comes back with Critical flags and real dollars behind them, bring the scorecard to the free 30-minute diagnosis call and I will walk the failing metrics back to their structural causes with your file on the screen. Either outcome beats guessing.