Most Amazon PPC audits chase an arbitrary ACOS target. This one inverts that: every check is judged against your break-even ACOS, and it hunts the structurally invisible waste the console hides by default.
To audit an Amazon Ads account, anchor every judgment to your break-even ACOS (the ACOS at which an ad sale earns zero profit after COGS, FBA and referral fees, and returns), then work top down through five areas: account structure and hygiene, targeting and search-term waste, bidding and placement and budget mechanics, profitability and attribution reality, and measurement. The biggest leaks are rarely bad keywords. They hide in mechanics most sellers never inspect: out-of-budget hours when buyers are searching, placement multipliers up to 900 percent, loose-match and complements bleed in auto campaigns, and a 7-day attribution window that flatters performance.
An Amazon Ads audit is not about finding one broken setting. It is about pressure-testing every layer of the account against a single question: is this spend earning a profit at the margin you actually keep after fees, COGS, and returns. With Amazon's retail media ad revenues forecast to pass $60 billion in 2025 and reach nearly $70 billion in 2026, ad efficiency is now a board-level line item, and a profitable-looking account can still hide meaningful wasted spend in blind spots the console does not surface by default.
This checklist is ordered the way an experienced operator runs a real audit: cheapest, highest-impact fixes first, structural work last. Tick each item as you complete it. Your progress is saved in this browser, so you can close the tab and pick up where you left off. The same data points that drive this checklist are the ones our team reviews in a free audit. For the cross-platform version of this method, see the 20-point Google Ads audit checklist.
If structure is wrong, every fix below becomes harder. Amazon recommends including only ASINs from the same category in one campaign, and you can add up to 1,000 products to an ad group at once, so structure is also a relevance lever, not just bookkeeping. These four checks confirm you can see and control where money goes. Deciding whether to scale the account at all is a separate question, covered in when to invest in Amazon Ads beyond Google Shopping.
Each campaign contains ASINs from a single category, with a naming convention that makes reports legible at a glance.
Amazon advises same-category grouping; mixed categories blur relevance and make the search-term report impossible to read.
Adopt a fixed pattern like Brand_AdType_MatchType_Theme and split mixed campaigns by category before anything else.
Searches for your own brand, generic terms, and competitor terms each live in their own campaigns.
Branded terms convert cheaply and pull whole-account ACOS down, hiding weak non-branded and competitor performance.
Create a dedicated branded campaign and report non-branded ACOS separately from then on.
The same ASIN and the same exact keyword are not active in more than one campaign at once.
Internal overlap drives up your own CPC and splits performance data across duplicate entries.
Keep one canonical home per ASIN and keyword, then negate it everywhere else it appears.
New Sponsored Products campaigns start near the recommended $10 daily budget and are monitored at least twice a week in week one.
Starved or runaway budgets in the first week generate misleading data you will later audit against.
Set conservative launch budgets, check performance twice weekly early, and scale only on proven data.
Structure is the lever almost nobody touches because it feels like plumbing. But you cannot steer spend you cannot see, and a clean, single-category structure is what turns a search term report from noise into a map. Cortex, Capconvert paid media desk
This is where most low-intent waste hides. Automatic targeting runs four distinct groups, close match, loose match, substitutes, and complements, each bid independently. Loose match and complements are the most common sources of low-intent spend, and you can bid them down or turn them off separately rather than abandoning auto altogether.
The loose match and complements groups are bid down or off where they spend without converting.
These two groups are the most common low-intent waste in auto targeting and can be controlled independently.
Split the four auto groups and reduce or disable loose match and complements while keeping close match running.
Proven terms from the search-term report have graduated from the auto R&D lab into tightly managed exact-match campaigns.
Leaving winners in auto means paying broad CPCs for terms you could bid on precisely with full control.
Move any term with repeated profitable orders to exact, then negate it in the source auto campaign.
Negative phrase, negative exact, and negative product or brand exclusions are in place where intent is wrong.
Negative phrase blocks queries containing the phrase, negative exact blocks the exact phrase, and Exclude stops ads on specific ASINs or brands you do not want to pay for.
Build negative-exact entries for zero-order terms and Exclude lists for irrelevant competitor ASINs and brands.
Auto and broad operate as discovery; phrase and exact operate as performance, with their own budgets.
Mixed match types share a budget, so high-intent exact terms get starved by exploratory broad spend.
Separate discovery and performance into distinct campaigns and fund the performance side to its proven ceiling.
Any search term that spent more than one target CPC with no orders is added as a negative.
These terms are pure waste and are usually the single largest bucket of recoverable budget.
Pull a 60-day search-term report, sort by spend with zero orders, and add negative-exact entries.
These are the mechanics generic checklists never inspect. A profitable-looking account can leak here through bid strategy, placement multipliers, dayparting, and rule-based bidding that silently switches off.
Dynamic bids down-only is the default on unproven campaigns; up and down is reserved for proven winners.
Per Amazon's dynamic-bidding guidance, up and down can raise a bid by as much as 100 percent for top of search (and a lower uplift on other placements) when a conversion looks likely, so a $1.00 bid can cost up to $2.00 at top of search.
Use down-only until a campaign has a stable, profitable conversion rate, then test up and down deliberately.
Top of search, rest of search, and product-page modifiers reflect actual placement-level conversion data.
Placement bid adjustments can be set as high as 900 percent, a multiplier most audits never reconcile against where conversions actually happen.
Pull placement reports and set modifiers only on placements with proven ROAS, removing blind high multipliers.
Any rule-based bidding is still active and not paused after a sustained stretch below its target.
Per Amazon's rule-based bidding documentation, a rule can auto-disable after running below target for a sustained period, quietly reverting bid control.
Audit rule status, re-enable or replace disabled rules, and confirm eligibility before relying on automation.
Campaigns converting below break-even ACOS are not exhausting their daily budget before the day ends.
A profitable, budget-capped campaign leaves guaranteed sales on the table every single day.
Check the budget-limited report and raise caps on any campaign profitable at scale.
Budget is not exhausted in low-conversion hours, leaving the account dark when buyers are active.
Since November 30, 2023, hours-of-day for schedule-based budget rules lets you raise budgets at specific times and pair them with hourly reporting, the lever behind dayparting fixes.
Pull the hour-of-day report and use schedule-based budget rules to shift spend into proven high-intent windows.
Every bid traces back to a documented break-even ACOS derived from true unit margin, not Amazon's suggested bid.
Suggested bids chase competition, not your margin, so following them drifts spend above your profit line.
Compute break-even ACOS as profit margin before ad spend, subtracting COGS, FBA and referral fees, and returns, then set ceilings from there.
The most dangerous waste is the kind that looks like success. Attribution truncation and same-SKU thinking can make you kill campaigns that actually drive organic and first-time buyers.
You know whether reported sales use the shorter Seller-side or longer Vendor-side window and adjust expectations.
Sponsored Products attribution is widely reported as 7 days for Seller Central and 14 days for Vendor Central; a 7-day truncation understates the halo and a longer window flatters cash-basis ACOS.
Report on a consistent window, confirm the current default on Amazon's attribution help page, and reconcile ad-reported sales against settlement revenue.
Total advertising cost of sales is tracked monthly as ad spend divided by total revenue, trended against ad ACOS.
Ad-only ACOS misses the organic halo, so a campaign that looks expensive can be lifting total revenue and lowering TACOS.
Build a TACOS trend line and judge spend changes by whether total ACOS improves, not just ad ACOS.
Campaigns acquiring first-time buyers are judged on repeat value, not single-order ACOS.
New-to-brand metrics, available for Sponsored Brands, use a 12-month look-back: a shopper counts as new if they have not bought your brand on Amazon in the prior 12 months.
Use the new-to-brand metric and a known repeat rate to set a higher acceptable ACOS for acquisition campaigns.
The final checks make the audit durable: a clean view of impression-share and SQP gaps, and a remediation sequence that stops the bleeding before it scales. For packaging findings, see how to present ROAS, CPA, and attribution clearly.
You know your ad-versus-organic share on priority queries and where impression share is being left to competitors.
Without the Search Query Performance view, you cannot see which high-value queries you under-serve in ads while ranking organically, or the reverse.
Use the SQP data layer to prioritize spend on high-share-opportunity queries; the deep-dive is in using Amazon Search Query Performance.
Findings are sequenced stop-the-bleeding first (negatives, budget caps, placement fixes), then scale (harvesting, acquisition).
Without a system, audits become one-off heroics instead of an ongoing discipline that compounds quarter over quarter.
Adopt a fixed naming convention and weekly metrics cadence, and tie ad efficiency to organic and AI visibility via Rufus-driven ASIN discovery.
A good ACOS is any figure below your break-even ACOS, which equals your profit margin before ad spend. Calculate it by subtracting COGS, Amazon FBA and referral fees, and returns from your selling price to find the margin you keep; that margin percentage is your break-even ACOS. A product with a 40 percent margin runs profitably at a 35 percent ACOS, while a 15 percent margin product loses money at the same number. Always calculate break-even first, then judge every campaign against it.
Start with the search-term report and negate any term that spent more than one target CPC with zero orders. Then inspect the mechanics generic checklists skip: loose-match and complements bleed in auto campaigns, placement multipliers set as high as 900 percent without supporting data, dynamic up-and-down bidding inflating top-of-search CPCs by up to 100 percent, and profitable campaigns running out of budget during peak buying hours. These structural blind spots usually hold more recoverable budget than bad keywords do.
ACOS is ad spend divided by ad-attributed sales; TACOS is ad spend divided by total revenue, including organic sales the ads help drive. Optimize for TACOS as the true health metric, because ad-only ACOS ignores the organic halo and can make a healthy launch look unprofitable. Use ACOS at the keyword level to control individual bids, and watch the TACOS trend line at the account level to confirm spend is lifting the whole business.
Sponsored Products attribution is widely reported as 7 days for Seller Central accounts and 14 days for Vendor Central, with Sponsored Brands and Sponsored Display commonly cited at 14 days. Amazon has changed attribution defaults before, so confirm the current window on Amazon's attribution help page. The practical point for an audit is that a 7-day window understates later conversions and the organic halo, so reconcile ad-reported sales against actual settlement revenue.
Default to down-only on unproven campaigns and reserve up-and-down for campaigns with stable, profitable conversion rates. Per Amazon's dynamic-bidding guidance, up and down can raise your bid by as much as 100 percent for top-of-search placements (and a lower uplift on other placements) when a conversion looks likely, so a $1.00 bid can cost up to $2.00 at top of search. That uplift is only worth paying on terms with proven top-of-search ROAS.
Run a full 20-point audit once a quarter and a lighter search-term, bid, and budget review every week. Amazon's auction shifts constantly, so the negative-keyword, placement, and budget checks need weekly attention, while structural and measurement checks can hold for a quarter. With Amazon retail media forecast to approach $70 billion in 2026, the cost of leaving these blind spots unchecked compounds, so a fixed cadence pays for itself.