An Amazon PPC audit is the annual checkup you've been putting off. You know the type. You scheduled it. Then rescheduled it. Then decided the account was "basically fine" and went back to refreshing your ACoS dashboard like that was going to tell you something new. It wasn't. It never does.
[Deep breath. This is a safe space.]
Here's what we actually find when we open most accounts for the first time: auto campaigns eating 50–70% of total ad spend, negative keyword lists that could charitably be described as "decorative," and bids set by gut feeling in Q4 of some year that rhymes with "twenty-twenty-something" that nobody has touched since. The account isn't broken. It just hasn't had its checkup. The good news: there's a 9-point framework for that. The slightly less good news: you're about to find out how long you've been walking around with a cavity.
We run these nine checks on every account before we change a single bid. Not because we enjoy making checklists — well, maybe a little — but because optimizing a leaky structure is like mopping your floor with the faucet running. The mop is fine. The faucet is the problem.
- 1–4Structural Checks: The Foundation
- 01Campaign Hierarchy
- 02Match Type Distribution
- 03Search Term Report: The Waste Check
- 04Negative Keyword Coverage
- 5–6Economic Checks: Getting the Math Right
- 05Bid Strategy vs. Unit Economics
- 06Budget Allocation: Branded vs. Non-Branded
- 7–9Performance Checks: Trends and Ceilings
- 07TACoS Trend
- 08Placement Modifier Performance
- 09Listing Quality as a PPC Ceiling
- —What to Do With the Findings
- —FAQ
The 9 Audit Points
Run these in order. The first four are structural — they shape everything downstream. Get the structure right before you touch bids or budget. Skipping ahead to bid optimization on a broken campaign hierarchy is like rearranging deck chairs. You know the rest.
Structural Checks: The Foundation (Points 1–4)
Structure is everything. A broken campaign hierarchy can't be rescued by better bids — it just finds new and creative ways to waste your money. These four checks come first, every time, no exceptions.
Is your campaign structure doing any actual work?
The most common first finding in any Amazon PPC audit: everything living inside one or two auto campaigns, no separation between discovery and conversion, and zero ability to control spend at the match-type level. Amazon is making all the decisions. [Squints at screen.] That's not a campaign structure — that's a very polite subscription to Amazon's quarterly earnings.
A functional hierarchy separates four campaign types: auto (discovery only), broad (intent expansion), phrase (volume), and exact (conversion). Each does one job. Auto feeds broad. Broad feeds phrase. Phrase feeds exact. Terms that convert get promoted upward; terms that waste spend get negated downward. This is the machine that compounds over time. It sounds simple because it is — which makes it all the more remarkable how rarely anyone actually builds it.
When you audit campaign structure, look for three red flags: campaigns that mix branded and generic terms in the same ad group, ad groups with 20+ keywords jammed together like sardines in a can, and ASINs sharing campaigns despite very different margin profiles. Any of these means the structure is working against you.
Separate campaigns by match type and intent. Branded terms in their own campaign with a defensive bid. Auto capped at 15–20% of total budget. Top-performing exact match terms in dedicated campaigns with bid control. This is the structure that most PPC mistakes trace back to not having.
What percentage of your spend is in auto campaigns?
Pull your spend by match type for the last 90 days. If auto campaigns represent more than 25% of total spend, you have a structural problem. If they're above 50%, you are essentially leaving your wallet on the counter with a note that says "help yourself" — and Amazon, to their credit, absolutely will. Amazon's incentive is to spend your budget. Your incentive is to optimize your TACoS. These are not the same incentive. One of you wins this arrangement.
Auto campaigns are valuable for exactly one thing: discovery. They surface search terms you didn't know to target. That's worth 15–20% of your budget. Beyond that, you're handing control of your CPCs, your keyword selection, and your match logic to an algorithm that doesn't know what your contribution margin is, doesn't care what your contribution margin is, and — if it could talk — would probably say something deeply unhelpful like "have you tried increasing your bid?"
Auto: 15–20% (discovery only, capped daily budget). Broad and phrase: 30–40% (scale and harvest). Exact: 40–50% (high-intent, conversion-focused). Branded: 10–15% (defensive, low CPC). These aren't hard rules — they shift by category and stage — but they're a useful diagnostic benchmark.
Where is spend going with zero return?
The search term report is the most important document in your account. Most brands open it with the specific energy of someone checking their credit card statement — you click through, you see something alarming, you close it, and you decide to deal with it next month. That's the wrong cadence and the wrong impulse.
The right question isn't "what's working?" The right question is: what is Amazon showing my ads for that has never once generated a sale? Open your search term report, filter for spend above $15–20 with zero orders, and sort by spend descending. Every line on that list is money that left your account and came back as nothing. Not a sale. Not a lead. Not even a lesson. Just gone.
On a typical account that hasn't been audited in 6 months, this exercise surfaces 15–40% of total ad spend on terms that have never converted. In one account we opened last year, the first pass found $4,200 in spend over 90 days on zero-order terms — a competitor's brand name, some irrelevant category terms, and one very determined search for "fidget spinner" that kept triggering a supplement ad. [We did not keep that term. The fidget spinner community was going to be disappointed regardless.]
Every high-spend, zero-order term gets negated immediately. Every high-spend, high-order term gets extracted from auto and added as an exact match in a manual campaign. Run this weekly. The accounts that compound the fastest have weekly search term review baked into the operating system, not the monthly audit.
Is your negative keyword list doing real work?
Negative keywords are the dental floss of Amazon PPC. Everyone knows you should use them. Almost nobody does it consistently. And when someone asks how often you're adding them, you say "regularly" while making very confident eye contact. You're not adding them regularly. I can see your campaign from here.
Pull your negative keyword lists for every campaign. If they're empty — or fewer than 30–40 terms in a mature account — your campaigns have no filter. Here's the compounding damage: every irrelevant impression your ads serve is inventory that doesn't convert, which tanks your CVR, which raises Amazon's effective cost to serve you, which raises your CPC. Negative keywords aren't optional. They're the immune system of the account. A campaign without negatives is a body without an immune system, and Amazon's search algorithm is a very enthusiastic virus.
One more thing: check whether you're negating at the campaign level or the ad group level. Campaign-level negatives apply everywhere. Ad group negatives apply only to that group. Most accounts use one or the other. The right answer is both, layered deliberately — like actually flossing, not just flossing the teeth you think your dentist is going to check.
A mature account has campaign-level negatives blocking obvious irrelevant terms, cross-campaign negatives preventing cannibalization between broad and exact campaigns, and active weekly additions from the search term report. Negative keywords are never "done" — they're a live system.
The four structural checks above determine roughly 80% of account performance. Bid optimization is a refinement. Placement modifiers are a refinement. Smarter budget splits are a refinement. None of them can rescue a broken structure — they just help you lose money slightly more efficiently.
Economic Checks: Getting the Math Right (Points 5–6)
Once the structure is clean, the next question is whether your bids are grounded in actual math — or in vibes and Amazon's suggestions. [Spoiler: it's usually Amazon's suggestions.] These two checks close the economic gap.
Are your bids grounded in math, or in Amazon's suggestions?
Most accounts set bids one of two ways: Amazon's suggested bid, or a number that "felt right" at some point in 2024 that nobody has revisited since. Neither is a strategy. Amazon's suggested bid reflects what Amazon thinks will maximize impressions — not what will protect your margin. Their own Sponsored Products guide acknowledges bid strategy as the primary efficiency lever and then, helpfully, declines to tell you how to set it correctly. Thanks, team.
The correct approach is math: target ACOS × conversion rate × product price = max CPC. Walk through it. If your product sells for $35, your target ACOS is 12%, and your CVR on that keyword is 8%, your max CPC is $0.34. If you're bidding $1.20, you are not running a campaign — you are making a very generous and entirely voluntary donation to Amazon's quarterly earnings. The path to lower ACOS runs through bid math, not bid instinct.
During the audit: pull your top 20 keywords by spend, calculate the math-derived max CPC for each, compare to actual bids. The gap is your problem set. Keywords where actual CPC is materially above the math need reductions, not excuses. Keywords where CPC is below the math have room to scale. That second group is usually smaller than people expect.
Set a math-derived max CPC (the ceiling) for each keyword based on target ACOS and CVR. Set a minimum bid floor to maintain impression share. Adjust within that range based on performance data, not instinct. Review monthly — CVR and price change, so the math changes.
How much are you spending to defend what you already own?
Branded campaigns — ads on your own brand name and ASIN-specific terms — should be your cheapest, most efficient spend. CPCs are low, conversion rates are high, and you're protecting real estate that competitors will happily occupy the moment you stop showing up. But spending 30–40% of your total budget on branded terms is like spending 40% of your Saturday defending friendships that were never under threat. Your existing customers know who you are. Use that budget to go find new ones.
During the audit, pull branded vs. non-branded spend as a percentage of total budget. If branded is above 20%, one of two things is happening: either your campaign structure is bleeding non-branded spend into branded terms (a structural leak you can fix with a negative keyword), or you've genuinely over-indexed on defending known demand instead of creating new demand. Both are problems. One is a five-minute fix. The other requires a strategic conversation.
Non-branded spend — generic category terms, competitor terms, top-of-funnel discovery — is where growth comes from. It costs more per conversion, but it builds organic rank and grows the top of the funnel over time. A rough benchmark for a scaling brand: 15–20% branded, 80–85% non-branded. If you're flipped, the account is protecting the past instead of building the future.
Run a 90-day report by campaign, tag campaigns as branded or non-branded, and calculate the split. If branded is inflated, look for cross-campaign targeting overlap — broad and auto campaigns often accidentally serve branded terms. Add branded terms as negatives in non-branded campaigns to break the bleed.
Performance Checks: Trends and Ceilings (Points 7–9)
The last three checks tell you whether the account is building momentum or just running in place. One of them has nothing to do with your campaigns at all — which is exactly why most PPC audits miss it entirely.
Is the ad account building the business, or just sustaining it?
TACoS — and yes, we mean Total Advertising Cost of Sale, not the $4 street food that is objectively superior to anything from a "modern Mexican concept," though both can leave you regretting your decisions — is the number that tells you whether your ad account is building a business or just renting revenue it already has.
ACoS tells you how efficiently your ads are converting. TACoS tells you whether the business is actually growing or just cycling money you already own. TACoS is total ad spend divided by total revenue — not just ad revenue. If TACoS is rising while ACoS is flat, your ads aren't creating organic demand. You're running faster to stay in the same place. If TACoS is falling while ACoS holds steady, ads are building organic rank and the business is compounding. That second scenario is what a healthy account looks like over time. It's achievable. Most accounts have never experienced it.
During the audit, pull a 12-month TACoS trend and look for the inflection points. Where did it start rising? What changed at that moment? New product launches, Q4 spend surges, and competitor entry all create distinct TACoS signatures you can diagnose and explain. We covered the full TACoS vs. ACoS distinction here — worth reading alongside this audit if you've been optimizing against the wrong metric this whole time.
For a healthy, scaling brand, TACoS should trend down over time as organic rank builds and organic revenue grows faster than ad spend. A TACoS above 15% on a mature product usually signals structural problems. We've taken accounts from 18%+ TACoS down to 3.54% — that delta is almost entirely structural, not seasonal.
Are you paying the same CPC for placements that convert at very different rates?
Amazon's placement report breaks performance into three buckets: top of search (first row), rest of search, and product pages. These three placements convert at very different rates and cost very different CPCs. Most accounts treat them identically because nobody has pulled the data to check. This is essentially a $20 bill sitting on the sidewalk — and most brands walk past it every single day because nobody thought to look down.
Pull the placement performance report for your top campaigns and compare CVR and ACoS across the three types. In most accounts, top-of-search converts at 2–3x the rate of product pages — but the default bid modifier is 0% for all three. That means you're underbidding where you perform best and overpaying for the placements that convert worst. [Takes a long, meaningful pause to let that land.]
Placement modifiers let you adjust your base bid up or down by placement type. If top-of-search converts at 12% and product pages convert at 4%, a +50% modifier on top-of-search and a −30% modifier on product pages gets you much closer to efficient allocation. This single audit finding typically reduces ACoS by 2–4 percentage points on an account that hasn't set modifiers. It's not complicated. It's just been sitting there, waiting for someone to check.
Run 30 days of placement data minimum before setting modifiers — you need statistically meaningful CVR by placement. Start with conservative adjustments: +20–30% on top-of-search, neutral on rest of search, −10–20% on product pages. Revisit monthly as data accumulates. Never set and forget.
Is the listing you're sending traffic to converting it?
PPC efficiency has a hard ceiling set by listing quality. You can build a perfect campaign hierarchy, set bids with mathematical precision, and negate every irrelevant search term in the known universe — and still have a mediocre ACoS if your listing doesn't convert the traffic you're sending it. This is the part of the checklist most PPC audits skip, because technically it's "not a PPC problem." It is your money, though. That makes it your problem.
Check four things on every ASIN you're advertising: title (does the first 70 characters include the primary keyword?), main image (white background, fills 85% of the frame, product clearly visible — not floating mysteriously in soft focus like a perfume ad), reviews (above 4.0, minimum 15–20 before you scale spend on it), and A+ Content (brand story, comparison module, something that signals this is a real brand and not a print-on-demand experiment). A listing missing any of these is converting at a fraction of its potential, and you're funding a bad CVR with ad spend. The CVR is not going to fix itself.
Practical audit: pull your top 10 most-spent ASINs and run each through a 30-second listing check. If any fail on the basics, pause spend on that ASIN until the listing is fixed. Sending traffic to a broken listing is not marketing. It is, generously, charity. And unlike most charitable contributions, it is not tax deductible.
Amazon average CVR is roughly 10–15% for Prime members. A well-optimized listing in a competitive category should convert at 12–18%. If you're seeing 5–7% CVR on a mature product with solid reviews, the listing is the problem — not the bid. Fix the listing before increasing the budget.
What to Do With the Findings
You've now done something most Amazon brands haven't done in 6+ months: you've had the checkup. You know where it hurts. Now the question is what to fix first — because trying to fix everything at once is how you end up unable to diagnose what actually moved the needle. Here's the sequence:
Making too many changes at once. When you restructure campaigns, adjust bids, and build negatives all in the same week, you lose the ability to know what actually worked. You've turned a carefully designed audit into a science experiment with no control group. Sequence the changes. Give each structural fix 7–10 days of data before layering the next one. The one exception: obvious waste (high-spend, zero-order terms) gets negated immediately. That one doesn't need a waiting period. Bleeding doesn't wait.
A note on what you'll find: most accounts we audit haven't had a systematic review in 3–6 months. First-audit findings are often large — a full restructure, a significant negative keyword build-out, a meaningful shift in budget allocation. That's completely normal. It doesn't mean the account is broken. It means it's been running on autopilot. Autopilot works great for planes. For Amazon ad spend, it's just a subscription to compounding inefficiency — one you can cancel, starting with any one of the nine checks above.
[And yes — the annual checkup callback. You've been waiting for it. You're welcome.]
If you want to see where your specific account's waste is concentrated before running the full framework, upload your Search Term Report from Amazon Advertising to our free analyzer below — it surfaces the top waste candidates in 60 seconds, no signup required.
FAQ
A thorough Amazon PPC audit takes 2–4 hours for a single product line, and 4–8 hours for a full catalog. The campaign structure check and search term waste analysis take the most time. If you're doing it yourself, block half a day. If you're hiring someone, expect a 2–5 business day turnaround for a written report.
The right cadence is a light review weekly (search terms, pacing, exceptions), a fuller audit monthly, and a deep structural audit quarterly. Most brands under-audit — they set bids in January and look again in Q4. The accounts that compound the fastest are the ones running systematic weekly search term reviews and catching waste early.
In our experience, the most common finding is an over-reliance on auto campaigns — often 50–70% of total spend — with minimal negative keyword coverage. Auto campaigns are useful for discovery, but should represent 15–20% of budget at most. The rest should be in manual campaigns where bids are set based on actual conversion data, not Amazon's algorithm.
If you have internal capacity to act on findings, a DIY audit using this framework is a solid starting point. If your campaigns haven't been touched in 6+ months, or if you're spending more than $5,000/month on ads without a clear campaign hierarchy, hiring an experienced operator to audit and rebuild is typically worth the cost — the wasted spend usually covers the audit fee within 30 days.
A search term report review is one component of a full PPC audit — it looks at where impressions and clicks are going. A full PPC audit covers campaign structure, match type distribution, bid strategy, budget allocation, TACoS trend, placement modifiers, and listing quality. The search term report is the single most important document in an audit, but it's only one of nine areas to check.
Freelance Amazon PPC audits typically range from $500 to $2,500 depending on catalog size. Agency audits range from $1,500 to $5,000+. If you're spending $5,000+ per month on Amazon ads, a professional audit that cuts 20–30% of wasted spend pays for itself in the first month. For accounts under $2,000/month in ad spend, start with the free tools — upload your search term report to a free analyzer first.