Search "PPC audit checklist" and you'll find a dozen versions of the same list. Set your date range. Check your match types. Review your ad extensions. Look at location performance.
All reasonable. None of it will tell you whether the account is making money, which is the only thing a PPC audit is for.
Those checklists were written for lead generation, where a conversion is a form fill and success is more of them at a lower cost. Ecommerce is a different problem. A Shopify brand can pass every item on a generic PPC audit checklist and still be buying revenue at a loss, because the checklist never asks the two questions that decide it: are the conversion numbers real, and what does this account have to beat?
This is the PPC audit we actually run on inherited accounts, in the order we run it. Eight checks. The first three you can do yourself this afternoon.
What is a PPC audit?
A PPC audit is a diagnostic review of a paid advertising account: what is running, whether the measurement can be trusted, and whether performance clears the profitability threshold the business needs. It identifies problems and their order of priority. It is not the fix, and treating it as one is how accounts get optimised in the wrong sequence.
That distinction matters more than it sounds. A PPC audit that produces forty recommendations with no priority order is a list, not a diagnosis. The value is in knowing which three things to do first and which thirty-seven don't matter yet.
Why most free PPC audit reports are worthless
Free automated PPC audits are a lead generation artefact. You enter your account, a script runs, and you receive a report full of red warning triangles: low quality scores, missing sitelinks, single-keyword ad groups, unused extensions.
Every one of those flags is real. Almost none of them costs you money.
The deeper problem is structural. An automated tool reads the same numbers your dashboard reads. If your conversion tracking is double-counting purchases, the tool inherits that error and grades you against it, then congratulates you on a strong ROAS. It cannot detect that the input is wrong, because the input is all it has.
It also cannot know your margin. That single omission makes every performance judgement it offers meaningless, because a 4x ROAS is excellent at one margin structure and a slow loss at another.
The checks that matter most are the ones no script can run.
Check 1: Is the conversion data real?
Nothing else in the PPC audit means anything until this is settled, which is why it goes first and why almost no checklist includes it.
The reconciliation test, in ten minutes
Pick a clean 30-day window that has already closed. Then:
- In Google Ads, pull total conversions for that window, segmented by conversion action.
- In Shopify, pull total orders for the identical window.
- Compare them.
They won't match exactly, and shouldn't. Google attributes on click date rather than order date, and some orders arrive from channels Google never touched. But you're not looking for a match. You're looking for the size and direction of the gap.
If Google reports substantially more conversions than Shopify has orders, something is being counted that isn't a sale. If it reports dramatically fewer, tracking is firing inconsistently or not at all. Either way, stop the audit and fix that before reading another number.
Duplicate and soft conversions
Two failures cause most of the gap.
Soft conversions counted as sales. Add-to-cart, begin-checkout, and page-view actions marked as primary conversions. The account then optimises toward people who browse rather than people who buy.
Duplicate purchase trackers. The Google & YouTube channel firing alongside a manually installed gtag, or a theme snippet left in place after an app migration. Every order counts twice, and reported ROAS is double reality.
An Australian jewellery brand came to us with both at once: page views, carts, and checkouts all registering as conversions, plus duplicate purchase tracking on top. A competent agency had been optimising diligently against those numbers for months. Cut to a single primary purchase action with data-driven attribution, restructured, and rebuilt, it produced A$331,092 at 3.0x blended ROAS against a 1.3x break-even.
Check 2: Does the feed support the campaigns?
This is the step generic PPC audit checklists skip entirely. For any Shopify brand running Shopping or Performance Max, the Merchant Center feed is not a settings page. It is the targeting layer. There are no keywords in Shopping; Google reads your product data to decide which queries each item can appear for.
Check three things. Disapprovals and warnings in Merchant Center, product by product rather than as a headline count. Titles, specifically whether variants share a templated title, which makes forty products compete for one query while the long tail goes unserved. We covered that mechanic in detail in our guide to feed optimisation for product variants. And identifiers — missing or invalid GTINs suppress eligibility silently, with no error loud enough to notice.
You can score your feed against 40 checks free in about 15 seconds, which covers this step faster than doing it by hand.
Check 3: Is the account structured, or just running?
Brand versus non-brand
If brand and non-brand traffic sit in the same campaigns, the account's headline ROAS is being flattered by demand you already owned. People searching your name were going to buy anyway.
Segment them and look at non-brand alone. That number tells you whether the account is creating anything. We have seen accounts reporting 6x that were almost entirely brand harvesting, generating no new customers at all.
Performance Max alongside Standard Shopping
Check whether PMax has real structure or one asset group holding the whole catalogue, and whether brand exclusions are set. Note too that where both target the same products, Performance Max wins the auction over Standard Shopping, so running them in parallel isn't the test people assume it is.
Check 4: Are targets set against a real break-even?
Ask what target ROAS the account is managed to, then ask where that number came from. If the answer is an industry benchmark, a previous agency, or a round number that felt right, the account has no profitability anchor.
Break-even ROAS is roughly one divided by contribution margin, and contribution margin means after cost of goods, fulfilment, transaction fees, and returns. Most brands calculate it on gross margin alone and land 30% too optimistic. You can work yours out in about two minutes.
An audit that never asks your margin cannot tell you whether the account is winning. It can only tell you whether it changed.
The eight-point PPC audit checklist
The full PPC audit sequence, in priority order. Checks 1 to 3 are DIY; the rest usually need account history to judge.
- Conversion data integrity. Reconcile Google Ads conversions against Shopify orders for the same 30-day window. Investigate any large gap before proceeding.
- Conversion action hygiene. One primary purchase action. No soft conversions marked primary, no duplicate trackers firing.
- Feed health. Disapprovals, variant-level titles, identifiers, and image compliance in Merchant Center.
- Brand and non-brand separation. Segment them and judge non-brand on its own numbers.
- Campaign structure. PMax asset groups split by product line and region, brand exclusions applied, Standard Shopping used deliberately rather than by accident.
- Search terms and negatives. Where is spend actually going? Look for spend on terms that will never convert, and for a negative list that hasn't been touched in a year.
- Bidding strategy against data volume. Target ROAS bidding needs enough conversions to learn from. On a thin account it will chase noise, and a simpler strategy will beat it.
- Account ownership. Are the Google Ads and Merchant Center accounts owned by you, or by your agency's MCC? If it's the latter, your conversion history and bidding learning leave when the agency does.
That last one isn't a performance check, but it's the one most likely to cost you a year of progress at exactly the wrong moment.
What to do with what the audit finds
Fix order matters more than fix count
Tracking, then feed, then structure, then targets. Always that order, whoever runs the PPC audit.
Restructuring campaigns on top of broken conversion data means rebuilding against numbers that aren't real, and you'll do the work twice. Fixing the feed before tracking means you won't be able to tell whether the feed work helped. Each step makes the next one measurable.
Don't scale spend during the fix window
Every fix disturbs the bidding algorithm's learning. Doing that while also increasing budget produces a fortnight of data nobody can interpret, and the usual conclusion is that the fix caused the dip.
Hold spend flat, make the changes, let it settle, then scale. On the accounts we take over this window is typically two to four weeks, and it is the least glamorous and most valuable part of the engagement.
The takeaways
A PPC audit is only worth the time if it asks questions a script can't answer.
- Verify the conversion data before anything else. Reconcile against Shopify orders. Every downstream number depends on it.
- For Shopify brands the feed is the targeting layer, not a settings page, so feed health caps everything above it.
- Judge non-brand on its own. Blended figures hide whether the account creates demand or just harvests it.
- An audit that doesn't ask your margin can't tell you if you're winning. Break-even ROAS is the reference point.
- Order beats volume. Tracking, feed, structure, targets, with spend held flat until it settles.
Run the first three checks yourself. They cost an afternoon and they'll tell you whether your account has a strategy problem or an inputs problem, which are fixed very differently.
If you'd rather we ran it, that's what the free Google Ads audit is for. You keep the findings whether or not you work with us, and you're welcome to hand them to whoever manages the account today.