The Ecommerce PPC Audit Checklist: Five Checks in the Right Order

How Do You Audit an Ecommerce PPC Account?

Open two tabs: Merchant Center and Google Ads. Keep a notepad open. Every step ends with something you write down. Numbers you can check beat opinions you have to trust.

Why the Order Matters More Than the Checks

Bids sit on structure. Structure sits on tracking. Tracking sits on the feed.

Change the top layer first and you are tuning a machine that is reading bad data. You raise a target ROAS on a campaign where the conversion number is wrong. You split a campaign in half when half the products in it never sell. You feel busy. Nothing moves.

Start at the bottom instead. Fix what Google reads. Then fix what Google counts. Then look at what your products did with the money. Then fix how the money is grouped. Bids get easy once the layers under them are honest.

This is the same order we publish on our agency page and on our process page. It is not five random tabs opened in whatever order they load.

Here is the whole pass on one screen.

Some steps give you one number. Step 1 gives you three things, and two of them are words, not numbers. That is fine. The point is that every step ends with something written down instead of a hunch.

Step 1: Can Google Actually Read Your Product Feed?

Start in Merchant Center, not Google Ads. Most of the work sits here.

Count the disapprovals. Open Products, then the issues view. Write down the number of disapproved products and the top reason. Google is blunt about the cost here: disapproved products stop showing across Google, and an account-level suspension takes them all down at once. If you need the reason decoded, we built a free Merchant Center disapproval checklist.

Check for a product identifier. Filter for products with no GTIN and no MPN. A GTIN is the barcode number. An MPN is the manufacturer part number. Both help Google match your product to what a shopper typed.

Check your title lengths. Sort or scan them. Titles that open with your brand name and end at "Dress" are giving Google almost nothing to match.

We measured this. In a nightly snapshot of Merchant Center accounts on 2026-07-28, across 12,375 products in 251 stores, 81.3% carried neither a GTIN nor an MPN. Only 4.2% of titles came in under 50 characters.

Google reads your product data from the product data specification, and Shopping and Performance Max decide from it. That is why this step is first and bids are last.

Step 2: Does Google See Your Real Sales?

Open Goals, then Conversions, then Summary. That is the current route. If you are looking under Tools, you are on an old menu.

Read which actions are set to Primary. Primary actions are the ones your bidding chases. Google lets you have more than one, and a campaign can carry its own goals, so "exactly one" is not the rule. Two questions are. Is your purchase in there once, or is a second tag counting the same sale twice? And are browsing actions like add-to-cart or page views sitting in Primary too? Those teach the bidding to chase people who look instead of people who buy. Anything you want to watch but not bid on goes Secondary.

Compare Google's sales to Shopify's, on matched dates. Careful here, because the default column lies to you about dates. Google's standard conversion column dates a sale back to the day of the ad click. Shopify dates it the day of the order. Someone who clicked on the 29th and bought on the 3rd lands in a different month in each tool. So add the conversions by conversion time column in Google Ads first, then pull the same dates out of Shopify.

Then read the gap for what it is. Google reports the sales it can attribute to your ads. Shopify counts every sale from every channel. Your email list. Your organic traffic. Your repeat buyers. So the gap is never zero. And the gap on its own does not prove your tracking is broken. It turns into a finding two ways. One, it swings hard from month to month for no reason. Two, Google shows a sliver of the orders on days when ads were nearly all your traffic.

One more thing about that gap. Google does not only count clicks. It can credit a sale to an engaged view on a video ad. It also reports view-through conversions in their own column. Somebody saw the ad, never clicked, and bought later. Know which column you are reading before you call a number wrong.

Check that your tracking is reporting, not just installed. Most stores send the sale from the shopper's browser. Browsers now block a lot of that. Two fixes are common. Server-side tracking sends the sale from your shop's own server instead of the browser. An ad blocker living in the browser cannot eat it. Enhanced conversions take the email or address the buyer typed at checkout and send Google a scrambled version of it. Google uses that to match a sale it would otherwise miss. Whichever you run, open its diagnostics. Confirm it recorded sales in the last seven days. Installed is not reporting.

This outranks bids for one plain reason. Smart Bidding only optimizes toward what it can see. Feed it half your sales and it learns half your business.

Step 3: Which Products Take Money and Give Nothing Back?

The highest-value check on this page.

Open the Products tab in Google Ads. Date range: last 30 days. Sort by cost. Filter to conversions equal to zero. You are now looking at every product that took your money and returned nothing Google could track.

Add up the cost column. That is your number.

We ran this at scale for the Ecom Pareto study: 106,400 products across 137 stores. Of the 95,149 fashion products that received ad spend, 71.7% produced zero conversion value inside their 30-day window. Those zero sellers absorbed 14.7% of all fashion ad spend. In home decor it was 29.5%.

The two niches sat 14.8 points apart on the same measure. So this is not a fixed tax on running ads. That share shrinks when someone reads this list every month and acts on it. It grows when nobody looks.

Then run the same move on the search terms report. Open the search terms report for the same 30 days, sort by cost, filter to zero conversions. Write down your top five zero-sale terms.

Step 4: Does the Structure Match Your Margins?

Does your structure treat every product the same, when your products do not perform the same? Three checks answer that.

Are your winners and your dead stock in the same campaign? If yes, one budget and one target ROAS are being applied to products that behave nothing alike. Your best seller is being throttled by the same setting that props up a product nobody wants.

Do your products carry custom labels? Custom labels are tags you add in your feed, for things like margin band or sales velocity. Google still reads every product on its own either way, with its own ID, its own attributes and its own performance. What labels give you is a handle. With them you can split a campaign by margin, set a different target for your best sellers, and report on a group in one row. Without them you can see your products but you cannot steer them in groups.

Is there a plain Search campaign underneath Performance Max? One that catches your brand name and your highest-intent terms, where you can see the query and set the bid yourself.

Step 3 already told you your catalog is concentrated. In that same fashion sample, the top 1% of products carried 45.2% of all Google Ads revenue. An account that treats every product the same is fighting that curve with its own budget. Structure is how you stop.

Step 5: The Settings That Quietly Leak Budget

Four settings. None of them looks urgent, which is exactly why they get skipped. You are reading them, not changing them yet.

• Location targeting. Open it and read which option you are on. Presence or interest is Google's default and its recommended setting, and it reaches people outside your countries who look interested in them. Presence narrows it to people actually in the places you picked. Neither one is the right answer for everybody. It is the right answer when it matches where you ship. If you only ship domestically and you see spend from countries you cannot deliver to, that is your finding.
• Your CSS. If you advertise in Europe, check which comparison shopping service your Shopping campaigns run on and write the name down. It is a one-line answer and it is worth knowing.
• Automatically created assets and final URL expansion. Neither of these changes your campaign type. Performance Max already runs across Search, Shopping, YouTube, Display, Gmail, Discover and Maps by design. What they change is the raw material Google gets to use. Final URL expansion lets Google send traffic to other pages on your site instead of only the ones you gave it. Automatically created assets let Google write headlines and descriptions from your site content. On a feed-only setup, plenty of advertisers turn both off so the campaign sticks to the pages and the copy they chose. Write down whether they are on, and whether that was your decision or just the default.
• Budget access. Check whether one campaign is absorbing the budget the others cannot reach.

Keep this pass short, because the full walkthrough already exists. Read how to fix a Performance Max campaign that is wasting spend once, fix the settings properly, then stop rereading them every month.

What Not to Count as a Finding

Google's Recommendations tab hands you four things that look like discoveries. None of them are findings on their own.

1. Lost impression share to budget. This says: lift your budget cap. Worth a look, genuinely. A budget cap can hold back traffic that would have paid. A campaign can sit on its target and still leave money behind the cap. It is just not a finding by itself. It tells you Google wants to spend more. It does not tell you the extra spend earns. So go read the products behind that campaign. If they are already earning at your target and you have the stock, more budget is a fair call. If they are not, the number is Google's appetite, not your profit.

2. Limited by target ROAS. Same trick, different label. It says your target is too strict. Your target is a business decision. Google does not know your margin.

3. Improve Ad Strength to Excellent. Ad Strength scores how many assets you filled in and how varied they are. It counts inputs. It says nothing about sales.

4. Your optimization score is low. A percentage that goes up when you accept Google's recommendations. Accepting all of them gives you 100% and tells you nothing about where money is leaking.

Want a longer, deeper version of the audit itself? Dennis Moons at Store Growers publishes a 44-point Google Ads audit that goes further into Search campaign detail than this page does. He makes the same call on Recommendations and on optimization score. Different scope, same read.

What Do You Do with the List?

Fix in the audit order, not by how easy each item looks. Feed, tracking, products, structure, settings.

Some of what you found you can fix yourself. A batch of disapprovals. A second tag counting the same purchase twice. A list of products that spent and sold nothing. Clear those and get back to running your store.

If it turned up a broken feed plus tracking Google cannot read, that is a bigger job than a checklist. Rebuilding product data across a live catalog while sales keep running is real work, and doing it badly costs more than leaving it.

Under roughly EUR 5,000 a month in ad spend, an agency is bad math. A percentage fee eats budget that should be buying clicks. That is our own line, not an industry rule. Run it yourself. The free Google Ads eCom Lab has 1,200+ ecom operators in it running this same playbook, and it costs nothing. We run that room, so weigh that too.

Above that, here is our fee. Check it against anyone else's quote. 10% on the first EUR 10,000 of monthly Google Ads spend, stepping down bracket by bracket to 6% above EUR 150,000. No setup fee. No lock-in. Two tools get billed on top at a fixed published price. EUR 24.90 for the CSS partner and EUR 49.90 for the feed tool, per store per month, and only once they are installed. All of it sits on the pricing page next to what we actually do.

The five findings you just wrote down are yours either way. If you want a second read on them, send your store URL on WhatsApp and we will tell you which one to fix first.

Frequently Asked Questions

What does an ecommerce PPC audit checklist cover?

Five layers, in this order. One, the product feed in Merchant Center, because Google has to read your products before it can sell them. Two, conversion tracking, because bidding chases what it can see. Three, products that took spend and produced no tracked sales. Four, account structure, so your best sellers and your dead stock are not sharing one budget and one target. Five, campaign settings. Run all five before you fix anything. Otherwise you rebuild one campaign and never look at the other four layers. Every step ends with something you write down, so you finish with five findings instead of a feeling.

What order should you audit a Google Ads account in?

Bottom up. Product feed, then conversion tracking, then the products that spent and sold nothing, then account structure, then bids and settings. Each layer sits on the one below it. Structure cannot beat tracking that counts the wrong sales. Bidding cannot beat product data Google struggles to read. The product check sits in the middle because it reads the data you just checked and it tells you what the structure has to fix. Bids are the layer you can see, so they are the tempting place to start. They are the last one that matters. Fix the data underneath and the same bid buys you more.

What is the highest-value check in a PPC audit for an online store?

Products that took ad spend in the last 30 days and produced no tracked sales. We ran that check across 106,400 products in 137 stores for our Ecom Pareto study. Of the fashion products that took spend, 71.7% produced zero conversion value inside their 30-day window. Those products absorbed 14.7% of fashion ad spend. Home decor was worse: 80.5% of products and 29.5% of the spend. Read the number carefully though. It counts the conversion value Google Ads attributed inside the window. Some of those products sell elsewhere. Some convert later. So it is the shortlist to investigate, not a bill for waste.

Should I trust Google's Recommendations tab in an audit?

Not as findings. Four common ones: lost impression share to budget, limited by target ROAS, improve Ad Strength to Excellent, and a low optimization score. Three ask you to spend more or loosen a target. The fourth grades how many assets you filled in. Lost impression share to budget is worth a look, because a budget cap really can hold back traffic that would have paid. Check the products behind that campaign first. If they are already earning at your target and stock can take it, more budget is a fair call. If they are not, the number is Google's appetite, not your profit. Read all four once, then park them so they do not crowd out the real fixes.

Can I run this audit myself instead of hiring an agency?

Yes, and under roughly EUR 5,000 a month in ad spend you probably should. That is our own line, not an industry rule. A percentage fee on small spend eats budget that should be buying clicks. Our fee is published. 10% on the first EUR 10,000 of monthly Google Ads spend, stepping down to 6% above EUR 150,000. No setup fee. No lock-in. If you want people to check your work for free, the Google Ads eCom Lab on Skool has 1,200+ operators in it. We run that room, so weigh that accordingly.

Still not sure what your five findings mean? The single biggest mistake we see in ecom accounts is usually sitting inside one of them.