by Austin LeClear
on September 28, 2026
TLDR: We personally audited roughly $10 million in ecommerce ad spend in 2026, on top of close to a thousand recorded account audits before that. The same five mistakes kept showing up in accounts of every size, including brands doing $50 million a year in revenue. None of them are exotic. All five are checkable in your own account this week: brand terms leaking into Performance Max, ROAS treated as a north star instead of a guidance metric, oversegmented campaign structures, Target ROAS goals set so high they starve the account, and margin or AOV tiers mixed inside the same bid strategy.
Prefer video? Check out I Audited $10,000,000 in eCommerce Ad Spend: They All Made the Same 5 Mistakes on YouTube:
Still with us? Good, let’s dig in.
Mistake 1: Brand Terms Leaking Into Performance Max
A lot of Performance Max campaigns are still picking up brand searches without anyone noticing, because PMax does not separate brand from non-brand the way a Search campaign does.
Brand searches convert at a much higher rate than non-brand searches, since the person searching your brand name already knows who you are. When those searches flow into a PMax campaign, they inflate the campaign’s blended performance.
In several audits, filtering brand terms out of the search term report turned a campaign that looked excellent into one that was actually unprofitable. Take brand out of the numbers, and the real picture can look completely different.
The fix is simple. Add your brand terms as negative keywords or use PMax’s brand exclusion setting on every Performance Max campaign you run. Keep brand traffic in its own Search campaign, or its own brand Shopping campaign, where you can see and control it directly.
We’ve written the full walkthrough here: How to Exclude Brand Terms from Your Performance Max Campaigns
Mistake 2: Treating ROAS as the North Star
ROAS is a guidance metric inside the Google Ads account. It is not supposed to be the north star for the business.
Plenty of ecommerce brands, some of them highly successful, have never cross-referenced their in-platform ROAS against what is actually happening on the back end. The real north star is new customer growth at unit economics that make sense for the business.
That means pulling numbers directly from Shopify: new customers per week or month, and what it actually costs to acquire them profitably.
A lot of these brands have simply never built that cross-reference. They only ever looked at what Google Ads reported.
The fix is a simple sheet that tracks new customers and true acquisition cost from your backend data, then cross-references that against what your Google Ads campaigns are doing. When you increase spend or launch a new campaign, that sheet tells you whether it actually moved profitable new customer growth, not just whether the in-platform ROAS number went up.
Mistake 3: Oversegmentation
Even in 2026, oversegmented account structures are common. One audited account was spending $20,000 to $30,000 a month and running 25 separate Performance Max campaigns.
Oversegmentation splits your conversion data across too many campaigns, which starves each one of the volume it needs to optimize well. An account can plateau on structure alone, independent of budget size.
If your account is running more than about 10 PMax campaigns, that is a signal you are probably oversegmented. That threshold does not change just because you are spending more. Segmentation is justified when there is a real strategic reason: a different country, a large margin difference between product lines, or a large AOV difference.
Without a reason like that, consolidate. Consolidating an oversegmented account has repeatedly improved performance rather than hurt it.
Here’s what that actually looks like: How We Added $1.3M in Revenue for an E-commerce Brand With Just $60K More in Spend
Mistake 4: Unrealistic Target ROAS Goals
Some accounts are running Target ROAS bid strategies as high as 1,500%, expecting extreme efficiency, and instead starving the campaign of the auctions it needs to spend in at all.
Rising costs, including tariff pressure and the steady year-over-year rise in Google Ads CPCs, have squeezed a lot of brands. The instinct is often to chase the high returns the account used to get by pushing the Target ROAS goal even higher.
That instinct is the same mistake as treating ROAS as the north star. A campaign asking for a 1,500% return will not enter enough auctions to spend, let alone scale.
The fix is to re-evaluate any bid strategy set to an extreme Target ROAS. Your bid strategy should correlate with what the backend data says is actually profitable for the business, not with what the account used to return in a different cost environment. A campaign that is plateaued or shrinking under an extreme ROAS goal needs a new target, not a higher one.
Mistake 5: Mixing Margin and AOV Inside the Same Campaign
Products with a 10% margin sitting in the same campaign as products with a 70% margin. Ten-dollar items in the same campaign as thousand-dollar items. Both show up often, sometimes in the same account that was just oversegmented somewhere else.
A bidding algorithm chases the path of least resistance. Given a single Target ROAS and a mix of a $10 product and a $1,000 product, it will lean toward selling more of whichever is easier to convert, as long as it still hits the target. That limits exposure for higher-ticket products.
It can also swing the other way. One high-AOV sale spikes the campaign’s ROAS, and the bidding algorithm gets aggressive off a single data point.
The same logic applies to margin. A 70% margin product supports aggressive bidding. A 10% margin product does not, and mixing the two into one bid strategy usually costs money on the low-margin side.
This is not a case for oversegmenting everything into its own campaign. It is a case for thinking strategically about where real AOV and margin differences exist in your catalog, and not defaulting to one campaign, one bid strategy, one target for products that do not share similar economics.
Mistakes 3 and 5 Are Not the Same Thing
Mistake 3 says don’t split your account into too many campaigns. Mistake 5 says don’t put very different products in the same campaign. Read quickly, those can sound like they disagree. They don’t.
Ask one simple question: is there a real, meaningful difference in margin, AOV, or geography between what you are splitting? If yes, keep it separate. If no, put it together. A few campaigns split for a real reason is not oversegmentation. Twenty-five campaigns split by minor subcategory, with no real difference between them, is.
What to Check in Your Own Account This Week
- Filter your Performance Max search terms for your own brand name. If it shows up, you have mistake 1.
- Pull your actual new customer count and acquisition cost from your backend data and compare it against your in-platform ROAS. If you have never done this, you have mistake 2.
- Count your active Performance Max campaigns. More than 10, on anything short of an unusually large and genuinely diverse catalog, is worth a structural review.
- Check your Target ROAS goals against what your account is actually returning right now, not what it returned a year or two ago.
- Look at whether your campaigns mix meaningfully different margins or AOVs under a single bid strategy.
None of these require a rebuild to check. They require about twenty minutes and access to your own account.
FAQs: Google Ads Mistakes
Why would brand terms end up in my Performance Max campaign if I never targeted them?
Performance Max uses automated targeting instead of the keyword lists you control in a Search campaign, so it can serve on brand searches without you ever adding them. Unless you explicitly exclude your brand as a negative keyword or through PMax’s brand exclusion setting, this can happen without any warning inside the interface.
Is ROAS a useless metric?
No. ROAS is a genuinely useful guidance metric for managing bids and budget inside the Google Ads account. The mistake is treating it as the business’s north star instead of new customer growth at profitable unit economics, which is a backend calculation ROAS alone cannot give you.
How many Performance Max campaigns should an ecommerce account run?
There is no fixed number, but more than about 10 is a signal worth investigating unless there is a clear strategic reason, like separate countries or sharply different product margins. Most oversegmented accounts we have audited had no such reason, just campaigns split by minor subcategory.
What Target ROAS should I actually be using?
It depends on your unit economics, not on what the account returned in a past cost environment. Rising CPCs and other cost pressures mean an old target that used to work can now be unrealistically high, which will starve the campaign of the auctions it needs to spend in.
Should I put all my products in one campaign to avoid oversegmenting?
No. The goal is not to avoid segmentation entirely. It is to segment for real economic reasons, like a large margin or AOV difference, rather than splitting by minor subcategory. Mixing a $10 product with a $1,000 product under one bid strategy causes its own problems.
Final Thoughts
None of these five mistakes require new tools, new platforms, or a rebuild from scratch. They require actually looking at the search terms feeding your Performance Max campaigns, at the backend numbers your Google Ads reporting never shows you, at how many campaigns you are really running and why.
The accounts that keep scaling profitably are not the ones with some hidden advantage. They are the ones where somebody checked.
Get a personalized, 1-on-1 Google Ads account review.
Not sure if one of these five mistakes is hiding in your own account?
We offer a free account analysis for advertisers who qualify. It’s a custom-tailored, in-depth review of your Google Ads account that reveals day-one improvements to your campaigns, the same kind of audit behind everything in this article.
We don’t take every account that applies. We work with businesses where we’re confident we can deliver real results, so we’ll need a few details on your business and goals before we book the call.
If that sounds like a fit, click below to apply.
