5 Google Ads Mistakes Burning Your Budget

Your Google Ads ROAS can look great and you can still have no profit.

We run ads for 200+ ecom brands at ZenoX. Chris sees these same five leaks in new accounts and in stores that spend millions a year. You can still make money with them. You just do not see how much profit they take.

Why These 5 Leaks Hide Behind a Good ROAS

Return on ad spend (ROAS) is money back per dollar you put into ads. A strong number can still hide a bad mix.

Google, Meta, TikTok, and Pinterest all want that number to look strong. Maximize Conversions and Maximize Conversion Value tend to favor the easiest conversions. In a mixed audience, repeat buyers can win more bids unless new-customer controls separate them. Performance Max can mix cheap YouTube clicks into a Shopping campaign. The average cost-per-click (CPC) still looks fine.

That is why these five survive in accounts that already "work." A good-looking ROAS can hide all five.

Chris walks through all five in this video:

G

Once the leaks are plugged, the growth rules sit in how to scale Google Ads.

Mistake 1: Paying Twice for Customers You Already Own

The biggest leak we see, even in million-dollar brands, is overpaying for warm traffic you already own.

You retarget past the sweet spot. Subscription brands and stores with lots of return buyers do this the most. They spend hard to pull back people who were going to buy again anyway.

Paid ads should hunt new customers. Repeat buyers are where the money lives. That is exactly why you should not overpay to chase them with ads. There are smarter ways to keep them in the loop.

Google does not care about that split. If your bid strategy is Maximize Conversions or Maximize Conversion Value, Google focuses on conversions. A bid strategy is the rule Google uses to pick which clicks to buy. You have new people and existing customers. Repeat buyers convert more easily, so Google keeps buying them unless you tell it to stop.

The video describes one account with an unusually high share of spend going to existing buyers. Treat it as an example, not a benchmark. Audit your own new-versus-existing split before deciding whether retargeting is too high. You may pay two, three, four, or five times what you should for someone who already loves the brand.

This matters more as the brand grows. If you are still small, you do not have a big existing base yet. You will. Set the split before it gets expensive.

You need control over where the money goes. Separate new and existing buyers before their costs blur together. We will not turn this into a setup course here.

Mistake 2: Never Letting a Campaign Settle

The second leak is never letting a new campaign, a new creative, or a new account settle. Give it a different change every day and it cannot learn or see patterns, because that needs stability. We have written the full deep-dive here: the biggest Google Ads mistake ecommerce brands make.

Mistake 3: One Target ROAS for Every Product

A flat target ROAS across the whole catalog builds your own ceiling.

Target ROAS means you tell Google the return you want, and it bids to hit that number. People ask Chris how to set it. The honest answer is that it depends. Product mix, the feed, and how much data you have all change the right number. Targets behave differently across product groups, so judge each change against sales volume and profit. A product group is a slice of the catalog, like winners, new items, or wasters.

You want a 3x ROAS, so you set the target to 3 and wait to grow. That target may cap sales before you learn the potential.

Winners, new products, and wasters should not share one target.

Winners have data. You can push them more. Push too hard and volume drops sharply.

New products need room. If you want to test and grow new winners, a high target chokes them.

Wasters and low performers belong in a tighter box. These are products that spend without making money. Isolate them. Control the budget. Set a higher target so they cannot wander.

This sits next to campaign structure. Judge each product group by how much data it has and whether it makes profit. A high target can leave you with few sales. A low target can bring volume without profit.

There are exceptions. Sometimes we take a bigger step. Usually we do not.

And remember what you are steering. Target ROAS is an easy campaign metric. Chris's line is the one we use internally: ROAS is a vanity metric. You want to optimize for profit. A 3x ROAS means nothing until you know the margin.

Mistake 4: Trusting Google to Grade Its Own Homework

Google tracks the sales you then use to judge Google. Meta does the same. So do TikTok, Pinterest, and Snapchat. Every platform has a reason to report a great ROAS.

You see a 5x in Pinterest. You double the budget. Meta looks amazing too. Google looks fine. Then you look at the bank. No profit.

It happens because you run more than one platform. A buyer might see you on Pinterest, click Google later, and see a Meta ad somewhere in between. All three want that conversion. You have one sale, and you see three in the ad accounts. The 5x was probably closer to a 2x.

This gets worse when view-through is on. That means they saw the ad but never clicked, and the platform still takes credit if they buy later. A window is how long after someone sees or clicks an ad the platform still claims the sale. Pinterest uses a 30-day window for a click and a 30-day window for a view. Either one counts as a conversion for Pinterest. Someone can see the ad, then buy through Google or organic 28 days later. Pinterest still counts the sale.

Tools like Triple Whale help. Understanding the attribution settings in each ad account matters more than any tool.

Blended ROAS compares all ad spend across platforms with real sales.

The Client Story: A 0.8 ROAS Hiding Behind a Great Number

A client came to Chris and said we make no profit. Scale Google back immediately.

That instinct was wrong. Google was delivering at least a 30% net profit margin. After costs, about 30 cents of every dollar from Google stayed as profit. The results were genuine.

He was also running Meta, TikTok, and Pinterest. Chris got on a call, walked through how attribution works, and they installed Triple Whale.

Then the picture was obvious. Google attributed 3x to 4x ROAS, sitting near that 30% net margin. TikTok was running at a ROAS of 0.8. That is a straight loss. Pinterest was 1-point-something. Also a burn.

The dashboards on TikTok and Pinterest still said they were crushing it. They wanted more budget. The money was leaving through those channels, not through Google.

After they checked blended results, they cut spend on the loss-making channels instead of Google.

Google ROAS

3-4x

Net from Google

~30%

TikTok ROAS

0.8

Pinterest ROAS

1.x

“

Always look at the blended ROAS, understand math in general, understand your own data, and don't scale blind. The number there is not real in the first place.

We still check ad account ROAS. We never decide from it alone.

Mistake 5: The Performance Max Black Box

Performance Max is not Shopping only. People treat it like it is. By default it can spend on Shopping, Search, Display, YouTube, and Discover.

We run most clients on a PMax feed-only setup. Feed only means we do not add headlines, images, or videos, so Google stays mostly on Shopping. It is not new. People who work with us already know it.

You can have a strong Shopping ROAS while cheap Display and YouTube clicks burn money beside it. Search terms are another black box because Google shows only part of them.

Remember Mistake 1 here too. PMax will go where it gets the best ROAS, which is often your own brand and existing customers. Split that if you are a bigger brand.

How to Check a Feed-Only PMax Campaign

There is no sealed feed-only box. Open the channel report and find where PMax spent. Check Shopping CPC by itself. Cheap Display and YouTube clicks can make the blended CPC look lower than the Shopping CPC.

Review the setup boxes that can let other placements into the campaign. Then use the channel report to find Display or YouTube spend and correct that setup.

Review search terms separately. Add account-level negatives for irrelevant or competitor queries. A negative is a search phrase you do not want to pay for.

For the settings that leak PMax budget, use Fix Your Performance Max Campaign That Is Wasting Spend. For why the catalog is often the constraint, read Performance Max isn't the problem, your feed is. The structure data behind our asset-group splits is in the study of 12,000 Performance Max campaigns.

How to Check Your Account for All 5 This Week

Do not look at ROAS and call the week. Break the account into variables. Find the pattern. That is the same point Chris makes in this short: "performance is bad" is a lazy answer until you know which lever broke.

Y

Run these five checks.

1. New vs existing. Check how much Maximize Conversions spends on repeat buyers. Split campaigns or turn on new-customer targeting so Google seeks new people.

2. Stability. Review the changes made this week. If you changed the campaign every day, stop and let it settle. The full operating rules live in the biggest Google Ads mistake ecommerce brands make.

3. Targets by group. Check whether every product group shares one target ROAS. Split them when their data and profit differ. Change one target at a time, in small moves. Details are in when and how to use target ROAS.

4. Blended vs dashboard. Add up Google, Meta, TikTok, and Pinterest claimed revenue. Compare it to real orders and bank profit. If claimed revenue is far above real orders, stop raising channel budgets until you check attribution.

5. PMax location of spend. Run the feed-only check above and record where the campaign spent its budget.

Frequently Asked Questions

What is the biggest Google Ads mistake ecommerce brands make?

Overpaying for customers you already own. Maximize Conversions hunts the easiest sales, which are often repeat buyers. We have seen accounts where an unusually large share of budget went to people who already bought. That is one example, not a rule for every store. You can still pay two to five times what you should for someone already loyal. Split campaigns or use new-customer targeting so Google hunts new buyers.

Why does my Google Ads ROAS look great but I have no profit?

Every platform reports its own ROAS, and each one wants that number to look strong. One buyer can see you on Pinterest, click Google, and still get counted three times. A client wanted to cut Google because he had no profit. Google attributed 3x to 4x ROAS with about 30% net margin. TikTok was 0.8. Pinterest sat around 1x, also a loss. The dashboards looked fine. The bank did not. Use blended ROAS and real profit.

Should I set the same target ROAS for every product?

No. One flat target across the catalog builds your own ceiling. Winners have data, so you can push them, but a hard push can kill volume. New products need a looser target so they can grow. Wasters need a higher target, isolation, and a controlled budget. Move targets in small steps. Do not jump from 190% to 250% and expect the same output. Steer for profit. ROAS is only the campaign metric.

Why do Meta, TikTok, and Pinterest all show a better ROAS than my bank account?

They each want to claim the conversion. When you run several platforms, one sale can show up in three ad accounts. Pinterest is especially heavy on view-through. With 30-day clicks and 30-day views, someone can see the ad and convert through Google or organic 28 days later, and Pinterest still counts it. Attribution tools like Triple Whale help you see this. You still have to read attribution settings. Trust blended ROAS and the bank.

Is Performance Max "feed only" actually feed only?

No. We run most clients on a PMax feed-only setup to push spend into Shopping. Google can still drop in cheap Display and YouTube clicks, especially if setup boxes are left ticked or the algorithm is confused. Blended average CPC can look fine while the Shopping CPC that matters is much higher. Audit search terms, add account-level negatives for junk and competitor queries, and check where spend actually goes.

How much should I spend retargeting customers who already bought?

There is a sweet spot, and going past it wastes money. Keep some retargeting. Do not let it eat the account. Paid ads should focus on new customer acquisition. Repeat buyers are where profit lives, which is why you should not overpay to recapture them. The bigger your repeat-buyer base gets, the more this costs you. If you mix new and existing in one Maximize Conversions campaign, Google will keep buying the easy repeats. Split them.