We Audited a $50K/Month Google Ads Account. Here's the 7 Things We Found.

Key Takeaways

• We pulled apart a real $50K/month Google Ads account and found seven specific issues, each with a dollar figure attached. The biggest single line item: $19K/month spent cannibalizing organic search.

• One structural pattern was responsible for more than 60% of the wasted spend. Not a tactical mistake. A wiring problem in how the account was built.

• Three of these seven findings show up in nearly every account we audit at $30K/month and above. If you're spending at that level, the odds you have at least one of them are very high.

• Total implementation took two weeks. First measurable lift inside 30 days: blended CAC dropped from $52 to $34 with the same monthly spend.

• The post ends with a short list of things you can check in your own account this week. No tools required, no consultant needed for the first pass.

The Account Before We Touched It

The brand: a DTC apparel company doing healthy revenue, AOV around $95, and spending right at $50K/month on Google Ads. They'd been with their agency for 18 months. The reporting they were getting back every month said 4.1x ROAS and what the agency called "stable performance." Nobody on either side was raising alarms.

The founder reached out because something felt off. CAC had been climbing for two quarters and the agency's answer was always some version of "the platform is more competitive now." He wasn't sure if that was true or if he was being managed. So he asked us to take a look.

The first hour of the audit was the part I want you to picture. Not the deep analysis. The first hour. Search terms report, campaign list, conversion settings, placement reports. That's it. Inside that first hour we'd already flagged four of the seven findings below. By the end of the second day we had all seven, with dollar amounts attached to each one. The agency wasn't doing anything malicious. They were just running on autopilot, which is what happens when accounts get reviewed quarterly instead of weekly.

What follows is the actual list. I've kept the brand anonymous but every number is real.

Finding #1: 38% of Spend Was on Branded Search Cannibalizing Organic

The single largest line item in the account was branded search. Roughly $19K per month was going to keywords like the brand name, the brand name plus product, and the brand name plus "discount code." Click costs ranged from $1.20 to $2.80. The campaign was hitting an 11x ROAS, which on its own looks fantastic.

Here's the catch. We pulled organic search performance for the same queries. The brand was already ranking #1 organically for nearly all of them. The same searcher who clicked the paid ad would have clicked the organic result if the ad hadn't been there. Estimated true cost-per-click in the absence of the paid campaign: about $0.40, because that's what the organic infrastructure was already producing.

So roughly $19K/month was buying clicks the company would have gotten for free. The 11x ROAS was an accounting illusion. The revenue was real, but the spend was incremental in name only.

The fix was to dial branded down to a defensive minimum. We kept just enough budget on it to cover competitor bids on the brand term (which is a real thing and worth defending against), but cut the rest. Roughly $15K/month freed up without losing any actual customers.

Finding #2: Smart Bidding Optimizing for Conversions, Not Profit

The shopping campaigns were running on Target CPA. The target had been set 14 months ago and never revisited. AOV had drifted up since then, but more importantly, the product mix had shifted. The bestsellers were now in a category with lower contribution margin than the products the target was originally calibrated against.

What that meant in practice: Smart Bidding was hitting its target CPA almost perfectly. Every month, right on the number. But the orders it was bringing in had a contribution margin below the breakeven point by about $6 per order on the dominant product line.

The bidding was successfully optimizing for a goal that was no longer aligned with the business.

This is the kind of finding that doesn't show up in a standard agency report because the report says the campaign is hitting its target. It only shows up if someone is checking whether the target is the right one. The fix was rebuilding the bid strategy around target ROAS calibrated to contribution margin, not Target CPA. We also segmented the shopping feed by margin tier so the algorithm could weight high-margin products higher.

Net effect: contribution margin moved from negative to positive on the channel within 21 days.

Finding #3: Display + YouTube Running with No Placement Exclusions

The account had display and YouTube campaigns running for prospecting. Combined spend on those two: about $8K/month. Standard supporting layer, totally reasonable in concept.

The execution was the problem. We pulled the placement report and found that roughly $2.4K/month was running on mobile app inventory inside kids' games. We're talking apps with names like "Bubble Pop Kingdom" and "Hungry Princess Adventure." The clicks were coming from kids tapping on ads while playing games on their parents' phones. Conversion rate from those placements: effectively zero. Some sessions were under three seconds.

There were also placements on news sites with auto-refresh inventory and a long tail of made-for-advertising content farms. None of these had been excluded because the account had no placement exclusion list at all. The agency had set up the campaigns 14 months earlier and never revisited the placement report.

Adding a baseline exclusion list (mobile apps, content farms, kids' content categories, low-quality domains we maintain across all client accounts) cut $2.4K/month of pure waste without affecting the campaigns' ability to find real customers.

Finding #4: 19 Active Campaigns. 5 Doing 92% of the Revenue.

This is the structural finding I mentioned at the top. The one responsible for more than 60% of total wasted spend across the whole account, when you add it all up.

The account had 19 active campaigns. Of those, 5 campaigns were generating 92% of the revenue. The other 14 were collectively producing 8% of revenue while consuming about 31% of total spend. Some of those 14 had been launched as tests 4, 6, even 9 months prior and never paused.

Why does this matter beyond the obvious "you're spending money on bad campaigns"?

Two reasons.

First, every active campaign is competing for the same Smart Bidding learning attention. Google's algorithms perform better when budget and signal are concentrated. Spreading $50K across 19 campaigns gives each one less data per unit of spend, which means less reliable optimization. The 5 winning campaigns were performing despite this dilution, not because of it.

Second, account managers tend to default to "let it ride" with low-volume campaigns because pausing them feels like an admission that the test failed. So they sit there, slowly bleeding budget, until somebody from the outside actually looks. After the audit, we paused 11 of the 14 underperformers immediately and consolidated the budget into the 5 winners. The 3 we kept active had legitimate strategic reasons (a new product launch test, a competitor conquest test, and a geographic expansion test). The other 11 should have been turned off months ago.

Want to know what's hiding in your account?

If you're spending $30K or more per month on Google Ads and it's been a year since anybody outside your agency looked at the account, this is the same audit we just walked you through. We'll show you exactly where the spend is leaking, what the structural patterns look like, and what we'd change if we were running it. No deck. No pitch. No two-week pitch process.

Get Your Free Audit →

Finding #5: Conversion Tracking Was Counting View-Through Twice

This one was the most embarrassing for the previous agency, because it inflated their reported ROAS.

The account had Google Ads conversion tracking installed and also had GA4 conversions imported as a secondary source. Both were configured to count purchases.

The Google Ads tag was set to count view-through conversions on a 30-day window. GA4 was also passing conversions back, including data-driven attribution credit for view-throughs. Same purchase, two systems, both claiming credit.

After de-duping the conversion data and pulling actual purchase records from the Shopify backend, the real ROAS was around 3.2x, not 4.1x. About a 25% inflation in the reported number. The agency had been showing the founder a 4.1x and patting themselves on the back. The actual performance was still positive but materially weaker than what was being reported.

The fix was straightforward: turn off one of the two conversion sources, settle on Google Ads tag as the source of truth for the platform, and reconcile back to actual Shopify revenue weekly. The interesting part wasn't the technical fix. It was that this configuration error had been live for over a year and nobody on the agency side noticed because nobody was reconciling against actual sales. Reporting was being generated from the platform, by the platform, and presented as truth.

Finding #6: No Negative Keyword Maintenance in 9 Months

The negative keyword list hadn't been updated since the account was originally built.

We pulled the search terms report for the trailing 90 days and the picture was rough.

The brand was paying for clicks on:

• Job-search queries (people looking for employment at the brand, not buying from it)

• Internal operational queries (returns, customer service phone number, tracking)

• Competitor research queries from people clearly comparing to other brands

• A long tail of mismatched intent terms that had crept in as Google's match types loosened

Total cost of those clicks over the trailing 90 days: about $7,400, or roughly $2,500 a month. Not the biggest line item in the account but pure waste, dollar for dollar. None of those searchers were going to convert.

The fix was a negative keyword sweep and a recurring 30-day review cadence baked into the account workflow. This is one of those things that takes maybe 90 minutes a month to maintain and saves multiples of that in waste. The previous agency just wasn't doing it.

Finding #7: Audience Layers on Top of Broad-Match Generic Keywords

The last finding is the most strategic. Several campaigns were running broad match on generic, top-of-funnel keywords ("women's apparel," "summer dresses," that level of generality), and the agency had layered "in-market for apparel" audiences on top.

The thinking, presumably, was that broad reach plus audience targeting would find buyers efficiently. In practice, this combination nullifies both. Broad match on a generic keyword tells Google "go wide." The in-market audience tells Google "find buyers." But Google's bidding algorithm has to choose which signal to weight, and with broad match on a generic keyword, the keyword signal is so noisy that the audience signal can't compensate. The result was a campaign with low CTR, low conversion rate, and high CPCs because the auctions it was entering were the wrong auctions.

Two cleaner setups: phrase or exact match keywords with audience layering, OR broad match with strong conversion data feeding Smart Bidding (no audience layer). Mixing the two was the worst of both worlds. We rebuilt the campaign with phrase match keywords and kept the in-market audiences as observation only. Performance jumped within the first week.

What Changed in 30 Days After the Fixes

Implementation took two weeks across all seven findings. Most of the changes were either pauses, exclusions, or reconfigurations, so the account didn't have to learn anything new for most of them. The bid strategy rebuild was the slowest piece because Smart Bidding needs about 14 days of conversion data to recalibrate.

Here's what the numbers looked like at day 30, with monthly spend held constant at $50K:

• Blended CAC dropped from $52 to $34

• Channel contribution margin moved from negative to positive by about $4.20 per order

• Revenue grew 18% on the same spend

• True ROAS (reconciled against Shopify, not platform-reported) climbed from 3.2x to about 4.6x

The point of sharing those numbers isn't that we're miracle workers. It's that the previous agency had been reporting 4.1x ROAS the whole time the real number was 3.2x, and a meaningful portion of the "improvement" we delivered was just measuring honestly.

The Three Findings Most Likely to Be Hiding in Your Account

We've run versions of this audit on about 40 accounts in the $30K to $250K monthly spend range over the last two years. Three of the seven findings above show up in nearly every one of them.

Branded search overspend (Finding #1) is almost universal. Most agencies leave branded campaigns running at full tilt because they juice the reported ROAS. If you're running paid branded campaigns, pull the organic ranking for the same terms and ask whether you'd lose those customers without the ad. In about 70% of cases, the answer is no.

Stale Smart Bidding targets (Finding #2) show up in roughly 60% of accounts we audit. Targets that were set during onboarding and then never revisited as the business evolved. If your AOV, product mix, or margin profile has shifted in the last year and your bid targets haven't been touched, this is almost certainly happening.

Campaign sprawl (Finding #4) shows up in about 55% of accounts. The pattern is consistent: more campaigns than the account can support, most of the budget being carried by 20-30% of them, the rest left on because nobody wants to call the test a failure. If you have more than 8-10 active campaigns in a single account, run a quick Pareto check on revenue contribution. The answer is usually uncomfortable.

You can do that check this week without any tools beyond the Google Ads UI. Pull the search terms report and look for off-intent queries. Pull the campaign list sorted by revenue and see how the curve looks. Look at when your bid targets were last modified.

Those three checks alone will tell you most of what you need to know.

FAQ

What does a Google Ads audit actually look at?

A real audit goes through the search terms report, campaign structure, bid strategy configuration, conversion tracking setup, placement reports for display and YouTube, audience layering, and reconciles platform-reported revenue against actual sales data from your eCommerce backend. The walkthrough above covers most of it. What it doesn't cover is creative review (a separate process) and landing page review (also separate, though we touch on it when it's clearly a problem).

How long does a Google Ads audit take?

Ours takes 3-5 business days from the time we get account access. The first hour usually surfaces 60-70% of the findings. The remaining time is for verifying the dollar amounts, pulling reconciliation data from your Shopify or other commerce platform, and writing up the findings in a way that's actionable rather than a 40-page deck.

How can I tell if my Google Ads account has wasted spend?

Three quick checks. One: pull your search terms report for the last 90 days and scan for off-intent queries (job searches, returns, support, competitor research). Two: look at your campaign list sorted by revenue and see what percentage of campaigns are producing 90% of the revenue. Three: check when your bid strategy targets were last modified. If any of those three turn up something uncomfortable, you have wasted spend.

Is a free Google Ads audit actually free?

Ours is. We don't charge for the audit, we don't require a contract to look at the account, and we don't gate the findings behind a sales process. You'll get the actual list of what we found. If you want us to fix it, that's a separate conversation. If you want to take the findings to your existing agency or fix it yourself, that's fine too. The audit is its own thing.

How often should I audit my Google Ads account?

If you have a competent agency or in-house team running the account, an outside audit once a year is enough. If you're at $30K/month or above and haven't had outside eyes on the account in the last 12 months, you're overdue. Internal reviews should happen monthly at minimum, weekly on the search terms report and placement reports specifically.

Want to See What's in Your Account?

If your account is over $30K/month and hasn't had outside eyes on it in the last 12 months, you're almost certainly sitting on findings that look a lot like the ones above. We've seen this pattern enough times to know it's not exceptional. It's just what happens when accounts don't get reviewed by someone whose job isn't to keep them looking good.

We'll show you exactly what's happening in your account. No contract, no pressure, no two-week pitch sequence. Just the actual findings, with the actual dollar amounts, in language you can take back to your team or your agency.

Get Your Free Audit →