Est. reading time: 11 minutes
The most expensive mistakes in Google Ads don’t happen inside the ad account. They happen in the meeting where someone looks at a report, draws the wrong conclusion, and makes a budget decision based on that conclusion.
We’ve sat in those meetings. A business owner sees the total spend number, divides it by the revenue they can directly trace to Google Ads, and decides the channel isn’t working. Budget gets cut. Leads dry up. Six weeks later, revenue declines across the board and nobody connects it back to the budget decision because the effects are indirect and delayed.
Or the opposite: the in-platform ROAS looks incredible, so budget increases. But the ROAS was inflated by brand campaigns cannibalizing organic traffic, and the incremental revenue from the additional spend is a fraction of what the dashboard suggested.
Both scenarios play out constantly. The problem isn’t that businesses are looking at the wrong dashboard. It’s that they’re interpreting the data without understanding what it actually represents, what it leaves out, and where the numbers lie to you if you take them at face value.
Brand Campaigns Distort Everything
This is the single biggest source of misread Google Ads performance, and almost every account we audit has this problem.
Brand campaigns are campaigns that bid on your own company name and close variations. Someone types “Tailored Edge Marketing” into Google, your brand ad appears at the top, they click it, and they convert. In the Google Ads dashboard, that conversion is attributed to the brand campaign with a very low cost per click and an extremely high ROAS.
The problem: most of those people were going to find you anyway. They already knew your name. They were navigating to your site. Without the brand ad, the vast majority would have clicked the organic result directly below it. The brand campaign didn’t generate that demand. It intercepted it, and then Google Ads took credit for the conversion.
We’re not saying brand campaigns are worthless. There are legitimate reasons to run them: protecting your brand name from competitors bidding on it, controlling the messaging in the top position, and capturing clicks on branded queries where the organic result might appear below competitor ads. These are valid defensive strategies.
But when you combine brand campaign performance with non-brand campaign performance in a single report, the blended numbers are meaningless. Brand campaigns will always have a lower CPA and higher ROAS because they’re converting people who were already looking for you. Non-brand campaigns, the ones actually generating new demand, will always look worse by comparison.
The businesses that make good budget decisions separate these completely. Brand performance is reported on its own, evaluated as a defensive cost, and kept at the minimum budget needed to maintain coverage. Non-brand performance is reported separately and evaluated as the true measure of whether Google Ads is generating new business. When these numbers are blended, the brand campaigns inflate the overall metrics and mask whether the growth-driving campaigns are actually performing.
We’ve seen accounts where the blended ROAS was 8x, the leadership team was thrilled, and the non-brand ROAS was 1.2x. The account wasn’t driving profitable growth. It was spending money to intercept branded searches at a great reported return, while the campaigns meant to generate new customers were barely breaking even. Without the split view, that problem was invisible.
Last-Click Attribution Lies to You in a Specific Way
Google Ads defaults to last-click attribution within the platform, which means conversion credit goes to the last ad click before the purchase or lead submission. This creates a systematic bias that’s important to understand.
Paid search often captures demand at the bottom of the funnel. Someone sees a Meta ad, visits your site, leaves, thinks about it for a few days, searches your product category on Google, clicks your search ad, and buys. In Google Ads reporting, that conversion is 100% attributed to the search click. Meta gets nothing. The Google Ads ROAS looks fantastic, and the Meta ROAS looks weak.
But Google didn’t create that customer’s interest. Meta did. Google converted the interest into a transaction, which is valuable, but the full story of how that revenue was generated spans both channels. Evaluating each channel in isolation, using its own attribution, overstates Google’s contribution and understates Meta’s.
This isn’t a theoretical problem. It drives real budget misallocation. We’ve worked with businesses that shifted budget from Meta to Google based on platform-reported ROAS comparisons. Initially the results looked positive because Google’s reported ROAS held strong. But within four to six weeks, Google’s performance started declining because Meta had been generating the top-of-funnel demand that Google was converting. Less Meta spend meant fewer people entering the funnel, which meant fewer branded and category searches for Google to capture.
The practical solution isn’t to abandon platform-level attribution. It’s to supplement it with metrics that reflect the full picture.
Blended ROAS (total revenue divided by total ad spend across all channels) tells you whether your overall marketing investment is generating an acceptable return, regardless of which channel gets credit. If blended ROAS is healthy and growing, your channel mix is working even if individual channel attribution is messy.
New customer acquisition cost matters more than overall CPA. If Google Ads is efficient at converting existing demand but isn’t bringing in new customers, the efficiency is less valuable than it appears. Segment your Google Ads conversions by new versus returning customers wherever possible.
Incrementality questions keep you honest. For any channel, the key question isn’t “what does this channel claim it generated?” It’s “what revenue would we lose if we turned this off?” You can approximate this with controlled spend-down tests: reduce budget on a campaign for two to four weeks, measure what happens to total business revenue (not just attributed revenue), and compare. The gap between the dashboard number and the business-level impact is your incrementality reality check.
The Campaign Types Tell Different Stories
Not all Google Ads campaigns should be evaluated the same way, and applying a single ROAS or CPA target across campaign types is one of the most common evaluation mistakes we see.
Search campaigns targeting high-intent keywords are your closest-to-revenue campaigns. Someone searching “buy [your product]” or “[your service] near me” has strong purchase intent. These campaigns should be held to your strictest efficiency targets because they’re capturing demand that’s ready to convert. If these aren’t performing, the issue is usually landing page quality, offer competitiveness, or ad relevance, not the channel itself.
Search campaigns targeting informational or research-stage keywords are further from revenue. Someone searching “best CRM for small business” or “how to fix a leaky faucet” is earlier in the decision process. These campaigns will always have a higher CPA and lower ROAS than high-intent campaigns because the searcher needs more time and touchpoints before converting. Evaluating them against the same benchmarks as bottom-funnel campaigns will always make them look like failures.
The right approach is to evaluate these campaigns on their contribution to the pipeline, not on immediate conversion. Track assisted conversions, not just last-click conversions. Look at whether these campaigns are introducing new users who later convert through other channels or campaigns. If a research-stage campaign generates a first visit that leads to a branded search conversion two weeks later, that campaign contributed real value that last-click attribution completely misses.
Performance Max campaigns are Google’s automated, cross-channel campaign type that runs across Search, Display, YouTube, Gmail, Maps, and Discovery simultaneously. They’re increasingly pushed by Google and increasingly adopted by advertisers who like the simplicity.
The evaluation challenge with Performance Max is transparency. Google provides limited visibility into which channels and placements are driving the results. A Performance Max campaign might report strong ROAS, but if the majority of conversions are coming from branded search queries that would have converted anyway, the reported efficiency is misleading.
We approach Performance Max with cautious optimism. It can work well, especially for ecommerce with large product catalogs. But we always run it alongside standard search campaigns rather than as a replacement, and we closely monitor the search terms it’s matching to (using the Insights tab, since the search terms report for PMax is limited) to ensure it’s not just cannibalizing branded traffic at a premium.
Display and YouTube campaigns are awareness and consideration channels, not direct response channels. Evaluating them on last-click ROAS will always produce disappointing numbers because that’s not how these channels work. They create demand that other channels convert. View-through conversions, audience list building for remarketing, and top-of-funnel reach are the appropriate metrics. If your leadership team expects display campaigns to produce the same ROAS as branded search, the problem is the expectation, not the campaign.
Spend Efficiency vs. Spend Level: The Scaling Trap
This is a subtler evaluation mistake, but it costs businesses significant money.
There’s a natural ceiling on paid search efficiency at any given spend level. For most businesses, the highest-intent, lowest-competition keywords get captured first. As you increase budget, you’re expanding into more competitive auctions, broader keywords, and audiences with progressively lower intent. Cost per acquisition naturally rises as spend increases. This isn’t the campaign failing. It’s the economics of scaling any auction-based channel.
The mistake happens when a business sees strong CPA at $10,000/month, increases to $30,000/month expecting the same CPA, and then panics when CPA rises by 40%. They conclude the campaign “stopped working” and either slash back to the original budget or restructure the account, when the real answer is that $30,000/month simply accesses a different efficiency frontier than $10,000/month.
The question isn’t whether CPA is the same at both spend levels. It’s whether the incremental conversions at the higher spend level are profitable. If your target CPA is $50, you were achieving $30 at the lower budget, and you’re achieving $45 at the higher budget, the campaign is still performing within target. You’re getting more customers at a slightly higher cost each. That’s usually a good trade.
We model this out before recommending budget increases. For any given account, there’s a curve of diminishing returns. The first $5,000 in monthly spend captures the highest-intent queries at the lowest cost. Each additional $5,000 increment produces incrementally more expensive conversions. At some point, the marginal CPA exceeds the threshold where additional customers are profitable. That’s the point where the budget should stop, not the point where CPA matches the original low-spend level.
When we present budget recommendations to clients, we show this curve explicitly. Not “spend more and get more,” but “here’s what the next $5,000 increment is likely to produce, here’s the expected CPA at that level, and here’s whether that’s profitable based on your unit economics.” This framing prevents the cycle of increase, panic, slash that we see in so many accounts.
What Good Google Ads Evaluation Actually Looks Like
The businesses that make consistently good decisions about Google Ads share a few practices.
They separate brand and non-brand performance in every report. They never allow blended metrics to obscure whether their growth campaigns are actually generating profitable new business.
They evaluate campaign types against appropriate benchmarks. High-intent search gets strict efficiency targets. Research-stage search gets evaluated on pipeline contribution. Display and YouTube get evaluated on awareness and audience-building metrics. Performance Max gets scrutinized for branded search cannibalization.
They supplement platform-reported attribution with blended business-level metrics. Total revenue relative to total ad spend. New customer acquisition cost. Periodic incrementality checks through controlled spend changes.
They understand the relationship between spend level and efficiency, and they make budget decisions based on whether incremental conversions are profitable, not on whether CPA matches a number it was never going to match at a higher spend level.
None of this requires sophisticated attribution modeling or expensive analytics tools. It requires separating the data correctly, asking the right questions, and resisting the temptation to make decisions based on whichever number makes you feel best or worst on any given day.
The biggest risk in Google Ads isn’t a poorly structured campaign. It’s a well-structured campaign that gets killed or starved because someone read the data wrong. Getting the evaluation right protects the investment, and that’s worth more than any bid adjustment or keyword optimization.
