I spent nine years inside Google. Not running campaigns from the outside, but actually inside the company, managing relationships with some of the largest advertisers in North America, watching how the platform really works from the inside out.

And here's what I noticed consistently: the things Google tells advertisers publicly and the things that actually drive performance are not always the same conversation.

Most agencies don't know this because they've never been on the other side of the table. They learn Google Ads from help docs, certification courses, and what Google reps recommend on calls. I learned it from the people who built it.

I’m here to tell you seven things I saw at Google that most advertisers and agencies still get wrong, and some of them are quietly costing businesses thousands of dollars every month.

1. Google Reps Are Incentivized to Increase Your Spend — Not Your ROI

I’m going to start with the most uncomfortable truth. 

When you get a call from a Google account rep recommending you increase your budget, enable broad match, or turn on a new campaign type, you must understand what's happening behind that call. Google reps have revenue targets. Their performance is measured by how much advertiser spend grows in their portfolio, not by whether your CAC goes down.

I was inside that system. I watched reps push recommendations that helped their numbers, not the advertiser's. The advice isn't malicious; it's structurally incentivized in a direction that doesn't always align with your goals.

The rule I'd apply: treat every recommendation from a Google rep as a hypothesis, not a directive. Test it with a small budget before you scale it. And whenever a rep tells you to "let Smart Bidding do more," ask yourself: more of what, exactly?

2. Quality Score Is the Most Undervalued Lever in the Entire Platform

Most advertisers treat Quality Score as a report card. Something you glance at and move on from. In reality, it's the single most impactful variable in determining how much you pay per click.

Here's the number that should stop you: a Quality Score of 7 reduces your CPC by 28% compared to a baseline score of 5. A Quality Score of 10 reduces it by 50%. 

On a $10,000/month account, closing that gap from a 5 to a 7 is $2,800 back in your pocket every single month, without changing your bids, your campaign structure, or anything else. Just improving relevance.

Quality Score is determined by three things: expected click-through rate, ad relevance, and landing page experience. Most agencies focus almost entirely on bidding and ignore the fact that improving these three components would reduce their client's cost-per-click faster than any bid adjustment.

When PAM, a voice AI platform for car dealerships, came to me, their previous agency had built campaigns with generic ad copy and mismatched landing pages. Quality Scores were low, CPCs were high, and the account was burning through budget with no clear returns.

After rebuilding the account with tight keyword-to-ad-to-landing-page alignment, PAM generated $234K in contract value from $72K in ad spend; a 3.25x ROAS. The campaign mechanics were not radically different. The Quality Score was.

3. The Auction Doesn't Work the Way Most People Think It Does

People assume Google Ads is a bidding war. Whoever pays the most, wins. That's not how Ad Rank works.

Ad Rank = your bid × Quality Score × expected impact of extensions and formats.

This means a competitor bidding $10 with a Quality Score of 4 has an Ad Rank of 40. You bidding $5 with a Quality Score of 9 have an Ad Rank of 45. You win the auction at roughly half the CPC.

The implication is significant. You can outrank better-funded competitors without matching their budgets if your relevance signals are stronger. I watched this happen repeatedly at Google with clients who thought they needed to spend more to compete. Usually they needed to optimize more.

The agencies that understand auction dynamics build accounts differently. Tight ad groups. High CTR signals. Landing page copy that matches search intent exactly. This is not theory; it's how the auction is designed.

4. Smart Bidding Is Only as Smart as the Data You Feed It

There's a belief in the industry that Smart Bidding is a magic box. Set it to Target CPA, walk away, results follow. I need to correct this directly.

Smart Bidding is a machine learning algorithm. Like every machine learning system, it learns from the data it receives. If the data is bad, wrong conversion actions are set as primary, form fills from unqualified leads are counted the same as demo requests from enterprise buyers, no CRM signal about which leads actually closed.

In simple words, the algorithm learns to optimize toward the wrong outcomes.

I've audited dozens of accounts where Smart Bidding was confidently, efficiently producing leads that sales couldn't close. The algorithm was working perfectly. In reality, the inputs were broken.

Smart Bidding needs at minimum 30–50 quality conversions per month in a 30-day window to train properly. Before you reach that volume, manual CPC with conversion tracking in place is more predictable than handing control to an algorithm that doesn't yet have enough signal. This is something Google reps rarely tell you because it means spending less until the account is ready.

The deeper fix: Connect your CRM so that closed-won revenue flows back to Google Ads as an offline conversion. This is what actually makes Smart Bidding intelligent for B2B. And that's what Google Ads attribution for B2B SaaS is designed to solve.

5. Budget Pacing Can Make Your Campaigns Look Like They're Working When They're Not

Here's something most advertisers never look at: how Google paces your budget throughout the day.

Google Ads uses a pacing algorithm that distributes your daily budget across the day. If your budget is limited relative to your potential impression volume, Google starts throttling. This means it will start showing your ads selectively based on predicted performance windows.

The problem is that this selective delivery can create the appearance of strong performance metrics (high CTR, decent conversion rate) on the portion of traffic Google does show your ads to, while hiding the massive volume of high-intent queries your ads never appeared for.

This is why a limited budget account often shows better-looking numbers than an account with a larger budget running on the same keywords. You're only seeing the cherry-picked impressions. The moment you scale the budget to match full demand, average performance often drops. Not because something broke, but because you're now reaching the full distribution of searchers, not just the best-predicted slice.

The diagnostic: compare your Impression Share Lost due to Budget against your Search Impression Share. If you're losing more than 30% of potential impressions to budget, your performance metrics are not representative of your true market position.

6. Performance Max Is Not What Google Says It Is

Performance Max is Google's most aggressively pushed campaign type right now. The pitch is full-funnel coverage across all Google surfaces: Search, Shopping, YouTube, Display, Gmail, Maps, with AI handling creative assembly and bid optimization automatically.

What most advertisers don't realize is that PMax almost always includes branded search traffic in its conversion reporting. Your brand keywords, the ones people use to find you after they already know you,  are high-conversion, low-competition, and cheap to win. When PMax captures those conversions, its reported ROAS looks extraordinary. But that ROAS is largely credit for conversions that would have happened anyway.

The way to see this clearly: run a PMax campaign alongside a branded Search campaign and check the search terms PMax is actually bidding on. You'll often find it's competing heavily for your own brand. That's not new customer acquisition. That's cannibalizing organic brand traffic and counting it as paid performance.

My approach with clients: I always run a dedicated branded Search campaign with high priority to isolate brand performance. Then run PMax with brand terms excluded, or at a minimum suppressed, using negative keyword lists. You'll see a more honest picture of what PMax is actually doing for you.

Here is a deeper context on how AI for Google Ads is actually changing the platform versus what the marketing suggests, which covers the full landscape.

7. The Most Important Number in Your Account Is Not CTR, CPC, or Even Conversion Rate

Every agency reports on CTR, CPC, conversion rate, and ROAS. These numbers matter. But none of them tell you whether Google Ads is actually growing your business.

The number that matters — and the one almost no agency tracks properly — is cost per closed deal.

B2B SaaS has CPCs in the $5–$15 range for competitive keywords. CPC increased across 87% of industries in 2025. At those costs, optimizing for form fills without connecting to revenue means you can spend $10,000 a month and have absolutely no idea whether you acquired a single customer.

When FYXER came to me, their Google Ads account showed a healthy cost per conversion on paper. But their CRM told a different story: their conversions weren't matching to closed revenue. The campaigns were generating noise, not customers.

After rebuilding the attribution layer, connecting their CRM, and restructuring the account around quality signals rather than volume signals, Google Ads became responsible for 12% of their total ARR and grew to over 10,000 new customers. The CPCs didn't change dramatically. What changed was what the account was optimizing toward.

Look here for the full step-by-step framework on Google Ads optimization for SaaS, including what to fix first and in what order. This is the most detailed version I've put together.

The Thing No Certification Teaches

Nine years inside Google taught me one thing above everything else: the platform is designed to serve Google's revenue goals alongside advertiser goals. Those two things are often aligned. But sometimes they're not.

The agencies winning for their clients aren't the ones following Google's recommendations to the letter. They're the ones who understand the system well enough to know when to follow it and when to push back.

Knowing the difference between those two situations is worth more than any certification badge. It comes from inside experience, watching how the auction really behaves, how reps are actually incentivized, how Smart Bidding actually learns, and how Performance Max actually reports.

That's what I bring to every account at Scalix AI.

If you want to know whether your current Google Ads setup is working with the platform's mechanics or against them, book a free audit. I'll show you exactly what I'd change if this were my own money on the line.