Does Google Smart Bidding optimize for Google revenue instead of mine?

Question 1

Does Google's AI work for my business or for Google's revenue?

Accepted Answer

According to the article, Google's advertising AI serves a platform whose primary obligation is to its shareholders, not to advertiser P&L. The article states trusting the algorithm without strategic oversight is like asking a landlord to also set a rent budget. The question is not whether the AI is powerful, but whether its power is pointed in the advertiser's direction.

Question 2

How can Smart Bidding hit my Target CPA but still overpay for conversions?

Accepted Answer

The article states systems like Target CPA, Target ROAS, and Maximize Conversions are designed to hit an average target over time, not to optimize every individual auction. When an advertiser sets a Target CPA of $50, the system does not refuse auctions where the projected CPA exceeds $50 and bids on cheaper and expensive auctions alike as long as the blended average lands near target. As a result, the article states the algorithm routinely overpays for individual conversions when it has budget from cheaper wins earlier in the period.

Question 3

Why can't I see where my Performance Max budget actually goes?

Accepted Answer

According to the article, Performance Max consolidates Search, Display, YouTube, Discover, Gmail, and Maps into a single campaign and does not show meaningful performance breakdowns by channel. The article states that without that visibility the advertiser cannot see if budget is going to lower-intent inventory and has no lever to reallocate spend to Search where ROAS is strongest. The article also states Performance Max absorbs branded search traffic that would have converted anyway and attributes those conversions to itself, overstating its incremental value.

Question 4

How does Broad Match expansion increase my Google Ads costs?

Accepted Answer

The article states Google pushes Broad Match on the premise that Smart Bidding can bid low on irrelevant queries and high on relevant ones, but in practice expansion floods accounts with tangentially related queries that do not convert at rates that justify cost. The article states more queries mean more auctions and more revenue, while control requires complex negative keyword strategies that most advertisers do not implement rigorously. In 2020, Google restricted search terms reporting to only show “significant” queries, so according to the article advertisers lack the full data needed to control the expansion Google recommends.

Question 5

Why does my Optimization Score go up when I spend more, not when ROI improves?

Accepted Answer

The article describes Optimization Score as presented as a health check where implementing suggestions raises the score. It states the most heavily weighted suggestions almost always involve spending more money: adding new keywords, raising budgets, adopting automated bidding, enabling ad extensions, expanding to new campaign types. Recommendations that would reduce spend, like pausing underperforming keywords or tightening geographic targeting, are rarely featured and carry minimal score impact when they appear.

Question 6

What happens to my budget during Smart Bidding's learning phase?

Accepted Answer

According to the article, every significant change to a Smart Bidding campaign triggers a learning phase that typically lasts one to two weeks, during which performance is volatile. Google explicitly warns against making further changes and recommends maintaining or increasing budgets during learning. The article states advertisers therefore pay full rates for suboptimal performance while the algorithm recalibrates, and attempting to course-correct triggers another learning phase.