Why Your Google Ads AI Recommendations Are Wrong

Every Google Ads manager knows the feeling. You log in and there it is: the recommendations tab glowing with that optimization score, silently judging you for not accepting every suggestion.

But here’s what Google won’t tell you: their AI recommendations optimize for Google’s revenue, not yours.

The Misaligned Incentive Problem

Google makes money when you spend money. Their recommendations are trained on one primary objective: increasing ad spend across their platform. Sometimes that aligns with your goals. Often, it doesn’t.

Consider the most common recommendations:

“Raise your budget” — Of course they want you to spend more
“Add broad match keywords” — Expands reach but often tanks conversion rates
“Enable auto-applied recommendations” — Lets Google make changes without your approval

What to Actually Do

Not all recommendations are bad. Here’s how I evaluate them:

Accept These (Usually)

Fixing disapproved ads
Adding obvious negative keywords
Removing redundant keywords

Scrutinize These

Bid strategy changes (test in a campaign experiment first)
Audience expansion suggestions
New keyword ideas (check search intent carefully)

Almost Always Reject

Broad match keyword additions without testing
Budget increases without performance justification
Auto-apply settings

The Real Framework

Before accepting any recommendation, ask yourself:

Does this expand reach or improve efficiency? Google loves reach. You should love efficiency.
Can I measure the impact? If you can’t isolate the variable, don’t change it.
What’s Google’s incentive here? Follow the money.

Your optimization score is vanity. Your ROAS is sanity.