Which Google Ads AI Features Should You Turn Off?

Is Google’s AI Optimizing for Your Goals or Google’s Goals?

Google’s AI is fast and data-rich, but it’s optimizing for Google’s metrics — not yours. If you let it run unsupervised, your budget will disappear without much to show for it. Dan Butler, who has managed over $2M in annual ad spend across fifteen years, draws a clear distinction: Google’s incentive is to spend your budget. Your incentive is to generate qualified pipeline or revenue. Every Google AI feature should be evaluated through that lens. The goal isn’t to refuse the AI — it’s to keep it pointed at your goals instead of Google’s.

Which Google Ads AI Features Should You Keep On?

Smart Bidding with target CPA or target ROAS is worth keeping on when you have at least 30 conversions per month and your conversion tracking accurately reflects real business value. Dan’s framework for which features earn their place: if the feature has enough data to optimize against, if it’s optimizing toward a metric you actually care about, and if you can set guardrails that prevent runaway spend. Smart Bidding meets these criteria for accounts with sufficient conversion volume. Responsive Search Ads also earn their spot because Google’s ability to test headline and description combinations at scale genuinely outperforms manual A/B testing.

Which Google Ads AI Features Should You Turn Off?

Auto-applied recommendations, broad match keywords without negative keyword lists, and asset auto-creation should be turned off in most accounts. Dan is direct about what quietly burns budget: auto-applied recommendations let Google make changes to your account without your approval, often in ways that increase spend without improving results. Broad match without robust negative keyword lists matches your ads to queries that have nothing to do with your business. Asset auto-creation generates ad creative that’s generic and off-brand. Each of these features is designed to be helpful, but the default settings prioritize Google’s revenue over your performance.

How Should You Use Performance Max?

Use Performance Max only if you constrain it with audience signals, negative keywords, and budget caps — and monitor its search term reports weekly. Dan’s view on Performance Max is nuanced: it can work well for e-commerce with product feeds, where Google’s AI has clear product data to work with. For lead generation, it’s more dangerous because Google will optimize for the cheapest conversions, which are rarely the most qualified. The key constraint is never letting Performance Max run without audience signals that tell it who you’re actually trying to reach. Without those signals, it will find the cheapest clicks, not the best customers.

What Should a Weekly Google Ads Review Look Like?

Every Monday, check search term reports for irrelevant queries, review auto-applied changes Google has made, verify that conversion tracking is still firing correctly, and compare actual CPA to target CPA by campaign. Dan’s weekly cadence is designed to catch the ways Google’s AI drifts away from your goals. The search term report reveals whether your keywords are matching to relevant queries or burning spend on junk. The auto-applied changes log shows whether Google has made unauthorized adjustments. Conversion tracking verification catches pixel breaks before they corrupt a full week of data. The CPA comparison by campaign identifies which campaigns need intervention versus which are performing within tolerance.