Google's auto-apply recommendations promise easier ad management, but they can raise your costs or shift your strategy without your say-so. Here's what to lock down first.
Google Ads changes fast. Every quarter brings new settings, new bid strategies, new automation features. And with each wave of change comes the Recommendations tab, nudging you to adopt the latest Google-blessed practice.
The problem: not all recommendations are equal. Some genuinely improve performance. Others increase your costs or shift your strategy in ways that don't serve your business. And if you've enabled auto-apply, Google is rolling out some of these changes without your approval.
According to Search Engine Land, there are four recommendations that deserve manual review before you let them run. The source doesn't itemize all four in the summary, but the pattern is clear: if a recommendation touches bidding, budget, or creative scope, test it manually first.
Why? Google's algorithm optimizes for Google's revenue (more clicks, more impressions, more competition). Your algorithm should optimize for your revenue (conversions, profit margin, customer lifetime value). Those two don't always align.
Search Engine Land's source notes that trial and error is part of the job in Google Ads. That's honest. But trial and error works only if you control the variables. Auto-apply stacks multiple changes at once, which means you can't tell which recommendation actually moved the needle, or which one burned the budget.
Google controls the algorithm, the targeting options, and the recommendation engine. You control the budget, the strategy, and the decision to act. Keep it that way. Auto-apply surrenders that control in exchange for convenience, and convenience is expensive when Google's incentives don't match yours.
Audit your account today. If auto-apply is on, turn it off. Then run through your pending recommendations one at a time. You'll spend a few hours now to avoid wasting thousands in misaligned spend later.
Auto-apply lets Google automatically implement recommendations without you approving them first. The risk: Google's algorithm prioritizes clicks and impressions, not your profit margin, so an 'optimization' can raise costs or change your bidding strategy in ways that hurt your bottom line.
Per Search Engine Land, four key recommendations warrant manual review before any automation: settings that affect bidding strategy, budget allocation, and broad creative or keyword changes. The source recommends testing each one independently for at least two weeks rather than stacking them.
Measure by conversion value and cost per conversion, not clicks or impressions. Google's default view emphasizes volume metrics; you must flip to your own conversion tracking to see if a recommendation is actually making you money.
Turn off auto-apply, create a test period of 2, 4 weeks per recommendation, isolate one change at a time so you can measure its real impact, then decide manually whether to keep it. Trial and error is part of the job, but you control the variables.