Everyone's talking about what to do in AI search. Let's talk about what not to do.
These are the mistakes I'm seeing repeatedly — some are old habits that no longer work, others are new traps specific to AI-era advertising. Each one is costing you money.
Mistake 1: Blindly Trusting AI Max
Google claims AI Max delivers 14-27% more conversions at similar efficiency. Independent testing tells a different story.
Smarter Ecommerce analyzed 250+ retail campaigns and found AI Max delivers conversions at approximately 35% lower ROAS compared to traditional targeting methods. Separate testing showed AI Max cost per conversion at $100.37 vs. $43.97 for phrase match — 90% higher.
What's happening: AI Max claims conversions from your existing keywords. It uses "inferred intent" matching — when a user types a partial query and clicks autocomplete, AI Max can take credit even if your exact match keyword would have caught it anyway.
The fix: Don't just look at AI Max's reported numbers. Pull search term reports. Cross-reference against your existing keyword coverage. Calculate what's actually incremental vs. cannibalized.
Mistake 2: Ignoring Search Term Reports
With broad match, AI Max, and Performance Max expanding what queries you show for, search term reports are more important than ever. Yet most advertisers check them monthly at best.
What's happening: The algorithms are matching your ads to queries you never intended. Competitor terms, irrelevant categories, queries in different languages, low-intent informational searches — all eating your budget.
The fix: Weekly search term audits minimum. Build robust negative keyword lists. Use n-gram analysis to find patterns. One analysis found 99% of AI Max impressions generated zero conversions across 30,000 search terms.
Mistake 3: Over-Segmenting Performance Max
The instinct to segment makes sense — different products, different audiences, different creative. But with Performance Max, it backfires.
What's happening: Each PMax campaign needs approximately 30 conversions monthly just to function, closer to 300 for genuine optimization. When you split campaigns excessively, each segment struggles to accumulate enough data.
The fix: Consolidate where possible. If you have separate PMax campaigns for t-shirts, jeans, jackets, and accessories, consider combining into broader asset groups within fewer campaigns. Feed the algorithm more data.
Mistake 4: Chasing Ad Strength Scores
Google pushes Ad Strength as a key metric. Many advertisers optimize for "Excellent" scores, believing it improves performance.
The problem: Ad Strength measures Google's ability to control your messaging, not your ad's actual performance. Research consistently shows lower Ad Strength ads often have higher conversion rates than high Ad Strength ads.
The fix: Ignore Ad Strength. Focus on actual conversion metrics. If a "Poor" Ad Strength ad converts better than an "Excellent" one, run the poor one. Your job is results, not Google's approval.
Mistake 5: Auto-Applying Google's Recommendations
The problem: Google's recommendations often prioritize spend increases over profitability. Suggestions like broadening match types, enabling audience expansion, or adding Search Partners frequently hurt ROI.
The fix: Never auto-apply. Review every recommendation manually. Ask: does this serve my goals or Google's? Most recommendations should be dismissed.
Mistake 6: Measuring the Wrong Things
If you're still optimizing purely for CTR and CPA, you're optimizing for a shrinking slice of reality. With 60% of searches ending without a click, traditional click-based metrics capture less and less.
What to add:
- Brand search lift (does branded search volume increase when you're active?)
- Impression share in key positions
- Conversion rate (should be up if CTR is down)
- AI visibility metrics (citation rate, share of voice in AI responses)
More Critical Mistakes
7. Treating Paid and Organic as Separate
Being cited in AI Overviews improves paid CTR by 91%. Your organic presence directly affects your paid performance. Joint strategy sessions are essential.
8. Neglecting Creative in Favor of Targeting
The AI handles most targeting now. Many advertisers spend 80% of time on targeting, 20% on creative. In the automation era, that ratio should flip.
9. Using Max Bidding with Broad Match
Max Conversions + Broad Match tells Google to spend aggressively on whatever vaguely related query it can find. Use target bidding (tCPA, tROAS) with broad match instead.
10. Waiting for Perfect Measurement Before Acting
AI search measurement is imperfect and will be for years. But waiting means competitors build presence while you optimize for yesterday's reality.
The Common Thread
Most of these mistakes share a root cause: treating AI-era PPC like it's 2019.
The new reality:
- • Algorithms handle tactics; humans handle strategy
- • Measurement is messier; directional decisions still matter
- • Creative differentiation is the remaining human edge
- • Paid and organic are deeply intertwined
- • Visibility sometimes matters more than clicks
Every mistake on this list comes from applying old thinking to new realities. Find one mistake you're making. Fix it this month. Then move to the next.
