AI Max campaigns fail in predictable ways. Most performance problems trace back to a small set of root causes: insufficient conversion volume, misaligned bidding targets, poor asset quality, landing page issues, or structural problems with campaign configuration. A systematic diagnostic approach resolves most issues faster than trial-and-error.
Problem 1: Campaign Stuck in Learning Phase
Symptoms:
- "Learning" status persists beyond 4-6 weeks
- Conversion volume is very low (< 5 per week)
- CPA is extremely variable (10x swings between days)
Root causes and fixes:
Insufficient conversion volume:
AI Max needs ~30 conversions in 30 days for effective Smart Bidding. If your campaign generates fewer, the learning phase extends indefinitely.
Fix: Lower the conversion event threshold. If you're optimizing toward "purchase," temporarily switch to "add to cart" or "contact form submitted" as the primary conversion to increase volume, then graduate to higher-intent events once the algorithm has learned.
Target CPA or ROAS set too tight:
A Target CPA of $10 when your actual conversion cost is $45 will keep the campaign in perpetual under-delivery while the algorithm tries to hit an impossible target.
Fix: Switch to "Maximize Conversions" (no target) for 4-6 weeks to accumulate data without budget restriction. Set Target CPA once you have real CPA data.
Budget too low relative to CPC:
If your daily budget is $20 and average CPC is $8, you're getting 2-3 clicks per day — not enough data for the algorithm to learn.
Fix: Increase daily budget to support at least 10-15 clicks per day. For a $8 CPC environment, that's $80-120/day minimum.
Conversion tracking is broken:
If Google Ads isn't receiving conversion signals, the algorithm can't optimize.
Fix: Test your conversion flow end-to-end (complete an actual conversion). Check Google Ads conversion column for "No recent conversions" status. Use Google Tag Assistant to verify tag firing.
Problem 2: CPA Too High
Symptoms:
- Actual CPA consistently 50%+ above target
- Campaign is spending budget but conversions are low-quality or insufficient
- You're within budget limits but ROI is negative
Root causes and fixes:
Wrong primary conversion event:
If you're optimizing toward a low-quality conversion (any contact form), the algorithm maximizes that event including unqualified submissions.
Fix: Switch primary conversion to a more qualified event (demo request with company size, phone call > 2 minutes). Accept lower volume for higher quality.
Target CPA set unrealistically low:
The algorithm will try to achieve your target even if it's below what the market allows, resulting in restricted volume and poor-quality traffic.
Fix: Calculate your maximum acceptable CPA from economics (not a guess or competitor benchmark). Set Target CPA at 15-20% above your absolute maximum.
Poor asset quality:
Low-quality assets attract clicks from less-relevant audiences.
Fix: Check asset performance ratings. Replace all "Low" rated assets. Analyze what "Best" rated assets have in common. Rewrite assets to be more specific about who your offering is for.
Landing page conversion rate problem:
High CPC × Low conversion rate = High CPA. The algorithm can't fix a 0.5% landing page conversion rate — that's a landing page problem.
Fix: Measure landing page conversion rate directly (Google Analytics or your analytics platform). If conversion rate is below 2-3% for most verticals, optimize the landing page before adjusting campaign settings.
Audience signal mismatch:
If your audience signals are attracting users who are interested in your topic but not ready to convert, CPA will be high.
Fix: Review AI Max Insights Report > Audience Insights. If high-spending audience segments have low conversion rates, refine your audience signals to target more purchase-ready segments.
The DSA-to-AI-Max migration (https://yositeup.com/blog/google-ads-dsa-ai-max-migration-february-2027) sometimes creates CPA spikes because the new campaign has no historical optimization data. Expect 4-6 weeks of elevated CPA after migration while the algorithm relearns.
Problem 3: Low Conversion Volume (Campaign Spends but Doesn't Convert)
Symptoms:
- Campaign spends budget daily
- Click volume is reasonable
- But conversion count is near zero
Root causes and fixes:
Conversion tracking is broken (most common):
Spending without converting is the primary symptom of broken conversion tracking. The algorithm sees clicks but no conversions and keeps bidding.
Fix: Verify conversion tracking fires on your confirmation/thank-you page. Use Google Tag Assistant Real-Time mode while completing a test conversion. Check Google Ads Conversions report for "No recent conversions."
Landing page disconnect:
Traffic is reaching a page that doesn't match ad intent. Users leave immediately without converting.
Fix: Check bounce rate and time-on-page for AI Max landing pages in GA4. If bounce rate is > 80% and time-on-page is < 30 seconds, users are arriving at the wrong page or the page isn't loading correctly.
Wrong audience:
AI Max is finding users who click but don't have purchase intent.
Fix: Review search category data in AI Max Insights. If the top categories are informational (research, comparison) rather than commercial (pricing, buy, quote), add negative keywords for pure research intent queries.
Form or checkout is broken:
Users try to convert but can't complete the process.
Fix: Test your conversion form or checkout flow yourself. Check for JavaScript errors in browser console on the landing page. Test on mobile specifically.
Problem 4: Campaign Runs Out of Budget Too Quickly
Symptoms:
- Budget exhausted by 10 AM
- CPA is reasonable but volume limited by budget
- ROAS is good but revenue potential is capped
Root causes and fixes:
Budget is too low for the opportunity:
This is often a positive problem — the campaign is efficient and could profitably scale.
Fix: Increase budget. Calculate maximum acceptable spend at current CPA/ROAS and scale toward that level.
Overly broad targeting without adequate budget:
If you're targeting nationally with a $50/day budget, you'll exhaust it on a small fraction of available impressions.
Fix: Either increase budget or tighten geographic targeting to concentrate spend in your highest-performing areas.
Bidding strategy is too aggressive:
Target CPA set very low forces the algorithm to bid on high-competition (higher CPC) opportunities to hit volume goals.
Fix: Slightly increase Target CPA to allow the algorithm to find more volume at slightly higher costs.
Problem 5: Sudden Performance Drop
Symptoms:
- Previously stable campaign suddenly underperforms
- CPA spikes or conversion volume drops without obvious cause
Root causes:
Conversion tracking broken by website update: A site update can break Google Ads tags.
Competitor behavior changed: New competitive entrant or competitor budget increase.
Seasonal/market shift: Demand drop, external news events.
Campaign learning reset: A significant campaign change triggered a new learning phase.
July 2026 ToS policy change impact: Assets or landing pages may have been flagged under updated policies. Check Google Ads for policy notifications. (https://yositeup.com/blog/google-ads-tos-july-2026-ai-automation-what-changed)
Diagnostic sequence:
- Check conversion tracking status first (most common cause)
- Check for policy notifications or disapprovals
- Check if recent campaign changes were made (even small ones can trigger learning)
- Check external factors (seasonality, news events, competitor activity)
- Check June 2026 reporting gap context (https://yositeup.com/blog/google-ads-reporting-data-deleted-june-2026) — if you're comparing to a period affected by the data deletion, your benchmark may be wrong
The systematic diagnostic approach — start with tracking, then targeting, then assets, then external factors — resolves most AI Max performance issues within one to two optimization cycles.
Top comments (0)