Google AI Max for Search: Complete Setup Guide 2025
Set up Google AI Max for Search: what Google's 14% and 27% conversion figures actually mean, setup steps, case studies, and optimization tips.
Typical conversion uplift Google cites at similar CPA/ROAS (non-Retail)
Google’s figure for campaigns with over 70% exact/phrase conversions (non-Retail)
Announced May 6; global beta from late May
Features: search term matching, text customization, Final URL expansion
Key Takeaways
When Google announced AI Max for Search at Google Marketing Live in May 2025, the headline claim was that advertisers who activate it will typically see 14% more conversions or conversion value at a similar CPA/ROAS, with 27% for campaigns still mostly using exact and phrase match keywords. Both figures are Google internal data for non-Retail advertisers. Case studies showcased conversion increases of 263% and a 73% lower cost per acquisition, but those are selected examples, not typical results.
This isn’t a story of AI Max being worthless or of Google making false promises. It’s a story about the gap between selected success stories and what a typical advertiser should expect. This guide separates what Google has actually claimed, and under which conditions, from what you can reasonably infer for your own account, so you can decide whether AI Max belongs in your advertising strategy.
AI Max for Search is a suite of AI-powered enhancements for Google Search campaigns that uses keywordless matching, generative ad copy, and dynamic landing page selection to expand your campaign reach beyond traditional keywords. When it works, it discovers high-converting queries you'd never think to target manually. When it doesn't, it burns budget on irrelevant clicks while providing limited transparency into why. Success depends heavily on your industry vertical, campaign structure, budget size, and willingness to actively manage an AI system that requires human oversight.
Critical Context: Google’s 14% and 27% figures are separate, self-reported, and exclude Retail advertisers. Test AI Max with a campaign experiment before relying on either number.
Corrected November 5, 2026: Google’s 14% figure is a typical result for non-Retail advertisers, and the 27% figure applies only to campaigns with more than 70% of conversions or conversion value from exact or phrase match keywords; they are not a 14–27% range. The global beta began in late May 2025. We removed the $50/day minimum, the 30-conversion requirement, the $750/day budget rule, and community poll and agency figures that we could not trace to Google or a published source.
What is Google AI Max for Search?
AI Max for Search is not a new campaign type—it’s an optional suite of AI-powered enhancements that can be enabled within new or existing Google Search campaigns. Google announced it on May 6, 2025 on the Google Ads & Commerce blog and featured it at Google Marketing Live on May 21, 2025, with a global beta rolling out from late May 2025. At announcement, Google said support in Google Ads Editor and the API was expected later in 2025.
The core premise is straightforward: AI Max applies Google's latest artificial intelligence technology to three key campaign elements—search term targeting, ad copy creation, and landing page selection—to find conversions your traditional keyword-based campaigns miss. Unlike Performance Max, which operates as a black box across all Google properties, AI Max maintains the transparency and control of Search campaigns while adding AI-powered expansion capabilities on top of your existing keyword structure.
The Three Core Features
- Search Term Matching: Expands beyond your keywords using broad match enhancement and keywordless technology to match queries you haven't explicitly targeted, similar to Dynamic Search Ads but more sophisticated.
- Text Customization: Uses generative AI to create ad headlines and descriptions dynamically based on your landing pages, existing ads, and the specific search query, optimizing messaging in real-time.
- Final URL Expansion: Analyzes your entire website and dynamically sends users to the most relevant landing page based on their search query, rather than using only your manually specified URLs.
Each feature can be toggled on or off independently at the campaign or ad group level, giving you granular control over how aggressively AI Max expands your campaigns. You can start conservatively with just search term matching, or enable all three features together for maximum AI-powered expansion. Google’s Klook case study describes using the full suite of features, while a staged rollout lets you isolate which feature drives results.
Google says the AI Max features need conversion-based Smart Bidding (such as Max Conversions, Max Conversion Value or target CPA/ROAS) to function effectively, and that search term matching doesn’t work with Manual CPC bidding. Google also says AI Max won’t be effective if campaigns are limited by budget. We found no minimum budget or minimum conversion count in Google’s AI Max documentation, so size the budget from your own target CPA and goals. This is one reason AI Max isn’t suitable for every advertiser.
AI Max vs Performance Max: Complete Comparison
The most common misconception about AI Max is that it replaces Performance Max. It doesn't. These are complementary tools serving different strategic purposes, and Google markets them together in its “Power Pack” (AI Max, Performance Max and Demand Gen). AI Max enhances Search campaigns specifically, while Performance Max operates as a standalone multichannel campaign type across all Google inventory. Understanding when to use each—or both—is critical for effective Google Ads strategy.
| Aspect | AI Max for Search | Performance Max |
|---|---|---|
| Campaign Type | Enhancement to Search campaigns | Standalone campaign type |
| Inventory | Google Search only | All Google properties (Search, YouTube, Display, Gmail, Maps, Discover) |
| Transparency | Full search terms reporting with source attribution | Limited transparency ("black box") |
| Control | Keyword-level control maintained, campaign/ad group negatives | Campaign-level negative keywords only, limited keyword control |
| Targeting | Keyword-based + AI expansion | Goal-based, fully automated |
| Best For | Search transparency with AI performance | Maximum reach across all channels |
| Setup | Add to existing Search campaigns | New campaign structure required |
- You want to maintain Search campaign structure and transparency
- Keyword-level control is important for your business
- You need detailed search terms reporting for optimization
- You're primarily focused on Search inventory rather than multichannel
- Your campaigns heavily use exact and phrase match keywords (Google’s 27% figure applies)
- You're in a regulated industry requiring tight messaging control
- You want maximum reach across all Google channels simultaneously
- You're willing to sacrifice transparency for automation and scale
- You have strong creative assets for display, video, and discovery formats
- Your goal is multichannel brand awareness and conversions
- You need to scale quickly without manual campaign management
The Portfolio Approach (Recommended)
Rather than choosing one or the other, Digital Applied suggests most advertisers consider a portfolio: traditional Search campaigns for core brand and high-intent terms, AI Max to expand and discover, and Performance Max for multichannel reach and remarketing. We have no sourced benchmark for the right budget split, so set it from your own experiment results. Google has promoted AI Max, Performance Max and Demand Gen together as its “Power Pack” for performance across Search and YouTube; that is a product bundle, not a budget formula.
Key Features & Capabilities
AI Max for Search consists of three main features that work independently or together to expand your campaign reach beyond traditional keyword targeting. Each feature addresses a specific limitation of manual Search campaigns, and each comes with its own set of benefits and risks that require careful management.
What It Does: Expands beyond your existing keywords using both broad match enhancement and keywordless matching technology to find relevant search queries you haven't explicitly targeted. Think of it as Dynamic Search Ads combined with more sophisticated broad match—it analyzes your landing pages, ad assets, and campaign context to predict which searches are likely to convert, then matches your ads to those queries even without specific keyword triggers.
The Reality: Google presents this as intelligent expansion, but the queries it matches can include terms you wouldn’t have chosen. The search terms report displays “AI Max” as a match type for these incremental queries, with a source column showing whether a term came from broad match expansion or keywordless matching, so review them regularly for relevance.
What It Does: Uses generative AI to dynamically create and optimize ad copy including headlines, descriptions, and calls-to-action based on your website content, existing ad assets, and the specific search query. Rather than serving static responsive search ads, the AI assembles customized ad combinations in real-time to match user intent more precisely.
The Reality: Text customization can generate relevant ad copy, but it can also draw on content from unexpected places on your site. Because you don’t write each line, generated copy could include claims or phrasing that conflict with industry regulations, a particular concern for pharmaceutical, healthcare, and financial services advertisers.
What It Does: Analyzes your entire website and dynamically sends users to the most relevant landing page based on their specific search query, rather than using only the manually specified final URL. The AI evaluates page content across your site to determine which page best matches the searcher's intent, then generates headlines that align with that dynamically selected landing page.
The Reality: Final URL expansion only works if text customization is also enabled (to ensure ad copy matches the selected landing page), and it requires a comprehensive, well-organized website to function properly. The risk is sending traffic to wrong or low-converting pages—about pages, blog posts, outdated content, or pages not designed for conversion. One caution: if your website isn’t a strong representation of your business, consider leaving text customization and Final URL expansion off.
Campaign-Level Negative Keywords
One of AI Max's most significant advantages over Performance Max is full support for negative keywords at both campaign and ad group levels. This gives you the ability to proactively block irrelevant search terms, competitor brands, and low-quality queries before they waste budget. Comprehensive negative keyword lists before launch are essential—don't expect AI to understand what you don't want to target. Set exclusions first, then let AI explore within those boundaries.
Smart Bidding Exploration (Related Feature)
Announced alongside AI Max at Google Marketing Live 2025, Smart Bidding Exploration was described by Google as its biggest update to bidding in over a decade. It lets Target ROAS bidding accept some flexibility on the ROAS target so it can bid on less obvious queries, including more general or early-stage intent. Google reported, on average, an 18% increase in unique search query categories with conversions and a 19% increase in conversions (Google internal data). While AI Max handles targeting and creative expansion, Smart Bidding Exploration handles bidding on the newly reachable queries. The two features work together but can be used independently.
Complete Setup Guide: Step-by-Step
Setting up AI Max correctly from the start significantly impacts your results. The difference between successful implementations and failed tests often comes down to proper prerequisites, strategic feature selection, and choosing the right testing methodology. Here's how to set up AI Max for the best chance of success.
Prerequisites Checklist
- Smart Bidding Active: Conversion-based Smart Bidding (Max Conversions, Max Conversion Value, or target CPA/ROAS); Google says the features need it, and search term matching doesn’t work with Manual CPC
- Conversion Tracking: Reliable conversion tracking so Smart Bidding has a clear signal to optimize against
- Budget Adequacy: Campaigns not limited by budget (Google says AI Max won’t be effective otherwise)
- Negative Keywords: Comprehensive negative keyword lists in place before enabling
- Website Quality: Well-organized site with comprehensive content for AI analysis
- Enhanced Conversions: Enabled for better conversion tracking and optimization
Setup Option 1: New AI Max Campaign
- Create New Campaign: In Google Ads, click "+" → "New campaign" → Select "Search" as campaign type → Choose conversion goals
- Configure Bidding: Select Max Conversions or Max Conversion Value → Optionally set target CPA or ROAS → Set daily budget sized to your target CPA and conversion goals
- AI Max Opt-In: When you reach the “AI Max” section during creation, toggle it on → all three features are enabled by default. Update, November 2026: Google’s Help Center now says new Search campaigns have AI Max turned on by default, so check that the setting matches your intent.
- Configure Features: Choose which features to enable (conservative: search term matching only; aggressive: all three)
- Set Exclusions: Add brand exclusions, URL exclusions, and campaign-level negative keywords
- Complete Setup: Add keywords (will work alongside AI), create RSA ads, configure targeting
- Ad Group Settings: Toggle search term matching per ad group, add URL inclusions, set locations of interest if needed
Setup Option 2: Enable on Existing Campaign
- Verify Eligibility: Confirm the campaign uses conversion-based Smart Bidding, which Google says the features need
- Access AI Max Settings: Within campaign settings, locate "AI Max" section → Toggle to enable
- Configure Campaign-Level: Choose features to enable → Set URL exclusions and brand settings
- Configure Ad Group-Level: Toggle search term matching per ad group → Add URL inclusions
- Monitor Closely: Watch for performance changes during the learning period
Setup Option 3: Test with Experiments (Recommended)
This is the safest approach for testing AI Max without risking your existing campaign performance. Digital Applied recommends this method for all first-time AI Max implementations.
- Create Experiment: Navigate to "Experiments" in Google Ads → Select campaign to test → Choose "Create custom experiment"
- Set Parameters: Base Campaign = existing Search (control) → Experiment Campaign = duplicate with AI Max (treatment) → 50/50 traffic split
- Enable AI Max: In experiment campaign only, enable AI Max settings → Choose features to test
- Match Budgets: Apply same daily budget to both campaigns for fair comparison
- Run Several Weeks: Allow for the learning period, then evaluate over a further stretch at stabilized performance
- Evaluate Results: Compare conversions, CPA/ROAS, conversion rate, search terms quality
- Decision Point: If experiment wins, apply AI Max to base campaign → If underperforms, disable and optimize approach
Feature Selection Strategy
Conservative Approach: Start with search term matching only, monitor for several weeks, then add text customization if performing well, finally add final URL expansion only if website is comprehensive and well-organized. This staged rollout minimizes risk and isolates which features drive results.
Aggressive Approach: Enable all three features together from day one for maximum AI-powered expansion. Google’s Klook case study describes using the full suite of features, but this approach also carries the highest risk of budget waste if relevance issues occur. Only recommended for well-optimized accounts with adequate budgets and comprehensive websites.
Selective Approach: Enable different features at different ad group levels based on performance and risk tolerance. For example, enable all features for non-brand generic ad groups, but only search term matching for product-specific ad groups, and disable AI Max entirely for brand defense ad groups.
Performance: What Google Claims and What It Means
This section sets out what Google has claimed, with the conditions Google attaches, and what that does and doesn’t tell you about your own account. Google’s case studies are real, but they are selected examples.
Google's Official Claims (May 2025)
The Promise:
- 14% more conversions or conversion value, typically, at a similar CPA/ROAS (Google internal data, 2025; non-Retail advertisers)
- 27% for campaigns still mostly using exact and phrase match, defined by Google as more than 70% of conversions or conversion value from those match types (non-Retail)
- Important exclusion: both figures are for non-Retail advertisers
Google's Featured Case Studies
- 263% surge in conversions
- 73% reduction in CPA
- Best-case scenario results
- 161% increase in conversion value
- 25% increase in ROAS
- Within one month
- 2X higher conversion rate
- 31% lower cost-per-conversion
- Net-new query conversions
- 20% incremental conversion lift
- 16% higher incremental ROAS
- 22% lower CPA
These are Google-published case studies, selected by Google, and not typical results. Royal Canin’s result came from an A/B experiment at the same spend; Klook’s came within one month.
What Independent Data Exists
An earlier version of this section quoted a community poll, an agency sample of search terms and an anonymous advertiser. We could not trace those figures to a published method, so they have been removed. Google’s numbers are self-reported vendor data; the dependable evidence for your account is a campaign experiment.
Why Results Vary
- Selected Case Studies: Google’s case studies are chosen examples, so they show what is possible rather than what is typical.
- Retail Is Excluded: Google’s 14% and 27% figures are both for non-Retail advertisers, so they say little about retail accounts.
- Account Maturity May Matter: Accounts that already cover their search demand well may have fewer untapped queries for AI Max to find, while accounts relying on exact and phrase match have more. Google’s higher 27% figure is for the latter group.
- The Features Interact: Final URL expansion requires text customization, pinned assets aren’t respected with Final URL expansion, and Google says the features need conversion-based Smart Bidding to work effectively.
- Budget Limits Hold It Back: Google says AI Max won’t be effective if campaigns are limited by budget.
- The Method Isn’t Published: The announcement and Help Center cite only “Google internal data, 2025” for the 14% figure.
The Critical Retail Exclusion
Google states that both headline figures are for non-Retail advertisers. The footnote to the May 2025 announcement attaches that qualifier to the 27% sentence, and the Help Center repeats it for the 14% figure. Neither source explains why, so we won’t guess. The practical consequence is that neither figure is a benchmark for retail advertisers: test AI Max with an experiment rather than assuming 14% applies.
Success Factors & Best Practices
Google’s case studies and Help Center guidance point to several factors that make AI Max more likely to deliver. These aren’t guarantees, and Google’s figures are self-reported, but they improve the odds of a useful test.
Campaign Prerequisites
- Smart Bidding Already Working: AI Max amplifies Smart Bidding performance. If your Target CPA or Target ROAS isn't stable and delivering results, fix that before adding AI Max complexity.
- Quality Conversion Tracking: Google says the features need conversion-based Smart Bidding to work effectively, so reliable conversion tracking matters. Enhanced Conversions and Data-Driven Attribution are worth having in place.
- Comprehensive Negative Keywords: Don't expect AI to understand what you don't want. Build extensive negative keyword lists before enabling AI Max, including competitor brands, low-intent queries, and irrelevant variations.
- Budget Headroom: Google says AI Max won’t be effective if campaigns are limited by budget. We found no minimum budget in Google’s AI Max documentation, so size the budget from your target CPA and conversion goals.
Account Structure That Excels
Exact and Phrase Match Heavy Campaigns: This is where Google’s 27% figure applies: campaigns with more than 70% of conversions or conversion value from exact or phrase match keywords (non-Retail), versus a typical 14% overall. Such campaigns have the most untapped search volume for AI Max to discover. If you already rely heavily on broad match, AI Max may add less because you already capture much of that expansion.
Strong Existing Coverage: Campaigns already showing for most searches on their current keywords can’t grow without expanding targeting. AI Max can find incremental volume that traditional keywords miss, which suits accounts that have fully optimized their existing keyword strategy.
Website Content Quality
AI Max needs comprehensive, well-organized landing pages to function properly. Text customization draws on your site content and existing assets, so thin product descriptions or sparse content give the AI less material to work with. Final URL expansion needs multiple relevant landing pages for different offerings.
Warning: If your website isn’t a strong representation of your business, consider leaving text customization and Final URL expansion off, and use URL exclusions to keep unsuitable pages out. Poor website quality limits AI Max’s effectiveness and increases the risk of sending traffic to the wrong pages.
Best-Performing Industries
Clear conversion events, product feeds provide rich data, revenue tracking feeds optimization. Royal Canin achieved 263% conversion surge, L'Oréal saw 2X conversion rate improvement.
AI Max matches conversational search patterns characteristic of business software research. ClickUp achieved 20% conversion lift with 16% higher ROAS.
Long-tail conversational queries dominate travel searches. Locations of Interest targeting particularly valuable. Klook saw 161% conversion value increase within one month.
Lead generation when properly configured with offline conversion tracking. MyConnect achieved 16% more leads at 13% lower CPA.
Active Management Requirements
AI Max is not “set and forget.” Plan frequent search terms review in the first weeks, then a regular cadence. Watch for competitor brand queries, search term relevance issues and weak Search Partner performance. Add negative keywords promptly, review which headlines and landing pages are serving, and be prepared to pause or narrow features that spend without delivering conversions.
When to Use AI Max (and When to Avoid)
Based on Google’s published guidance and case studies, here’s when to test AI Max and when to wait.
Test AI Max If You Have:
Avoid AI Max If You Have:
Decision Framework
Ask yourself: (1) Do I have exact/phrase match-heavy campaigns with room to expand? (2) Am I using conversion-based Smart Bidding, with a budget that doesn’t limit delivery? (3) Do I have reliable conversion tracking? (4) Is my website comprehensive and well-organized? (5) Am I willing to monitor and optimize through a multi-week experiment? If you answer yes to all five, test AI Max with an experiment. If you answer no to several, fix those prerequisites first.
Optimization Strategies for Better Results
AI Max requires active optimization to succeed. Unlike traditional Search campaigns where you optimize keywords and bids, AI Max optimization focuses on guiding the AI's exploration through strategic exclusions and signal reinforcement.
Early Weeks: Frequent Monitoring
Search Terms Review (Daily): Download the search terms report filtered to "AI Max" match type. Look for three red flags: (1) Competitor brand terms consuming budget, (2) Queries that completely miss user intent, and (3) High-impression, zero-conversion terms. Add negative keywords immediately—don't wait for statistical significance.
RSA Headline Analysis: The search terms report shows which headlines served with each query. Identify headlines generating conversions versus those burning budget. While you can't remove AI-generated headlines directly, you can disable text customization if generic AI copy outperforms your manual assets (or vice versa).
Landing Page Performance: If using Final URL expansion, check which pages AI is selecting. Look for patterns like "AI keeps sending traffic to blog posts instead of product pages" or "About page getting Search traffic." Add URL exclusions for pages that shouldn't receive paid traffic, and URL inclusions at ad group level for pages that should be prioritized.
Evaluation After the Learning Period
Performance Comparison: If running experiments, compare control vs treatment once the learning period has passed. Look beyond aggregate metrics—segment by match type to see if AI Max traffic performs differently than traditional keyword traffic. Check conversion rates, not just conversion volume (AI Max can deliver more conversions at worse efficiency).
Budget Pacing: Is the campaign spending its full budget? Budget-limited campaigns prevent AI from exploring new opportunities. Consider increasing budget if seeing positive signals, or decreasing if burning through spend without conversions.
Search Partner Network Audit: Segment performance by network (Google Search vs Search Partners). If Search Partners show clearly weaker ROAS or CPA, exclude them from the campaign.
Ongoing Optimization (Weekly)
Negative Keyword Expansion: AI Max requires more aggressive negative keyword management than traditional Search. Build negative lists around competitor brands, low-intent queries, informational searches that don't convert, and any search patterns burning budget. Unlike Performance Max, AI Max respects both campaign and ad group level negatives—use this advantage.
Extract High Performers: Identify search terms from AI Max that consistently convert well, then add them as exact match keywords to your traditional Search campaigns. This "harvests" AI Max discoveries while maintaining control over bidding and messaging for proven queries.
Feature Toggle Testing: If all three features are enabled and performance is mixed, try disabling features one at a time to isolate what works. Start by disabling Final URL expansion (highest risk of wrong pages), monitor for a week, then re-enable or keep disabled based on results. Repeat with text customization and search term matching until you find the optimal combination for your account.
When to Pause AI Max
Consider pausing if:
- After a full evaluation period in an experiment, CPA is clearly higher than the control with no improvement trend
- A large share of impressions generating no conversions despite active negative keyword management
- Search terms report shows consistently irrelevant queries despite aggressive negative keyword management
- Budget consumption accelerates without corresponding conversion increases
- You're spending more time managing AI Max than it would take to manually manage traditional Search campaigns
Remember: pausing a test is a valid outcome. AI Max isn’t suitable for every account, and recognizing that early saves budget.
Advanced: Portfolio Rebalancing
If AI Max performs well in testing, shift budget gradually from traditional Search to AI Max, but keep both. Monitor overall account ROAS, not just campaign-level metrics, to confirm AI Max delivers true incrementality rather than taking conversions from other campaigns. If overall ROAS declines as AI Max budget increases, you have a cannibalization problem, not a growth opportunity.
Conclusion: Setting Realistic Expectations
AI Max for Search is Google’s attempt to extend Search campaigns beyond the keywords you chose: search term matching finds queries, text customization writes ad copy, and Final URL expansion picks landing pages. Google’s published case studies show what the features can do in the best-documented cases: Royal Canin’s 263% conversion increase in an A/B experiment, ClickUp’s 20% lift in incremental conversions, and Klook’s 161% rise in conversion value within a month.
But they’re selected examples, not typical results. Google’s own headline is that advertisers will typically see 14% more conversions or conversion value at a similar CPA/ROAS, and the higher 27% figure applies only to campaigns where more than 70% of conversions or conversion value came from exact or phrase match keywords. Both are Google internal data for non-Retail advertisers and neither is a guarantee for your account. Treat any single number, Google’s included, as a hypothesis to test.
At Digital Applied, our recommendation is to test AI Max with a campaign experiment instead of adopting it wholesale. Set up negative keywords, brand and URL exclusions before you start, review the search terms report regularly, and compare against the control on conversions and CPA/ROAS before scaling. Pause or narrow features that bring in irrelevant queries or pages.
AI Max is a layer within Search campaigns, not a replacement for Performance Max or for traditional keyword management. Decide the mix of Search, AI Max and Performance Max from your own experiment results rather than from a fixed budget split.
Google's push toward AI-powered advertising is inevitable—AI Max is just one step in that direction, alongside Smart Bidding Exploration and the broader automation of paid search. The question isn't whether to adopt AI tools, but how to do so strategically while maintaining human oversight and accountability. Test AI Max with realistic expectations, monitor performance honestly, and scale only what delivers measurable incrementality. That's how you succeed with AI Max in 2025, regardless of what Google's marketing materials promise.
Need Honest Google Ads Guidance?
Plan an AI Max test with experiments, clear exclusions and a measurement plan before you scale it. Get PPC strategy from an agency that prioritizes your results.
Frequently Asked Questions
Deep dives on AI, marketing and development.
Practical guides and fresh insights by email. No recycled takes.
Related Articles
Continue exploring with these related guides