More conversions at the same cost? This is what Google promises with AI Max.
After initial tests with several accounts, the results are mixed: In about one quarter of our test accounts, AI Max delivers clear performance improvements. In the remaining three quarters, however, results fell short of traditional campaigns or showed no significant improvements.
How does AI Max work?
AI Max combines three AI features that interact intelligently:
- Search Term Matching expands reach beyond existing keywords. Example: If you are bidding on the keyword “buy fitness tracker”, you may also reach “health wristband with activity tracking” – without having explicitly listed that keyword.
- Asset optimisation automatically creates tailored ad copy that matches search intent. The system uses existing ads, landing page content and generative AI for real-time adjustments.
- Final URL Expansion automatically directs users to the most relevant landing page – similar to Dynamic Search Ads, but seamlessly integrated.
Performance numbers: a mixed picture
With a fifty-fifty budget split between classic search campaigns and AI Max, successful tests show impressive results: roughly 150 percent more conversions at nearly half the cost per conversion. Click prices remain stable.
However, these numbers do not apply across all accounts. In around 75 percent of our tests, AI Max lagged behind classic campaigns or did not show meaningful improvements. The reasons vary, from insufficient data volume to highly specific niche markets to website content that is difficult for AI systems to interpret.
In other words: AI Max may work very efficiently and deliver significantly better results – but in many cases, it does not.
The key difference: contextual understanding instead of keyword matching
What makes AI Max different? It better understands context within search behaviour and therefore reaches relevant users who would often be missed by traditional campaigns. While classic keyword campaigns rely on exact or broad matching, AI Max analyses overall search behaviour contextually.
Returning to our example above: This contextual understanding allows AI Max to connect the keyword “buy fitness tracker” with queries such as “health wristband with activity tracking” or “which activity tracker is right for me?”. Under classic keyword matching, these queries would not be considered.
Looking at existing broad match keywords and their associated search terms often provides a first indication of whether an account offers a good foundation for AI Max. If this data is clean and well structured, AI Max can build on it effectively.
AI Max as a solution in a zero-click reality
The performance increases are no coincidence. They align with fundamental changes in online search – and with Googles strategic challenge.
The search landscape is shifting with ChatGPT, Perplexity, Gemini and other LLMs answering queries directly, without users visiting a webpage. At the same time, Google searches are becoming more natural and complex. The result is a zero-click world in which classic keyword marketing loses relevance.
Googles answer: AI Max is built for this new search logic. Instead of focusing on individual keywords, the system analyses users actual interests in a contextual way. This bridges traditional search marketing and a future dominated by AI – while also protecting Googles business model against competing LLMs.
The key question is no longer just “Which keywords should I book?” but rather:
“How do I reach users in a world where AI systems deliver the answers?”
What matters for successful AI Max campaigns
The critical foundation: Conversion Tracking
Without correctly implemented conversion tracking, no AI-based campaign can optimise effectively. AI Max is only as good as the data it learns from.
Technical requirements:
- Precise conversion tracking with clearly defined success metrics
- Cleanly implemented smart bidding strategies
- Thematically organised account structure without duplicate keywords
- High-quality website content
How to get started:
- Test gradually with Google Experiments (50-fifty split)
- Allow at least two weeks of learning and eight weeks of runtime for a solid evaluation
- Review search terms every one to two weeks
More transparency than Performance Max
AI Max offers significantly better insights than Performance Max:
- The Search Terms Report shows exactly which queries triggered the ads.
- The Keywords Report quantifies how much traffic comes from AI expansion.
- The Asset Report reveals which automatically created content performs best
When AI Max does not work
In our tests, we also identified accounts where AI Max did not meet expectations.
The most common reasons:
- Too little data: Accounts with few conversions per week do not provide enough learning signal. AI Max needs volume to recognise patterns.
- Highly specialised niche markets: For very specific B2B products or technical niche solutions, the AI often cannot find relevant expanded search queries. Classic keyword control works better here.
- Weak website content: If landing pages have little text, unclear structure or generic content, AI Max cannot perform effectively. The system relies on high-quality information.
- Missing conversion data: Without clean tracking, AI Max optimises blindly.
- Brand-only campaigns: Here, AI expansion adds no value, as users are already searching directly for the brand.
Testing pays off – but without guarantees
Our initial tests show: AI Max for Search can deliver impressive results. About 150 percent more conversions at half the cost per conversion – this works in roughly one quarter of our test accounts. In three quarters, however, results fell short.
Our recommendation: Use Google Experiments to test AI Max in parallel with classic campaigns. With a 50/50 budget split, you will have initial insights after eight weeks to determine whether AI Max works for your account. Without testing, you will not know.
AI Max is not a miracle solution, but an important tool in a changing search landscape. In a zero-click world in which LLMs increasingly answer queries, new approaches are needed. AI Max is one of them, but not the only one.
The classic keyword world is changing fundamentally. Anyone who wants to remain successful in Google Ads should test AI Max. But with realistic expectations: It works well, but not everywhere.
