30 Days of AI Max Testing: The Results Are In

Google unveiled its new automated features for Google Ads, AI Max, in May 2025 and rolled them out globally in September.

According to Google, data shows that advertisers are seeing 14% higher conversion rates¹ at comparable costs per acquisition, with some campaigns hitting 27%.¹ Those are tempting numbers, but is testing AI Max worth it? 

Our tests suggest that boosting your ROI may not be as easy as toggling on the Max button, sitting back, and raking in conversions. Results vary across different account types, like ecommerce and lead gen.  

At Coalition Technologies, we follow the data. Our decisions are based on real-world benchmarks. As always, you can count on us for unbiased information and guidance. 

In this AI Max case study, we’re going to help you determine if you should adopt, abandon, or somewhere between.

What AI Max Is and How It Works 

AI Max is a suite of AI-powered features for Google Ads Search campaigns. It’s not a new type of campaign; it’s an optimization layer that leverages existing account data to automate ad creation, targeting, and landing page selection.

If you’re considering testing AI Max, be aware that it relies heavily on existing conversion data for training. Sites with limited conversion data may not be able to take full advantage of the new features.

Frequently Asked Questions

How is AI Max different from Performance Max?

• PMax is a new campaign type designed for cross-channel advertising throughout Google properties, such as Search, Display, Discover, Gmail, YouTube, and Maps. 

• AI Max is a new layer that uses artificial intelligence to optimize existing Search campaigns, focusing solely on the Search network.

How will I know if AI Max is working?

Google added a new AI Max category in the Search Terms Report. When you’re testing AI Max, you can simply filter the Match Type for AI Max, which will reveal the new AI-generated queries.

Are there any real-world success stories?

AI Max is still relatively new to the scene, so there aren’t a whole lot of metrics available. However, we did use AI Max to increase one of our ecommerce clients’ revenue by an impressive 70.15% within just 30 days, as we’ll show in the AI Max case study below.

Here are two other noteworthy examples of AI Max success stories as reported by Google:

• L’Oréal’s conversion rate doubled, and the cost-per-conversion declined by 31%.¹
• MyConnect generated 16% more leads and reduced their cost-per-action by 13%.¹

Remember: Success isn’t guaranteed. Our AI Max testing showed that AI Max can be effective for some clients. However, simply toggling AI Max on doesn’t turn on the money faucet. You need a solid framework in place to take advantage of AI Max’s features.

Will AI Max replace my existing Google Ads campaign?

No, AI Max is an optional layer that enhances existing campaigns. It uses your current keywords and assets, as well as “keywordless” context clues, to find and predict new search queries you may have missed. The goal is to better match your ads with user intent.

Can I use AI Max features individually?

Yes, you can toggle each feature on and off separately in your campaign’s settings.

Is it worth testing AI Max if I don’t have a lot of conversion data?

Having a substantial amount of conversion data is very important, but it’s not the only factor to consider. Throughout our AI Max case study, we found that lead-gen accounts with a strong conversion history performed poorly. We’ll elaborate more below.

What features are included with AI Max?

Main Features:

Here are the core features AI Max brings to the table:

• Search Term Matching: Uses your data to discover new high-potential keywords.

• Text Customization: Uses your existing text to create new descriptions and headlines that align with the searcher’s query.

• Final URL Expansion: Rather than send the searcher to your homepage, AI Max directs them to the most relevant page.

Settings and Reporting Features:

Here’s how you can customize and analyze AI Max and live campaigns:

• URL Inclusions/Exclusions: Ensures AI Max only shows the pages you specify.

• Brand Settings: Lets you tell AI Max which brands you do and don’t want your ads to show up for.

• Ad Group Level Locations of Interest: Lets you target people who may be interested in a location, rather than just those who are physically there.

• Enhanced Search Terms Report: Shows the new keywords generated by AI Max.

• New Reporting Views: Shows the full scope of the searcher’s journey, including their query, text on the ad they were shown, and the page they arrived on.

Flip the Switch? Potential Benefits of AI Max 

Should you bother testing AI Max? The most compelling finding of our AI Max case study was the sharp contrast between the respective performances of ecommerce and lead-gen accounts. 

Ecommerce accounts with a strong conversion history consistently had impressive results. The same cannot be said for our lead-gen accounts, which performed poorly.

Ecommerce Case Study:

We tested multiple accounts. Here’s one example:

We activated AI Max on September 30 for an ecommerce client. 

To ensure that AI Max wouldn’t cannibalize branded search campaigns’ traffic, and to prevent traffic from reaching low-value pages, we excluded the brand list and irrelevant URLs.

Our AI Max testing showed a 70.15% increase in revenue with a 72.01% boost in its ROAS (return on ad spend) compared to the previous 30 days.

It’s important to note that while AI Max-driven search terms weren’t irrelevant or concerning, they also didn’t directly generate very many sales. 

However, activating the new layer did improve overall campaign performance, most likely by feeding Google’s bidding algorithms stronger conversion signals.

Conclusion: The numbers don’t lie. AI Max is worth taking for a test drive for ecommerce accounts with a strong conversion history. 

Although the AI-generated terms themselves didn’t produce exceptional results, AI Max can help to uncover high-intent audiences and expand reach, particularly when traditional Search has already plateaued and new avenues are required.

Risks and Trade-Offs

Our AI Max testing definitively showed that lead-gen accounts should keep the toggle off, at least for the time being. Even accounts with substantial conversion data underperformed due to ambiguous or weaker quality signals.

Lead-Gen Case Study:

Although we tested multiple accounts, we’re going to use the data from one of our biggest lead-gen clients to highlight the risk:

The client’s impression share from our tested branded search campaign declined sharply to 58% after activating AI Max. We also noticed a 15-20% drop in leads.

Conclusion: Our AI Max case study showed that a solid conversion history isn’t enough for lead-gen accounts to come out ahead. We strongly recommend sticking with traditional ads unless AI Max changes.

Other Risks to Keep in Mind

Even if you’re in the Goldilocks zone for AI Max and implementation, being an ecommerce business with a strong conversion history to pull from doesn’t guarantee success. If your formula is already working, changing it is inherently risky. With that said, great risks (sometimes) lead to great rewards.

Keep in mind that using AI Max means losing control over your ad copy and search terms. Many brands have a specific tone and feel they wish to maintain. While AI MAX does attempt to replicate existing copy, you risk losing the human touch your audience connects with.

To see AI Max’s true potential and, ideally, generate positive outcomes, you really need to pay attention to brand and URL exclusions. Throughout our AI Max testing, we found these settings to be vital for guiding AI Max in the right direction. There’s a high risk of wasted spend if you don’t exclude irrelevant and low-value pages and brand terms.

Pass or Pursue: Is AI Max Right for You?

Every AI Max case study we performed returned similar results: AI Max is suitable for mature ecommerce campaigns with strong site content and stable conversion tracking. The ROI can be significant, though not guaranteed.

On the other hand, we recommend that lead-gen accounts skip AI Max, at least for now. We also advise that niche accounts with scarce history avoid AI Max, as our testing showed that robust data is required to train the AI effectively.

Finally, we recommend that compliance-heavy accounts and businesses with unique, human-driven voices avoid AI Max, as its AI-generated ad copy may cause issues. 

Our Recommendation: An Incremental Approach

If you’re considering turning the toggle on, we recommend testing alongside traditional Search, not replacing it entirely. Pay attention to incremental lift, not just direct AI Max conversions. 

We’ve already used AI Max to generate revenue for our ecommerce clients. We’re confident that we can deliver real results because we’ve completed extensive AI Max testing. We know what works.

Ready to join a winning team? Partner with us to take full advantage of AI Max’s potential and make Google Ads work for you.

Source:

1. https://blog.google/products/ads-commerce/google-ai-max-for-search-campaigns/