With the global launch of Microsoft AI Max for Search campaigns, it’s worth spending some time unpacking where the features align and differ from Google’s version.
Much of the core functionality is the same:
- Search term matching expanding your reach beyond static keyword lists. AI Max uses your keywords, ads, and landing pages, along with intent and contextual signals, to uncover relevant searches you may not qualify for with keywords alone. This is especially helpful for complex conversational queries.
- Text customization allowing ads to adapt to high-value placements and prospects. AI Max uses your existing assets and website content to generate and test additional messaging variations. It then selects the most appropriate combinations at auction time, helping you to deliver more relevant ad creative, including in AI-native experiences.
- Final URL expansion that routes people to the page on your site that best matches their intent. Instead of always sending traffic to a static landing page, AI Max can route people to the page that best matches what they’re looking for. This ensures a more consistent user experience across query, creative, and website.
However, there are some key nuanced differences. We’ll dive into:
- Where Google and Microsoft AI Max are the same.
- Key points of differentiation between Google and Microsoft AI Max.
- How to leverage AI Max successfully in new and existing account structure.
(Note: I am a Microsoft Advertising employee and wrote this post as platform-agnostically as possible. Features discussed in this post are based on publicly available help documentation as of September 2026.)
What’s the same between Google and Microsoft AI Max?
Aside from the core functionality, the following AI Max mechanics remain the same across both Google and Microsoft.
AI Max is a setting, not a campaign type.
Unlike Performance Max (PMax), Demand Gen, Audience ads, and other unique campaign types, AI Max represents optional settings within Search campaigns. These settings are designed to work better together, and advertisers tend to see the most benefit when they opt into all three.
Here’s how advertisers benefit when they leverage all three core AI Max features across Google and Microsoft:
- Search term matching allows for net new queries and that can bring the advertiser into auctions their static ad might not accurately reflect. By allowing the text to adapt in the moment, you’ll ensure your ads are relevant for those new queries.
- To deliver the most relevant landing page for unique queries, Final URL expansion can be really helpful. Final URL expansion is linked with Text asset generation as platforms need to be able to adapt creative to make a promise the dynamic landing page can deliver on.
- If you’re going to allow for text generation, you’ll get the most bang for your optimization buck by also allowing for intelligent matching of URLs and relevant queries.
That said, if advertisers want to begin with a more conservative test, they absolutely can.
For example, an ecommerce brand selling products with similar margins might feel more comfortable testing URL expansion because it allows them to get fuller coverage of their products without building out unique ad groups, but don’t want to leverage search term matching. While Brand Controls exist and can help ensure certain brands aren’t included (more on that later), it’s fair for the advertiser to start with just Final URL Expansion and Text customization.
Whether an advertiser tests just one, two, or all three parts of AI Max, these tests can be done safely with experiments. While both platforms support controlled testing, Google’s AI Max experiments are designed to run within the existing campaign by diverting traffic, whereas Microsoft Advertising’s Search Experiments compare a standard campaign against a cloned test version with AI Max enabled.
Here is guidance on how to experiment with AI Max:
- Start with your strongest campaign. Choose a campaign with stable performance and enough volume to generate results. Experiments work best when there’s enough traffic to detect meaningful differences.
- Create a 50/50 split. Split traffic evenly between the control and experiment. Keeping traffic balanced makes it easier to determine whether performance differences are due to the change you’re testing.
- Let the experiment learn. One common mistake is ending a test too quickly. New bidding strategies and AI-powered features need time to learn. If you’re testing AI Max specifically, go directly into an A/B test and allow sufficient learning time before drawing conclusions.
- Measure business outcomes, not just clicks. Focus on:
- Conversion rate
- Cost per acquisition (CPA)
- Return on ad spend (ROAS)
- Changes in revenue or conversion value
Conversion-based bidding is a core part of search term matching
AI Max’s search term matching relies heavily on conversion data in order to be successful. This is why both Google and Microsoft AI Max require conversion-based bidding when enabling this part of AI Max.
Conversion-based bidding works when there’s accurate conversion data flowing into the platform. It is ideal to have at least 15-30 conversions in a 30-day period before working with conversion-based bidding.
If your brand likely won’t hit the threshold, it makes sense to wait on trying AI Max until you can or to use micro-conversions with strategic conversion values. This means setting conversion values aligned to each stage of the journey and TROAS that reflects the focus on important steps.
For example, if a brand wanted to secure applications for their financial product, they might include conversion goals reflecting different completion milestones (beginning, mid-way, completed, and accepted). These conversion goals would get a profit-related conversion value associated:
- Beginning application: $10
- Mid-way: $20
- Completed: $50
- Accepted: actual conversion value using offline conversion uploads.
Brand controls exist on both, though there are some differences in what’s available.
Google and Microsoft both understood advertisers need messaging and branding controls to ensure ad creative stays within style guides. That’s why both AI Max variants allow advertisers to set brand inclusions, exclusions, term exclusions, and message constraints.
To make it easier to understand the mechanics for each, here’s the breakdown:
Google:
- Brand inclusions: 10 brand lists per campaign, with up to 5,000 brands per list
- Brand exclusions: 10 brand lists per campaign, with up to 5,000 brands per list
- Term exclusions: 25 per campaign
- Message constraints: 40 per campaign
Microsoft:
- Brand inclusions: 20 brand lists per campaign, with up to 100 brands per list
- Brand exclusions: 20 brand lists per campaign, with up to 100 brands per list
- Term exclusions: 25 per campaign
- Message constraints: 40 per campaign
Microsoft offers disclaimers that do not take up ad real estate, and work with AI Max. Google is piloting disclaimers as of this writing that would take up description line #2.
What’s unique between Google and Microsoft AI Max?
While most of the core AI Max functionality remains the same, there are a few key differences to account for when moving between ad platforms.
Consolidation of settings vs being able to pick and choose all three.
AI Max features work better together, so it makes sense to opt into all three. However, if you want to test individual settings before committing to the full suite, Microsoft makes this easier by offering all three features as opt-in toggles.
Google opts all advertisers who turn on AI Max at the campaign level into search term matching, which can be turned off at the ad group level.
While Microsoft maintains all existing ad group targeting settings (including location targeting, scheduling, and time zone selection), there are no AI Max-specific ad group settings.
Google also supports Locations of interest, URL inclusions, and Brand inclusions at the ad group level.
This means Google’s AI Max management asks advertisers to make more decisions at the ad group level, while Microsoft focuses AI Max settings at the campaign level.
Matching mechanics and search term transparency are different.
Both Google and Microsoft recommend leaning away from syntax-oriented keywords as they can get in the way of being eligible to serve for complex and longer queries, especially in AI experiences.
However, Google and Microsoft take different approaches to matching and search term reporting. Microsoft offers full search term reporting for any query resulting in a click for AI Max, PMax, and traditional Search and Shopping campaigns. These can be found in the reporting templates, Search term and Search term landing page.
Google hides some search terms for privacy reasons, which also means there isn’t full transparency on whether the query is relevant. To mitigate this, Google allows for more close variant mechanics in negative keywords.
When it comes to matching, due to different ecosystems, there are inherently different signals about how Google and Microsoft will match user queries to advertiser campaigns. Here’s a breakdown of the main signals by each ad platform:
Google:
- YouTube data
- Previous search behavior
- Conversion data
- Landing page
- Other keywords in the ad group
- Audiences: In-market, demographics, and other 1P audiences like Customer Match).
Microsoft:
- LinkedIn data
- Previous search behavior
- Conversion data
- Landing page
- Other keywords in the ad group
- Audiences: Impression-based remarketing, In-market, demographics, and other 1P audiences like Customer Match
How to leverage AI Max in existing and new account structures?
AI Max brings the best of AI functionality to search campaigns, and it’s understandable that advertisers will want to test some or all of the AI Max feature suite. However, there are some key considerations for accounts bringing AI Max into their structures.
Here are the top five considerations:
- Do you trust your conversions measurement and do you meet conversion thresholds?
- Will your landing pages be a help or a hindrance in conveying what your brand offers to AI systems as they address human questions?
- Are your existing ad assets on brand or do they have serious deviation?
- Is PMax already part of your account structure?
- Have you budgeted for the targets you’re setting?
Let’s dig deeper into each one.
Do you trust your conversion measurement and do you meet conversion thresholds?
AI Max and conversion-based bidding are joined at the hip because conversions are a critical signal AI Max uses to help advertisers connect with the right customers at the best ROI. If your campaigns don’t have accurate conversions feeding into the system, or don’t have enough conversion data (15-30 conversions in a 30-day period is the minimum), it will be very hard for AI Max to make intelligent matching choices.
If you have an existing account with at least 90 days of accurate conversion data, AI Max is a no-brainer.
Newer accounts should work on building conversion data so they can leverage AI Max and conversion-based bidding once the account ramps up.
Will your landing pages be a help or a hindrance in conveying who you are to the ad platform?
Landing pages are a critical signal that ad platforms use to understand your utility to a potential user. If your landing page is accessible in phrasing and visuals, this translates to easier content consumption for AI and humans alike.
A really great example of this is including alt-text on images and videos on the landing page, which ensures there can be no doubt about the subject matter. On the flip side, if you don’t make it clear what you’re offering and why your customers enjoy working with you, it can be hard to translate those messages to AI creative and matching.
A common mistake brands will make is disallowing all bots from crawling their landing pages, which deprives AI from critical insights into who you are and how you can help your customers.
A good way to understand whether your landing pages are AI Max compatible is to plug them into the PMax campaign creation flow. If you strongly disagree with the assets Google or Microsoft generated for you, that can be a sign to adjust landing page copy/mechanics before turning on AI Max.
Are your existing ad assets on brand or do they have serious deviation?
Both Google and Microsoft AI Max rely on existing text assets to inform potential new ad creative. Beyond adding in Brand Controls (including term exclusions and message constraints), it’s really important that your ad assets reflect the guidance you share.
For example, if you give a style guide note that all headlines should be sentence-cased, but your existing headlines are title-cased, that can cause confusion in the system. It’s important to audit your ad creative and landing pages for phrasing choices that might not align or were included unintentionally.
Is PMax already part of your account structure?
AI Max takes the best parts of PMax AI and layers them as optional boosts to Search campaigns. That means there are inherently fewer net-new opportunities for AI Max to unlock in accounts already running mature PMax campaigns.
However, AI Max can still add value when PMax is focused primarily on Shopping, is budget-constrained, or isn’t fully capturing your search opportunity. Rather than evaluating AI Max and PMax in isolation, focus on whether the combination is driving incremental account-level growth in conversions, revenue, or efficiency
PMax, by its nature, is cross-channel, and has a strong affinity for ecommerce. It can be useful to have AI Max as a search specific tool to go after parts of your business you don’t want exposed to non-search inventory.
In short, running AI Max and PMax in the same account isn’t inherently good or bad. If you want the full AI performance lift, Performance Max will have an easier time delivering that because it’s not restricted to search only surfaces. AI Max represents the useful AI gains of PMax in choice oriented and search-specific experience.
Have you budgeted for the targets your setting?
AI Max requires leaning into conversion-based bidding (“Smart” on Google and “Auto” on Microsoft). One of the biggest reasons any campaign can fail is it’s asked to go after too many targets for the budget.
For example, if you’re targeting customers for your plumbing business, it won’t be useful to ask the same campaign to go after minor repairs and a burst pipe. This is because the services have different costs, levels of urgency, and service capacities.
As a general rule, all services/products within a campaign should be within 20-30% of each other. If there’s too much variance, make sure you’ve set up accurate conversion values/TROAS goals, URL exclusions, and budgeted enough to cover the spread.
Final takeaways
Ultimately, Google and Microsoft’s AI Max are fairly similar. The differences have more to do with platform-specific mechanics.
Both agree that you’ll see the best results when you opt into all three core AI Max features.
Google puts more AI Max functionality at the ad group level, while Microsoft makes it a campaign-level choice.
Both platforms actively take on advertiser feedback, so if there is a preference for one style of management vs another, it’s worth sharing through support.
