Google’s AI Max for Search Campaigns is generating a wave of buzz across the PPC community. But between speculation, outdated ‘Search Max’ beta information, and a growing sense that “keywords are dead,” it’s easy to get lost.
Here’s everything you actually need to know about AI Max, straight from insights shared by Jyll Saskin Gales - former Googler, PPC consultant, and author of Inside Google Ads - to help you navigate this pivotal moment in search advertising.
Watch the full extended session with Jyll below - or keep reading for the written takeaways:
Timestamps:
0:00 - Intro
2:36 - ‘Inside Google Ads’ book discussion
4:08 - Why ‘the house always wins’ in Google Ads
6:45 - AI Max vs. Search Max
8:45 - What is AI Max?
9:57 - How transparent is AI Max?
11:34 - Is AI Max just broad match + DSA?
12:59 - Is this the ‘keywordless future’?
15:29 - Final URL Expansion + audience signals
22:20 - Should you test AI Max?
25:07 - Using Broad Match effectively
29:00 - Where Jyll uses Broad Match
31:24 - Key Broad Match advice & guardrails
34:14 - Pairing Broad Match with first-party data
39:00 - Final thoughts & outro
What is AI Max, really?
AI Max reflects Google’s shift to a search experience centered around intent prediction, rather than strict keyword matching.
It’s designed to consolidate multiple automation tools into a single, easy-to-enable bundle within standard search campaigns. This means advertisers don’t need to piece together disparate features to test more advanced automation; AI Max packages them so you can opt in with a single toggle.

Unlike Performance Max, AI Max focuses specifically on search inventory. Jyll noted that even Google insiders were calling it “Search-only PMax”, because it takes many of the dynamic elements of Performance Max (like final URL expansion) but restricts them to search results.
In theory, this allows you to leverage more aggressive automation without losing control of search-only campaigns, avoiding the uncertainty of Performance Max’s cross-channel reach.
In addition, Search Engine Land underscores that AI Max is not a standalone campaign type: it augments your existing search campaigns. This means you can roll out AI Max gradually across ad groups and campaigns you already trust, rather than rebuilding entire campaign structures from scratch.
This approach lowers the barrier to entry for advertisers eager to explore advanced automation but who may be wary of the black-box nature of Performance Max.
Clearing up the Search Max vs AI Max confusion
The confusion between “Search Max” and “AI Max” goes beyond a simple name change. Jyll shared her frustration with early speculation and misinformation:
“People were using AI to write posts about Search Max… Everyone was already giving advice about this thing that didn’t exist yet.”
During the closed beta, Search Max evolved rapidly under NDA, fueling rumors. Once the beta ended, Google rebranded it as AI Max to highlight its AI-driven search optimizations.
Understanding this timeline is important: Search Max was the experimental phase; AI Max is the real, publicly available solution. Focusing on AI Max ensures you’re acting on facts, not outdated speculation.
Is AI Max a black box?
AI Max reflects Google’s recent (welcome) trend of giving advertisers slightly more granular transparency compared to older automation tools like Smart Shopping or early Performance Max releases.
In order to win back the trust of marketers previously put off by the black-box, ‘trust us we got this’ nature of PMax, Google has introduced a few new transparency features unique to AI Max:
- Headline-level insights – you’ll see exactly which RSA headlines served alongside each search term, a level of creative performance detail not currently available.
- URL-level matching – you’ll know which landing page each query directed traffic to, helping identify gaps in your site or ad-to-page alignment.
This added clarity helps address one of the most persistent criticisms of automation: the feeling of flying blind. Instead of being limited to aggregate performance, you’ll know what messaging and destination pages are driving conversions (or wasting spend).
At GML 2025 when AI Max was announced, Google insisted that it ‘blends the best of Performance Max’s automation with the familiar transparency of search campaigns’.
Importantly, Jyll emphasized that AI Max is modular: you can enable or disable individual components like text customization or final URL expansion.
This means it’s not a take-it-or-leave-it black box (cough, PMax) - marketers can test AI Max selectively, layering automation where it makes sense and retaining manual control where it doesn’t.
As with the welcome addition of recent transparency features (such as campaign-level negative keywords in PMax), this is a promising sign that Google is listening to the marketers using their platform - or at least trying to create a show of good faith in the face of increased regulatory pressure.
Is AI Max giving or taking control?
One of the most common fears about AI Max is that it represents yet another step toward Google reducing advertiser control - especially given how black-box-y Performance Max can be.
Fortunately, Jyll made it clear: AI Max actually increases flexibility and control compared to many previous automated features.
Specifically, advertisers aren’t forced into an all-or-nothing adoption. AI Max allows you to opt into:
- Text customization – letting Google tweak your RSAs to better match predicted intent.
- Keyword list targeting – unlocking broader matching beyond your specified keywords.
- Final URL expansion – giving Google the freedom to send clicks to the most relevant page on your website.
Jyll stressed:
“You could, for example, just opt into final URL expansion, not the other two parts of it. It’s not like Performance Max, where you’re locked into a single approach.”
This modularity is important because it gives experienced PPC marketers a controlled testing ground to layer automation onto existing campaigns at their own pace.
A promising start, to say the least.
The shift from keyword matching to intent prediction
AI Max exemplifies the biggest paradigm shift in modern paid search: moving from keyword-centric to intent-centric advertising.
For years, advertisers have optimized around static keywords. But as Jyll explained, keyword definitions themselves have shifted - what “exact match” means today is wildly different from 5 years ago, and match types have become progressively broader.
She described AI Max as part of the second phase of the keywordless evolution:
“The first phase was when match type definitions all changed… Keywords have already died in the way we knew them.”
The real driver behind this shift is changing consumer behavior. As users interact with conversational AI like Gemini or make voice searches, they don’t always type neat keywords. And they expect search engines to infer intent from longer, more complex queries.
Search Engine Land’s analysis aligns with Jyll’s point: it highlights that AI Max’s keyword list targeting is designed to look beyond literal keywords and instead infer user needs, opening opportunities for advertisers to meet customers earlier in their journey.
But while intent-based targeting promises better alignment with modern user behavior, the keywordless future could backfire for marketers who lack clean data, struggle with offline conversion tracking, or operate in industries where precision is critical - like legal or healthcare.
Without keywords as reliable levers, advertisers may find it harder to exclude irrelevant traffic, diagnose campaign issues, or protect brand safety. This reduced ability to control exactly where and how ads appear can lead to wasted budget and poor lead quality if campaigns aren’t meticulously set up to guide Google’s automation.
This change is critical to understand because it signals the end of absolute keyword control. Advertisers who adapt will thrive; those who insist on rigid, outdated keyword structures risk being left behind as Google increasingly prioritizes understanding intent over matching words.
Opportunities and risks of keywordless PPC
Moving toward keywordless PPC has the potential to create major opportunities - but it’s not without serious risks, especially if you don’t feed Google the right data.
Potential opportunities include:
- Expanded reach into queries you may never have considered, as AI predicts intent beyond your keyword lists.
- Higher conversion potential from uncovering long-tail or related queries aligned with your goals.
- Reduced manual maintenance, freeing you to focus on strategy and creative assets instead of endless keyword pruning.
However, Jyll warned that these benefits hinge on your ability to provide Google with the right signals:
“It’s not that final URL expansion doesn’t work for lead gen… it’s that most lead gen advertisers don’t give Google the data it needs to give them the results they need.”
The most common risk is optimization for the wrong outcome. For example, if you’re only tracking top-funnel leads (like form fills) but never import offline conversions, Google might chase the cheapest, lowest-quality leads - often spam - because it doesn’t know what a qualified lead looks like.
In short, advertisers with limited or poorly structured conversion tracking are most likely to struggle with AI Max, since the system’s effectiveness depends on quality feedback loops.
To seize the upside while avoiding wasted budget, you must:
- Implement robust conversion tracking, including offline conversions where applicable.
- Regularly review AI Max’s detailed search term and RSA headline reports.
- Use your first-party data (like customer lists) to teach Google what signals indicate real business value.
Only by doing these can you ensure Google optimizes toward profitable actions rather than vanity metrics.
Final URL expansion: powerful but risky
One of AI Max’s most talked-about components is final URL expansion - first introduced in Performance Max, now arriving in search campaigns. With this feature enabled, Google can send clicks to any page on your site it thinks is most relevant.
While this can help you capture unexpected demand, it’s not without danger. For B2B lead gen advertisers with sprawling blogs or multiple solution pages, Google might send users to pages misaligned with your funnel, tanking lead quality.
Here’s Jyll’s advice:
- If you have offline conversion tracking set up and are feeding accurate data back to Google, final URL expansion can be powerful.
- If you can’t track which leads turn into real opportunities, you risk Google optimizing for spammy or low-value leads.
Should you avoid AI Max if you’re not ready?
If you’re still relying solely on exact match and are skeptical of automation, jumping straight into AI Max could be a recipe for frustration.
She recommended building towards AI Max readiness incrementally:
- Start by testing broad match in experiments.
- Run dynamic ad groups to get comfortable with looser keyword targeting.
- Use Performance Max (with tight target CPA or ROAS) to see how full automation behaves - keeping most traffic on search inventory.
AI Max’s unique features worth testing
Beyond what broad match and dynamic search ads can already do, AI Max introduces new capabilities worth exploring:
- Ad group-level location intent – You can specify an ad group focus on a specific location without reworking your keywords. E.g., target “hotels” with a Los Angeles focus by setting a location intent, rather than rewriting keywords with “Los Angeles” variants.
- More granular reporting – The ability to see RSA headlines and landing pages tied to specific search terms helps you pinpoint what’s working - and what’s wasting budget.
Using broad match effectively: lessons from AI Max
Broad match lies at the heart of AI Max’s ability to reach beyond rigid keywords:

But broad match’s reputation among marketers has long been checkered - and often for good reason. Many have burned budget on irrelevant clicks, seeing little ROI. As Jyll put it:
“If you are running broad match keywords and you look at your search terms report, you’re gonna see at least 50% hidden ‘other search terms,’ if not 80%, which means a lot of stuff you don’t even know what you’re advertising on.”
Yet broad match isn’t inherently bad. When used correctly, it can unlock incremental volume by finding high-intent queries you’d never think to add manually. But for broad match to deliver value, three critical conditions must be met:
- You have sufficient conversion data - Without enough data on what constitutes a valuable lead or purchase, Google’s algorithms can’t distinguish between good and bad queries.
- You have enough budget - Broad match relies on exploring many search variations. If your budget is too tight, it won’t gather enough signals to optimize performance.
- You give it time to learn - Early results can look messy, with seemingly wasted spend. But pulling the plug too soon will rob the system of its chance to figure out what works.
The best scenarios for broad match
So when should you consider using broad match? Jyll recommends it when you’re hitting the ceiling of your current performance, and want to scale:
- Maxed out on exact match - If your search impression share is already high on your exact match keywords, broad match can help capture additional relevant queries you’d otherwise miss.
- Large enough budget - For example, moving from $1,000/month to $1,000/day enables broad match to quickly gather the data it needs to optimize.
- Proper offline conversion tracking - Feeding Google signals on what real, revenue-driving conversions look like (not just form fills) helps ensure broad match spends your budget on the right prospects.
Jyll also noted her evolving perspective on match types, particularly phrase match:
“Phrase match has lost its sweet spot between control and reach, becoming the ‘good for nothing’ match type compared to exact or broad.”
In other words, marketers still relying on phrase match to play it safe may be missing opportunities - or wasting spend on a match type that no longer delivers meaningful differentiation.
Avoiding common broad match mistakes
Broad match works best when advertisers are realistic about the ramp-up phase. Expect some wasted spend early on. Pulling the plug too quickly prevents the system from learning.
Jyll’s (rather sweet) analogy frames it best:
“My son is two years old. He doesn’t say firetruck, he says ‘fire f***.’ I don’t yell at him to never try again - I correct him, and he learns. Broad match works the same: you need to give it time to learn.”
Instead of knee-jerk reactions to initial poor queries, monitor your campaign closely. Regularly review your search terms reports, add negative keywords thoughtfully, and look for improving trends before making major adjustments.
Feeding Google the right data
Whether you’re using broad match or AI Max, one truth remains: the quality of your inputs determines the quality of your results.
Jyll distilled her guidance into three essential elements - her three C’s - for Google Ads success:
- Conversion tracking - Don’t settle for measuring only leads or clicks. Prioritize tracking the real outcomes that generate revenue, such as qualified leads or completed purchases. Without conversion data, Google will optimize for the easiest or cheapest actions, which rarely align with your goals.
- Customer lists - Upload fresh customer lists to Google Ads regularly. Even if your lists are too small for direct targeting, they help train Google’s algorithms by providing better signals about your ideal customers.
- Creative - Ad automation can handle targeting and bidding, but it can’t write compelling ads or design effective visuals. Invest in high-quality ad copy, engaging headlines, and clear calls to action—these are what convince users to click.
Beware of audience signals illusions
A common misconception among advertisers is that feeding audience signals into Performance Max - or AI Max, for that matter - gives them meaningful control over who sees their ads. Jyll dispelled this myth:
“The campaign is going to show ads to whomever it thinks will hit your target CPA or ROAS, regardless of the signals. If the signals align, great - it would’ve found them anyway. If not, it’ll ignore them.”
While Google markets audience signals as a way to guide automation, in reality, they function more like suggestions. If your signals match what Google’s algorithm already wants to do, they’ll appear effective - but they don’t truly constrain or direct automated campaigns.
This means you can’t rely on audience signals alone to target niche segments or protect against irrelevant impressions. Instead, invest in robust conversion tracking and ensure your ad assets resonate with the types of customers you want to attract.
How to experiment with AI Max (responsibly)
If you’re planning to test AI Max, treat it as a scientific experiment—avoid rushing in blindly or assuming it will work out of the box. Set your campaigns up for meaningful learning with these best practices:
- Set clear benchmarks - Define what success looks like ahead of time, comparing AI Max performance to your current campaigns on metrics like CPA, ROAS, or lead quality.
- Isolate AI Max tests - Don’t mix AI Max ad groups with standard ones in the same campaign. Use Google Ads experiments or separate campaigns so you can attribute results accurately.
- Give it time - Like any automated system, AI Max needs enough conversions and budget to learn. Cutting it off prematurely will lead to false negatives about its effectiveness.
- Monitor new insights - Watch the detailed search term and RSA headline reports AI Max provides. These insights are unique opportunities to understand how Google is matching intent with your ads—and to refine your messaging.
By approaching AI Max methodically, you can get a clearer picture of whether it genuinely drives incremental value for your business.
AI Max: Final thoughts
AI Max represents the next stage of Google’s push toward intent-driven, keywordless search.
It’s far from a silver bullet - without solid data, conversion tracking, and patience, it can easily waste your time and budget. But used properly, and when fed the right information, it does have potential to unlock scalable, incremental growth.
Huge thanks to Jyll Saskin Gales for sharing her expert insights on AI Max and broad match with us. Make sure you’re following Jyll on LinkedIn if you’re not already, as she regularly shares valuable, no-nonsense takes on the latest ad platform updates, campaign types, and strategies.
If you’re looking for a deeper understanding of Google Ads (and advice on how to scale and succeed your campaigns effectively), Jyll’s website hosts a vast range of information, courses, and coaching available to all marketing professionals.
Want more practical advice from the top voices in the PPC industry? Subscribe to Lunio’s YouTube channel to ensure you don’t miss an episode - or find the Paid Media Lab wherever you get your podcasts.
Say goodbye to wasted ad spend
Discover how Lunio can help you eliminate invalid ad clicks and maximize paid media performance