AI hasn’t replaced traditional search – it’s expanding it, according to Robby Stein, Google’s VP of Product for Search, speaking in a new interview on Lenny’s Podcast.
Google is seeing more searches than ever as people ask harder, more conversational, and more visual questions powered by AI, Stein said:
“People come to search for just ridiculously wide set of things. … They want specific phone number. They want a price for something. They want to get directions. They want to find a payment web page for their taxes. Like every possible thing you can imagine.
“I think the vastness of that is underappreciated by many people. And what we see is that … AI hasn’t really changed those foundational needs in many ways. What we’re finding is that AI is expansionary.
“There’s actually just more and more questions being asked and curiosity that can be fulfilled now with AI. That’s where you get the growth. The core Google search isn’t really changing in my opinion, we’re not seeing that. But you’re getting this expansion moment.”
Stein pointed to Google Lens as proof: a 70 percent year-over-year surge in visual searches. “Billions and billions and billions of searching this way,” he said.
AI Mode and the future of search. Stein essentially called AI Mode a new layer of Search that gives searchers a “consistent, simple product experience,” where they don’t have to think about where they are asking a question:
“[AI Mode] creates an end-to-end frontier search experience on state-of-the-art models to really truly let you ask anything of Google Search. You can go back and forth. You can have a conversation. And it taps into and is specially designed for search.
“It’s able to understand all of this incredibly rich information that’s within Google. So there’s 50 billion products in the Google Shopping Graph, for instance. They’re updated 2 billion times an hour by merchants with live prices. You have 250 million places and maps. You have all of the finance information. Not to mention you have the entire context of the web and how to connect to it so that you can get context but then go deeper.
“And you kind of like put all of that into this brain that is effectively this way to talk to Google and get at this knowledge. That’s really what you can do now. You can ask anything on your mind and it’ll use all of this information to hopefully give you super high-quality and informed information, as best as we can.
“It’s also been integrated into our core experiences. You can get to it really easily. You can ask follow-up questions of AI Overviews right into AI Mode now. Same for the Lens stuff – take a picture takes you to AI Mode, you can ask follow-up questions and go there too. So it’s increasingly an integrated experience into the core part of the product.”
GEO and content advice. Stein was asked about the rise of AEO (answer engine optimization) and GEO (generative engine optimization) and what it means. The rules of showing up in AI answers haven’t changed as much as people think, Stein said, noting that Google’s AI still searches – just a lot faster and smarter.
“When our AI constructs a response, [it] does something called query fan-out, where the model uses Google search as a tool to do other querying. So, maybe you’re asking about specific shoes. It’ll add and append all these other queries, like maybe dozens of queries, and start searching in the background. And it’ll make requests to our data back end, so if it needs real-time information, it’ll go do that.
“And so, actually something’s searching. It’s not a person. But there’s searches happening. And then each search is paired with content.
Stein referenced Google’s quality rater guidelines and seemed to indicate that SEO best practices still apply in the evolving era of GEO/AEO:
“Do you satisfy the user intent of what they’re trying to get? Do you have sources? Do you cite your information? Is it original, or is it repeating things that have been repeated 500 times? And there’s these best practices that I think still do largely apply because it’s going to ultimately come down to an AI is doing research and finding information.
“And a lot of the core signals – is this a good piece of information for the question? – they’re still valid. They’re still extremely valid and extremely useful. And that will produce a response where you’re more likely to show up in those experiences.”
Stein’s advice for publishers and creators:
“Think about what people are using AI for. I mentioned this is an expansionary moment, right? Like seems to be that people are asking a lot more questions now, particularly around things like advice or how to or more complex needs versus more simple things. If I were a creator, I would be thinking, what kind of content is someone using AI for? And then how could my content be the best for that given set of needs now? And I think that’s a really tangible way of thinking about it.”
How Google AI search differs from competitors. AI Mode isn’t a chatbot – it’s designed and specially created for informational needs (planning, learning, verification), not therapy, productivity, or creativity, Stein said.
“We’re really focused on what people use Google for and making an AI for that so that you can come to Google, ask whatever you want, and get effortless information about that, and context and links to then also verify, dig in, and go to the authoritative sources ultimately that people want.”
So perhaps we should call it IEO (information engine optimization) instead of AEO, GEO, etc.?
Bottom line. Google Search isn’t shrinking – it’s expanding due to multimodal searches, according to Stein. It’s being rebuilt to be “the best at informational needs.” That means answering natural language questions, not making searchers speak “keyword-ese.”
https://i0.wp.com/dubadosolutions.com/wp-content/uploads/2025/10/konsqqvbiey-DPZ5yd.jpg?fit=1280%2C720&ssl=17201280http://dubadosolutions.com/wp-content/uploads/2017/05/dubado-logo-1.png2025-10-13 15:39:172025-10-13 15:39:17Google’s Robby Stein on AI Mode, GEO, and the future of Search
But engine numbers aren’t always the “source of truth.”
Sometimes they’re only directionally accurate. Other times, they barely correlate with actual performance.
Here’s the risk when you don’t define that source upfront: you build and present a solid report, only to have it derailed by, “I don’t think these numbers are right.”
A client questions whether Google Ads is inflating conversions, or a CFO insists revenue must come from the CRM.
Suddenly, the discussion shifts from strategy to data defense.
When stakeholders don’t trust the numbers, your report loses its power. You can’t drive action on data that no one believes.
So before building a report, clarify the source of truth.
A quick litmus test: if you said, “We generated $1 million in PPC revenue yesterday,” what system would leadership check to verify it?
Whatever they name is your source of truth.
You may never reconcile every dataset perfectly, but alignment matters most.
Pull numbers from that trusted system where possible, call out known gaps – like offline conversions lagging in Google Ads or modeled data in GA4 – and always identify data sources clearly.
Without one, visitors don’t know what to do next – and conversions drop.
Reports work the same way, only without a button to click.
That’s why I developed a framework I call “invisible CTAs.”
An invisible CTA is the intended outcome for each section of your report – the “conversion” you want your audience to experience.
It doesn’t appear in the report itself, but it guides how you build every chart, annotation, and insight.
There are three types of invisible CTAs:
Do: The next step they should take based on the data – fix a landing page, approve budget reallocation, or adjust strategy to defend against a competitor.
Know: What happened and why, even when there’s no immediate action – a holiday promo drove a 15% spike that won’t sustain, Apple’s privacy updates reduced match rates, or a tracking glitch underreported conversions.
Feel: The emotional response that drives urgency or confidence – concern that a competitor is outspending you, encouragement that a new strategy is working, or worry that rankings are slipping.
Don’t shy away from negative emotions.
When we hide problems to keep reports “positive,” stakeholders won’t commit the resources needed to fix them.
Think of it this way: which battery icon motivates you to get off the couch and grab your charger?
Not the full one.
Before building any section, ask:
What’s the one takeaway I want my audience to leave with?
Then design everything – your charts, metrics, headlines, and comparisons – around that invisible CTA.
When each section has a clear intent, your audience knows exactly what to do next, even without clicking a button.
The purpose of PPC reporting is simple – to help your audience understand what happened and what to do next.
If your reports don’t accomplish that, you’re not just wasting time. You’re leaving your readers without the clarity they need to act.
When you design reports around your audience’s needs, anchor them to a trusted source of truth, build invisible CTAs, apply conversion principles, and show results in context, you turn reporting into a decision-making tool.
Follow these steps, and your PPC report will stop being a monthly time-sink and start becoming a high-value asset that earns trust, drives action, and strengthens retention.
https://i0.wp.com/dubadosolutions.com/wp-content/uploads/2025/10/Identify-interview-your-audience-VWlIgW.png?fit=1320%2C633&ssl=16331320http://dubadosolutions.com/wp-content/uploads/2017/05/dubado-logo-1.png2025-10-13 13:00:002025-10-13 13:00:005 ways to drive action with your PPC report
Instacart today became the first retail media network to integrate directly with TikTok Ads Manager. This will allow CPG advertisers to use Instacart’s first-party retail data to target audiences, measure conversions, and drive shoppable experiences – all without leaving TikTok’s platform.
The integration marks a major step in the convergence of retail media and social commerce. By embedding Instacart’s targeting and closed-loop measurement capabilities into TikTok, brands can connect with high-intent consumers at the exact moment of inspiration and track their impact through purchase.
Why we care. For CPG advertisers, this partnership removes a friction point – tying social engagement directly to grocery purchases. It enables smarter audience targeting, more personalized creative, and real-time performance insights within TikTok’s ecosystem, where over 180 million monthly U.S. users can discover products.
The details. Advertisers can now:
Target high-intent shoppers using Instacart audience segments.
Power shoppable TikTok ads with grocery selection data from Instacart.
Measure campaign performance daily through Instacart’s closed-loop conversion data.
Microsoft is officially winding down its DSP, Microsoft Invest, and naming Amazon DSP as its preferred transition partner – a move that deepens ties between the two tech giants and reshapes Microsoft’s advertising strategy.
Starting Feb. 28, 2026, Microsoft will sunset Microsoft Invest as it shifts focus to its core advertising products: Microsoft Advertising Platform, Monetize, and Curate. The company says the partnership with Amazon DSP will ensure advertisers experience a smooth transition and continue achieving performance goals.
Why we care. Microsoft’s move to sunset its DSP and align with Amazon DSP means advertisers who use Microsoft Invest, will need to migrate campaigns, data, and workflows – but in return, they’ll gain access to Amazon’s powerful audience insights, premium inventory, and stronger performance capabilities across both Amazon and Microsoft ecosystems.
The details. Microsoft Monetize will also join Amazon Ads’ Certified Supply Exchange program, enabling Amazon DSP advertisers to access premium Microsoft inventory and curated deals that combine Amazon shopping insights with Monetize’s supply. Publishers using Microsoft Monetize gain expanded access to high-quality demand and improved monetization efficiency.
What they’re saying. Microsoft emphasized its commitment to supporting customers through the transition, calling Amazon DSP a “natural fit” for advertisers seeking scale, performance, and transparency.
Bottom line. As Microsoft pivots toward conversational and AI-driven advertising, this collaboration offers advertisers a seamless bridge between the two platforms.
OpenAI is turning ChatGPT into more than a chatbot. The company’s latest update lets users access third-party apps (e.g., Spotify, Canva, Zillow, Expedia) directly inside conversations.
Why we care. This shift could make ChatGPT a high-intent marketing channel, reaching 800 million users where they’re already engaging – in chat.
How it works. Users can call apps by name (“Figma, turn this sketch into a diagram”) or ChatGPT can suggest them automatically based on context – like surfacing Booking.com when discussing travel.
Apps act like actions, not separate tools, within ChatGPT’s conversational flow.
OpenAI calls this a step toward a “conversational operating system” – one interface where users access software and services.
What they’re saying. Nick Turley, ChatGPT’s product lead, described the goal:
“If we can evolve ChatGPT the right way… maybe you’ll be spending a lot of time in ChatGPT. But it won’t feel like you’re in a chatbot.”
The marketing angle. This platform play could turn ChatGPT into a valuable discovery and conversion engine for brands. Some potential upsides for marketers:
Massive reach: 800 million users. Apps appear at the moment of intent.
Contextual discovery: Brands show up naturally as users describe needs — not through search.
Interactive experiences: The Apps SDK supports visuals and dynamic UIs. Zillow can show listings with maps; Canva can design in chat.
What’s next. More brand integrations are coming – including Target, Uber, Peloton, and Instacart. Developers can start building now; OpenAI plans an app store and publishing reviews later this year. Apps meeting high design and functionality standards will get higher visibility.
OpenAI also hinted at “agentic commerce” – one-click transactions powered by AI. It already tested in-chat Etsy shopping for U.S. users and is hiring for ad tech and attribution tools.
The big picture. This evolution opens new paths to brand interaction, contextual engagement and possibly commerce – all within a single, AI-mediated interface. This gives marketers a chance to meet consumers where they act, not just where they search.
http://dubadosolutions.com/wp-content/uploads/2017/05/dubado-logo-1.png00http://dubadosolutions.com/wp-content/uploads/2017/05/dubado-logo-1.png2025-10-07 17:13:142025-10-07 17:13:14ChatGPT becomes a platform for apps – and maybe marketing
Google is piloting a Labs section inside the Google Ads platform, giving advertisers a centralized place to access early-stage experiments and tools.
Why we care. Instead of quietly testing features across scattered accounts, Google appears to be consolidating experiments into a single hub, making it easier for advertisers to discover and try new features before full rollout.
Details.
The Labs section shows up at the bottom of the left-hand menu in select accounts.
Early tests include features like “missed growth opportunities,” which highlight potential campaign improvements through bid or budget adjustments.
Not all advertisers will have access to it.
First seen. The update was first spotted by Vojtěch Audy, PPC specialist at Na volné noze.
https://i0.wp.com/dubadosolutions.com/wp-content/uploads/2025/09/labs-google-ads-r8l28t.png?fit=1280%2C720&ssl=17201280http://dubadosolutions.com/wp-content/uploads/2017/05/dubado-logo-1.png2025-09-30 16:52:412025-09-30 16:52:41Google Ads tests new Labs hub for experimental features
Google is enhancing Meridian, its open-source Marketing Mix Model (MMM), to help marketers make smarter, more precise budget decisions.
Why we care. Understanding ROI across channels is increasingly critical. These new Meridian updates allow for a more precise understanding of what drives sales, factoring in both media spend and non-media variables like pricing and promotions.
The big picture. Here’s what’s new:
Non-media variables: Marketers can now include pricing, promotions, and other business levers to measure their impact on sales more accurately.
Channel-level contribution priors: New features let you guide the MMM with your own business knowledge, improving the relevance of insights.
Longer-term effects: Enhanced binomial adstock decay functions track how upper-funnel media influences purchases weeks later.
Optimized spending: Marginal ROI (mROI) priors help pinpoint where the next dollar delivers the highest return, using past campaign performance.
Support and community. Meridian now has 30 certified global partners and an active Discord community, offering guidance to deploy the tool effectively and translate insights into business growth.
The bottom line: These updates aim to transform Meridian from a measurement tool to a strategic partner for budget allocation. This means Meridian can help marketers optimize spend across immediate conversions and longer-term brand impact.
https://i0.wp.com/dubadosolutions.com/wp-content/uploads/2025/09/google-chart-decline-1920-1-vHdqlf.jpg?fit=1920%2C1097&ssl=110971920http://dubadosolutions.com/wp-content/uploads/2017/05/dubado-logo-1.png2025-09-30 16:00:002025-09-30 16:00:00Google adds ROI insights to Meridian marketing mix model
The history of search offers clues about where we’re headed in the AI era – but there’s more to learn and more to do to move forward with practical steps.
AI changes the way people search. Instead of short queries that once required digging through blended results, users can now ask complex questions and get direct answers.
Much of the work to optimize for AI, though, overlaps with what SEO professionals have been doing for years. Our community is already adapting and well-positioned to take on this shift.
This article outlines practical steps to navigate the evolving landscape.
SEO, AEO, GEO: Defining the new terms
Before diving in, it’s worth addressing the terminology. It’s still new, and no single label has fully crystallized for this AI layer on top of SEO.
Two terms have gained traction:
AEO (answer engine optimization), which focuses on optimizing content so it’s chosen as the answer in AI-driven results like Google’s AI Overviews.
GEO (generative engine optimization), which describes a broader approach across generative AI platforms.
Neither feels perfect. AEO is a bit clunky, and GEO risks confusion with geography and local search.
Still, these are the terms currently in use. For simplicity, I’ll use GEO in this article.
While tools like ChatGPT are undoubtedly cool – and early adopters, myself included, are spending a lot of time tinkering – the broader story is different.
For most people, AI will matter less as a standalone tool and more through its integration into phones, web browsers, and search engines.
GEO is a new layer that sits above SEO.
Often, tools like ChatGPT will search and compile information.
These tools provide a layer of abstraction and do much of the grunt work.
They still scan the digital world and then collate that information to simplify things for the user, while search engines like Google continue to provide the best overall digital map of the world.
If traditional SEO was about matching keywords, AEO is about being the best answer – and the easiest to integrate into an AI response.
GEO: Strategic foundations
There’s no firm consensus on GEO tactics, but most of what’s recommended is simply good SEO.
That said, tactics that lost ground in the zero-click landscape may regain utility in AI Mode, where the AI does the deep dive and collates information for users.
Here are some basic strategic foundations to put in place to set yourself up for visibility in AI tools.
1. Focus on your customers
I’ve long championed bringing traditional marketing thinking into SEO, and GEO is the natural evolution of that approach.
Know your audience. Create personas, gather feedback, and define their goals, pain points, and the jobs they rely on your product or service to support.
Customer insight is key to building a customer-first strategy that helps you stand out in the age of AI.
2. Real expertise wins
The web is full of derivative content that does little to stand out. This creates a problem for AI.
Model collapse happens when AI keeps training on AI-generated content without new signals, leading to increasingly stale and inaccurate results.
The solution is what humans are still best at – fresh insights from:
Interviews.
Original research.
Proprietary data
These provide AI with something new – and worth citing.
That’s an opportunity. Have a voice, and bring something original to the table.
AI is helping people research and make purchase decisions. Your role is to be part of the discussion.
Modern SEOs are in good shape – much of what AI requires builds on the strategic SEO work we already do.
Add in some PR, social media, and content creation (which often sit under the SEO umbrella), and you’re well on your way to a functional GEO strategy.
Getting started is crucial. To stay ahead:
Create content worth quoting: Write the piece an AI (or a human journalist) would want to reference. That means clear answers, evidence, original insights, and a point of view – not filler.
Anticipate the full conversation: Don’t just answer the first question – answer the follow-ups, too. If someone asks, “How does AI change SEO?” they’ll also want to know, “What should I do about it?” Build that into your content.
Structure for machines and humans: Use headings, lists, FAQs, and concise summaries to help AI parse your work. But don’t forget the narrative depth that keeps people reading.
Diversify your discovery footprint: Don’t rely on Google alone. Research your audience, understand their hangouts, and publish in the formats and places where they ask questions today: LinkedIn, YouTube, podcasts,and industry forums. AI tools crawl all of it.
Focus on authority signals: Show the human behind the content. Add author bios, cite sources, and link to your work elsewhere. AI engines, like search engines before them, lean on trust and authority.
Experiment, measure, refine: Try different formats, test and measure how your content shows up in AI summaries, track brand mentions, and adapt. SEO has always been iterative – this new era is no different.
The opportunity in the chaos
As SEO evolves into GEO – or whatever it may end up being called – this really is the best approach.
There’s no doubt a lot of change is happening.
But much of it is part of the same gradual evolution we’ve seen before, where clicks declined and Google started answering questions directly.
AI now makes it even easier for customers to find the information they’re looking for.
They may not read it on your site – at least not initially – but the AI will, and that’s the point.
Another strength we have as SEOs is that change is constant.
If you’ve been in SEO for any length of time, you’ve lived through Panda, Penguin, Mobilegeddon, BERT, helpful content updates – the list is long (and may cause PTSD for many of us).
The key is to treat this as the next evolution. AI is being integrated into search, and it will likely become the way the masses adopt the technology.
Don’t see this as the death of SEO.
Instead, view SEO and AI (or GEO/AEO, etc.) as close, contributing partners, and evolve your plan to match the changing landscape.
Your job as a marketer is to feed these tools the information they need to point customers to you and your clients.
This shift will likely mean fewer short-term manipulations and tactical opportunities – but better results for businesses that do the basics well.
At its core, good SEO/GEO is just good marketing: understanding your customers, meeting their needs, and communicating clearly.
Amid the chaos lies opportunity. For those willing to embrace the challenge, experiment with new tools, and keep going.
That’s what we’ve always done as SEOs, which is why we’re best positioned to embrace this new world.
https://i0.wp.com/dubadosolutions.com/wp-content/uploads/2025/09/Thriving-in-AI-search-starts-with-SEO-fundamentals-4nKDbJ.jpg?fit=1920%2C1080&ssl=110801920http://dubadosolutions.com/wp-content/uploads/2017/05/dubado-logo-1.png2025-09-30 15:00:002025-09-30 15:00:00Thriving in AI search starts with SEO fundamentals
Before this, figuring out what AI Max was doing felt like looking into a black box.
Now, we can finally see the data and make smarter decisions.
Evaluating AI Max: Metrics that matter
When you’re looking at AI Max performance, start with the basics.
Open your search term tab and look for the Search terms and landing pages from AI Max option.
This lets you see AI Max results separately from your exact match, phrase match, and broad match keywords.
Compare conversion share and budget share
The first thing to check is how many conversions come from AI Max versus your regular keywords.
If AI Max is bringing in 30% of your conversions but eating up 60% of your budget, you know something needs attention.
Look at the cost per conversion, too.
AI Max might cost more at first, but that’s normal while Google learns what works for your business.
Look beyond cost – focus on conversion rates
Don’t just focus on the cost. Pay attention to conversion rates.
AI Max often finds people who are ready to buy but use different words than you expected. These new search terms can be gold mines if you spot the patterns.
One of AI Max’s biggest benefits is finding search terms you never thought to target.
Look at your search terms report and filter for AI Max queries. You’ll probably see some surprises.
Let’s say you sell running shoes and target “best running shoes.” AI Max might show your ads for “comfortable jogging sneakers” or “shoes for morning runs.”
These are people who want the same thing but use different words. The smart move is to add these high-performing terms to your regular keyword lists.
Identify irrelevant traffic
If you’re getting clicks from people searching for “cheap shoes” when you sell premium products, add “cheap” as a negative keyword.
AI Max respects negative keywords, so use them to guide the system toward higher-quality queries.
Every AI system needs time to learn, and AI Max is no different.
Plan for a learning phase
The first few weeks often look expensive because Google is still figuring out what works. Don’t panic if your costs jump initially.
Track your performance daily during the first month
As AI Max learns your patterns, you should see costs stabilize and conversion rates improve. If things keep getting worse after three weeks, then it’s time to make changes.
Keep an eye on the types of search terms AI Max finds over time
Early on, you might see lots of random queries. As the system learns, the terms should become more relevant to your business goals.
The biggest mistake people make is changing too much, too fast.
AI Max needs data to work properly, and constantly adjusting things prevents the system from learning.
That said, some problems need quick fixes.
If AI Max is spending money on completely irrelevant searches, add negative keywords immediately.
If the AI creates ads that violate your brand guidelines, remove those assets right away.
Watch your overall account performance, not just AI Max numbers.
Sometimes, AI Max might look expensive on its own, but it actually helps your other campaigns perform better by capturing different types of traffic.
Planning for the future
AI Max is still evolving, and Google keeps adding new features.
To adapt:
Build reporting systems that can grow with these changes.
Set up automated reports for the metrics that matter most to your business.
Don’t try to control everything.
The businesses seeing the best results from AI Max are the ones that:
Set clear goals.
Provide good data.
Let the system do its job.
Your role shifts from managing keywords to managing strategy.
Start testing AI Max on a small scale if you’re nervous about it. For example:
Create one campaign with AI Max enabled and compare it to your existing campaigns.
Run an AI Max for Search campaign experiment and let Google evaluate if the experiment is statistically valid.
Once you see how it works for your specific business, you can decide whether to expand.
Case studies: AI Max in action
These are early results from a limited data set and shouldn’t be viewed as statistically significant.
I’m sharing them to illustrate what I’ve seen so far – but the sample size is small and the timeframe short. Take these numbers with caution.
Case 1: Tourism and travel
This advertiser already had a solid search setup and was seeing good results.
Growth, however, was difficult because of heavy competition and the fact that strong keywords were already in play within a modern search structure.
Match type
Avg. CPC (€)
CVR (%)
AI Max
€0.11
1.47%
Broad match
€0.09
3.79%
Exact match
€0.53
9.00%
Exact match (close variant)
€0.22
7.11%
Phrase match
€0.16
6.25%
Phrase match (close variant)
€0.11
3.27%
AI Max generated additional conversions, but relative to the existing setup the impact was limited.
The conversion rate was much lower than other match types.
Because the average CPC was low, there was no cost spike, but performance still lagged.
Broad match – also known for surfacing broader, newer queries – had an even lower CPC (€0.09) and a conversion rate more than twice that of AI Max.
In this account, AI Max’s contribution was minor.
Search term overlap analysis showed AI Max had a 22.5% overlap rate, meaning 77.5% of queries were new to the campaign.
That’s a fairly good sign in terms of query discovery.
Case 2: Fashion ecommerce
This account focused on women’s clothing and already had a well-optimized campaign.
The goal was to expand reach during the competitive holiday season, when exact match keywords became increasingly expensive.
Match type
Avg. CPC (€)
CVR (%)
AI Max
€0.08
2.15%
Broad match
€0.12
2.89%
Exact match
€0.67
8.45%
Exact match (close variant)
€0.28
6.78%
Phrase match
€0.19
5.92%
Phrase match (close variant)
€0.14
4.11%
AI Max performed well here, delivering the lowest CPC at €0.08 and a respectable 2.15% conversion rate.
Although conversion rates were lower than exact and phrase match, the cheaper clicks kept cost per acquisition competitive.
AI Max also captured fashion-related long-tail searches and seasonal queries that the existing keyword set had missed.
Notably, AI Max outperformed broad match with both lower costs and better conversion rates.
This suggests its ability to better understand product context and user intent – especially important in fashion, where search terminology is diverse.
Search term overlap analysis showed only an 18.7% overlap rate, meaning 81.3% of queries were completely new.
That level of query discovery was valuable for extending reach in a highly competitive market.
Case 3: B2B SaaS
This account promoted project management software and had a mature strategy focused on high-intent keywords.
Conversion tracking was strong, measuring both MQLs and SQLs.
The client wanted to test AI Max for additional lead generation opportunities.
Match type
Avg. CPC (€)
CVR (%)
AI Max
€0.89
0.76%
Broad match
€0.72
1.23%
Exact match
€1.84
4.67%
Exact match (close variant)
€1.22
3.91%
Phrase match
€1.05
3.44%
Phrase match (close variant)
€0.94
2.88%
In this case, AI Max struggled. Despite a reasonable CPC of €0.89, the conversion rate was just 0.76%.
That pushed CPA well above the client’s target, making AI Max the worst-performing match type in the account.
It tended to capture too many informational searches from users not yet ready to convert.
Even broad match, typically associated with lower-intent traffic, outperformed AI Max with a 1.23% conversion rate at a lower CPC.
The complexity of the B2B buying cycle favored exact and phrase match keywords over AI Max’s broader interpretation.
Search term overlap analysis showed a 31.4% overlap rate, leaving 68.6% of queries as new.
However, these were mostly low-intent informational searches that didn’t align with SQL goals – underscoring the importance of high-quality conversion tracking when evaluating AI Max.
Wider industry sentiment
Advertiser feedback so far mirrors these mixed results.
In a recent poll by Adriaan Dekker, more than 50% of respondents reported neutral outcomes from AI Max, while 16% saw good results and 28% reported poor performance.
Tips to analyze AI Max search terms
You can analyze AI Max queries in Google Sheets using a few simple formulas. If your search term report has the term in column A and match type in column B:
To check whether a search term appears in both AI Max and another match type:
To count how many match types trigger a given term:
=COUNTIFS($A:$A;A2) can be used to count how many match types trigger on that search term.
To measure query length with an n-gram analysis:
=(LEN(A1)-LEN(SUBSTITUTE(A1," ","")))+1
These checks show whether AI Max is surfacing unique queries, overlapping with existing match types, or favoring short-tail vs. long-tail terms.
Because AI Max is still in early stages, it’s hard to draw firm conclusions.
Performance may improve as the system learns from more data, or remain flat if your setup already covers most transactional queries.
That’s the question advertisers will answer in the coming months as more tests and learnings emerge.
So far, results can be positive, neutral, or negative.
In my experience, neutral to negative outcomes are more common – especially in accounts with strong existing setups, where AI Max has fewer opportunities to add value.
A Google Ads script to uncover AI Max insights
To make analyzing AI Max performance easier, I created a Google Ads script that automatically pulls data into Google Sheets for deeper analysis.
It saves hours of manual work and includes the exact formulas mentioned earlier in this article, so you can immediately spot overlap rates and query patterns without manual setup.
The script creates two tabs in your Sheet:
AI Max: Performance Max search term data with headlines, landing pages, and performance metrics.
Search term analysis: A full comparison of all match types, including AI Max, with automated formulas.
The analysis covers:
Overlap detection between AI Max and other match types.
Query length analysis (short-tail vs. long-tail).
Match type frequency counts to identify competitive terms.
Automatic cost conversion from Google’s micro format into readable currency.
How to use it:
Create a new Google Sheet and copy the URL.
In Google Ads, go to Tools > Scripts.
Paste the script code and update the SHEET_URL variable.
Run the script to automatically populate your analysis.
With this setup, you can quickly calculate the same metrics I used in the case studies – like the 22.5% overlap rate in the tourism account or the 81.3% new query discovery in fashion.
The automated workflow makes it easier to see whether AI Max is surfacing genuine new opportunities or simply redistributing existing traffic.
https://i0.wp.com/dubadosolutions.com/wp-content/uploads/2025/09/LinkedIn-poll-AI-Max-performance-sT75AQ.png?fit=545%2C393&ssl=1393545http://dubadosolutions.com/wp-content/uploads/2017/05/dubado-logo-1.png2025-09-30 14:00:002025-09-30 14:00:00AI Max in action: What early case studies and a new analysis script reveal
Google AI Mode is getting more visual by providing a more graphical response to some of your queries, including your shopping search queries. Google can do this by using its new visual version of the query fan-out technique it has used with AI Overviews and AI Mode.
Google AI Mode, for certain queries, particularly those related to shopping, will respond with images and graphics. These are aimed at sparking inspiration, Robby Stein, VP of Product Management at Google Search, told Search Engine Land.
AI Mode will be able to not just understand your query in a text-based manner, but also understand your query visually and respond with both textual and visual responses. It is a new and updated fluid, ongoing conversation in AI Mode that “sparks inspiration,” Google said.
Visual search fan-out technique. Google’s fan out technique is able breaking down your question into subtopics and issuing a multitude of queries simultaneously on your behalf. Now, Google can also do this visually by looking at the image and text query input for query analysis and various image region analysis, including meta data and context around the image. Then Google AI Mode can render a visual grid of responses to your query.
“This means AI Mode can perform a more comprehensive analysis of an image, recognizing subtle details and secondary objects in addition to the primary subjects – and then runs multiple queries in the background. This helps it understand the full visual context and the nuance of your natural language question to deliver highly relevant visual results,” Google wrote.
AI Mode is more visual. Now when you search for some queries in AI Mode, the responses will be much more graphical and visual, right up front. Yes, AI Mode may have responded with images before, but now the images are higher and more prominent in some of the responses.
Plus, you can conduct follow up questions on the visual responses.
“You’ll see rich visuals that match the vibe you’re looking for, and can follow up in whatever way is most natural for you, like asking for more options with dark tones and bold prints. Each image has a link, so you can click out and learn more when something catches your eye. And because the experience is multimodal, you can also start your search by uploading an image or snapping a photo,” Google added.
Shopping in AI Mode. Lilian Rincon, VP Product Management Google Shopping, told us one of the best places for a visual experience is with shopping in Google Search. A more visual AI Mode helps you shop conversationally with fresh shopping data, can lead to a better shopping experience.
With the addition of Google’s Shopping Graph of 50 billion product listings, where 2 billion products are refreshed daily every hour, the responses are not just inspirational but detailed and helpful.
AI Mode for Shopping can not just give you ideas on what to put in your living room but also help you find the perfect article of clothing, in your color, style and fit.
Here is a video of this in action:
More details. This is launching today in Google AI Mode in the US in English. These are free listings, not shopping ads, and currently have no paid model, including no affiliate model. While Google has ads in AI Overviews, Google is only experimenting with ads in AI Mode, and there are no more details on ads right now for this experience.
Agentic experiences, like helping you buy and find what your looking for, is here for some areas now in Search Labs. But Google said they want the final purchase to happen directly on the retailer’s site.
Why we care. A new, more visual and graphical experience in AI Mode, may be a better search experience for some searchers and for some queries. Google is experimenting with a lot of changes to Google Search and is rapidly trying new interfaces and technologies.
https://i0.wp.com/dubadosolutions.com/wp-content/uploads/2025/09/google-robot-dressing-room-1920-IOZ8po.jpg?fit=1920%2C1097&ssl=110971920http://dubadosolutions.com/wp-content/uploads/2017/05/dubado-logo-1.png2025-09-30 13:00:002025-09-30 13:00:00Google AI Mode gets more visual, including inspirational shopping responses
