As conversational search gains traction, the bigger question isn’t who has more users, but who can monetize them.
Google enters this phase with a massive advantage: mature ad systems, deep advertiser adoption, and decades of optimization. Early AI Mode signals point to a measured rollout.
The panic phase is over
After a period of panic within the company, Google’s built-in advantages, coupled with massive capital expenditures, have helped it regain ground on category leader ChatGPT in LLM search.
In December 2025, Google’s own code red became OpenAI’s code red.
The dust will continue to settle, and analysts have different takes. But one signal stands out: in a major validation, Apple has chosen Google to power its own AI.
It was perhaps premature to assume Google Search would simply lose to ChatGPT on product. That was the consensus at the start of 2025. Google shares fell about 30% from peak to trough before rallying 130%. Today, the company is valued at roughly $3.6 trillion, just behind Apple.
The SEO toolkit you know, plus the AI visibility data you need.
Why monetization will decide the winner
Why did Google’s recent progress in LLM conversational queries — in the form of AI Overviews and AI Mode — have such a large impact on the company’s valuation in such a short time?
Ultimately, it comes down to visibility of financial projections. In a company with so much to defend, Google’s CFO and leadership team needed to determine whether shifts in user behavior — in how search works and how it makes money — would weaken the business model or reinforce it.
Net-net: Google before the shift: huge. Google after the shift: ditto.
Visibility — in the sense of financial planning, not in the SERP — means a great deal to Google’s advertisers, too.
A large proportion of your annual digital advertising budget is likely allocated to Google. You also still care about how you appear in organic results and increasingly, how your company appears in AI Mode, ChatGPT, Claude, and similar environments.
“I’m fine with 30% less of my business coming in from Google, and figuring out lots of complicated ways to replace it,” … said no advertiser ever.
How monetization will play out in AI search
The competition between monetization models in LLM conversations — especially between the two leaders, ChatGPT and Google’s AI Mode — will play out differently from the broader race for overall user share. There are several moving parts to keep an eye on:
- Overall assumptions about ad formats and “how to monetize.”
- Pace of rollout.
- Whether users and public opinion recoil at ads.
- Advertiser success rates based on performance measurement.
- Advertiser adoption, including adoption by the agency ecosystem.
- Platform targeting options.
- Advantages of fuller-funnel ad journeys and data collection.
- Privacy, safety, policies, and enforcement.
- An all-encompassing consumer brand vs. a better mousetrap.
- And a few other factors.
Right now, OpenAI is at a critical moment because it’s still so early in its monetization. It’s still testing an inefficient auction model confined to a small group of large advertisers. (Some ads, from their pilot, spotted here.) It may be some time before more mature tools and reporting emerge.
Most recently, OpenAI brought ad platform Criteo (often used for retargeting) on as a partner. The Trade Desk, the world’s largest non-Google DSP for programmatic, is also in the mix. Some observers have speculated about deeper partnerships or even an acquisition of The Trade Desk, though that seems unlikely.
In any case, outsourcing inventory to programmatic partners is a pragmatic step in OpenAI’s monetization strategy. It also underscores how early the company is in building a scalable ads business.
Despite a broad rollout with partners, OpenAI is stepping back from “checkout in chat” integrations after limited adoption from both merchants and consumers. When your primary competitor has a 25-year head start, the learning curve is steep.
So does it make sense now for advertisers to lean into evolving Google user behavior and figure out how to ride the wave?
AI Mode considerations for Google advertisers
Expect the transition to more AI Mode sessions — and eventual monetization — to be smoother than initially anticipated. If you’re an advertiser, AI Mode need not equal panic mode.
How do these LLM sessions look to users? Obvious to you and me, but likely less so for many searchers.
Depending on how you search, AI Overviews may appear above other results on the SERP. That’s becoming a natural extension of Google Search sessions.
But that’s not the real conversational layer. The LLM workflow happens in AI Mode. How often users go there remains to be seen.
It’s improving quickly. Unlike ChatGPT, Google AI Mode downplays how it finds information, whether it is “reasoning,” and which model is being used. The experience feels relatively seamless.
It’s still early, but ads are already appearing in some cases. The key question is how this evolves, and what advertisers should be paying attention to.
The key areas to watch are:
- Extent of monetization.
- Different ways to monetize.
- Advertiser control and campaign types.
- Reporting.
- Funnel stage.
1. Extent of monetization
AI Mode is in a popularity contest and a price war with ChatGPT. Google will likely try to grind down competitors in LLM conversations by monetizing lightly and gradually. Perplexity and Anthropic, for their part, are completely shunning ads.
The result will be less ad volume in this space than you might expect. It may also increase the commercial value of organic visibility in LLM-driven results, leading to renewed focus on content and reputation fundamentals.
Forget ad campaign FOMO, then. It will be interesting to place ads alongside AI-driven sessions, but don’t break the bank. Implement, watch, and learn at your own pace.
2. Different ways to monetize
Experienced advertisers know there are a few ad formats to consider in any situation like this. The main ones would be: text ads triggered by keywords or similar signals, in a reasonably native format, and feed-based Shopping type ads.
Another way to make money is to allow direct checkout — to take a cut of transactions. As noted above, OpenAI is backtracking on this approach, though not eliminating it entirely. How important it will be for Google merchants (and Google itself) remains to be seen.
Google’s experience likely allows it, again, to play the long game, study the data, and bring partners and advertisers along for the ride, on an impressive scale.
Recently, Loblaw inked an integration deal with OpenAI. A week later, it made a similar deal with Google.
3. Advertiser control and campaign type
In terms of execution, we’ll want to be on the lookout for which kinds of campaign types in Google Ads make your ads eligible to show in AI Mode.
You can learn everything you want about how ads will show in AI Overviews in Google’s help files. Unsurprisingly, text and shopping campaigns from Performance Max, standard shopping, and keyword campaigns make your ad eligible to show in AI Overviews.
Google says less about AI Mode in its documentation, for now.
Our agency recently received a Google deck outlining a “Shopping Expansion” beta. There’s little mention of AI Mode, though one table, in a subtle way, refers to both AI Overviews and AI Mode.
My expectation is that Google will gradually ease users into AI Mode and test ads sparingly. Even if ads appear in a small share of sessions — say 0.5% — that will still generate significant data and feedback.
Advertiser control will likely be even more limited than it is today. In the world of feed-based ads, you have some levers, but the massive machine learning that controls matching is held by Google and the real-world behavioral ecosystem.
To a lesser extent, that’s also how keyword matching works. Micromanagers won’t be too comfortable, but the impact of the ads could still be powerful, especially with data-driven attribution.
Here’s hoping new signals, new reporting breakouts, and new levers become available to advertisers. Namely: audiences including cool personas; demographics; novel larger buckets around life stages; novel characteristics we haven’t even dreamt of yet, such as their language ability level or aspects of how they interact with the LLM.
4. Reporting
The real question is: will reporting be transparent and insightful? We need to at least be able to look at all available metrics for ads that showed in AI Mode specifically. Time will tell.
Microsoft seems to be the first out of the gate with AI-conversation-specific reporting breakouts. We expect no less from Google and are impatiently awaiting further guidance on this front — primarily on what kind of reporting will be directly available in the Google Ads interface.
It would be easy for the casual observer to blindly believe that somehow, you’ll never be eligible to show up in AI Mode or AI Overviews unless you adopt certain Google Ads campaign types. There’s a lot of rhetoric around AI Max.
I’d advise advertisers to do their own research and run their campaigns to suit themselves. Hint: AI Max isn’t the only magical gateway to AI-using users and might not even be a good or appropriate one for many advertisers.
Once reporting is beefed up, you’ll want to know how well the AI-specific inventory is doing, however your campaigns wind up serving there.
5. Funnel stage
But that leads us to a wrinkle. Although ads appearing astride AI Mode conversations could certainly be low-funnel (think Shopping ads in high-intent situations), much of the opportunity here is thematic. Your company may now enjoy new opportunities to associate itself with higher-order thinking, new audience definitions, and new intent characteristics.
This opportunity probably comes to your door dressed up as “lower ROAS.” It may be tempting, therefore, to shy away.
That’s a mistake.
Why?
Like what happened when everyone started using mobile phones, that’s where the consumer will be. Ugly early numbers shouldn’t blind us to the imperatives associated with scale.
Track, optimize, and win in Google and AI search from one platform.
When the funnel moves, everything moves
Midsized to larger advertisers should step back and reimagine how they approach growth and market impact. There are meaningful opportunities for companies to align more closely with their audiences.
This has little to do with AI Max, and everything to do with how LLM-driven research works. Compare how publishers have traditionally assembled consumer personas — often from fragmented behavioral signals — with the much richer context that can emerge from ongoing interactions with an LLM.
A net shift up-funnel could follow. Imagine a world where a significant share of Google search sessions takes place within conversational experiences. Your ads will need to show up there, where appropriate. If that happens, your funnel — and your competitors’ — will move with it.
Will you be ready?
https://searchengineland.com/ai-mode-google-next-ads-engine-471967