The Great Search Shift, Episode 3: Agents, Gemini, and the ChatGPT ads opportunity

In Episode 1 we looked at where search has been. In Episode 2 we looked at where it is right now. Episode 3 is where Phillip Thune and Neil Patel stop hedging and make their calls.

Predicting five years out in search right now is, as Phillip puts it, “insane.” But looking to the end of 2026 or the next 12 months? That’s a manageable window. And the patterns forming right now are directional enough to act on.

Here’s what they think is coming.

The AI agents conversation tends to focus on ChatGPT and OpenAI. But Neil’s prediction for the biggest shift by the end of 2026 isn’t about OpenAI at all.

The corporate world is already spoken for

Google and Microsoft have something OpenAI doesn’t: penetration into the corporate world at scale. Every company running G Suite is already a Google customer. Every business on Microsoft Outlook, Word, and Excel is already in the Microsoft ecosystem.

Neil gives a telling example from a conference he attended. A publicly traded real estate company was paying ChatGPT for customizations. OpenAI’s response: we don’t do this unless you’re spending at least $10 million a year. Then Google came in. The company asked about costs. Google’s answer: nothing. You’re already on G Suite, we’ll take care of it.

That’s the structural advantage in plain terms. Google can offer AI agents for free because it already has the relationship and the revenue. OpenAI has to charge because it doesn’t.

Free today, monetized tomorrow

The playbook is familiar. Offer it for free, build the installed base, then monetize. It’s exactly what Microsoft did with Teams. Teams wasn’t necessarily the best product, but it was free for existing Microsoft customers. It won enterprise adoption anyway.

Google and Microsoft are running the same play with AI agents right now. Organizations are going to accept it because they’re already integrated. And by the time those companies want to switch, the switching costs will be too high.

For paid search managers, this matters because the agent layer is going to be built on top of the platforms your customers already use daily.

This is the prediction that generated the most discussion in Episode 3. Phillip’s take going in: yes, within 12 months. Neil’s response: not quite.

24 to 36 months, not 12

Neil’s view is data-driven. Gemini’s growth is fast, but not fast enough to surpass ChatGPT’s monthly active users within 2026. The charts don’t support it. They’d be bleeding into Q2 2027 at current trajectories.

But in 24 to 36 months? Both Phillip and Neil agree Gemini gets there. The reasons stack up:

Google won the Apple deal. Gemini is now the LLM of choice on Apple devices, a distribution advantage that would have changed the competitive picture entirely if OpenAI had secured it. It already has Android. It prints $140 billion a year in profit and doesn’t need debt to fund its data centers. It can outspend, outlast, and out-distribute almost everyone.

As Neil puts it: “At some point in the future, I’m pretty confident. Gemini.”

If Gemini grows and AI Mode becomes a primary way people interact with Google Search, does that mean paid search managers need to learn an entirely new platform?

Consolidation is coming. And that’s good news.

Phillip’s prediction here is straightforward: on Google products at least, it all gets consolidated. If you have an AI Max campaign running today, you’ll show up wherever Google’s AI surfaces show up. You won’t have to rebuild from scratch or learn a new buying interface.

Why would Google do it any other way? Creating a new product means convincing existing advertisers to sign up for something new, re-selling them, and risking losing their budget in the transition. Consolidating means more revenue and less friction. The incentives point one way.

The practical implication: advertisers who invest in Google’s AI-driven campaign formats now, PMAX and AI Max, are likely positioning themselves to automatically extend into Gemini inventory as it scales. Keep an eye on how AI Overviews are already affecting paid visibility to see where that’s heading.

With the ChatGPT ads beta threshold dropping from $200K to $50K as of May 2026, and set to open up further through partnerships like Criteo, the channel is becoming accessible to a much broader set of advertisers.

Why early movers win

Neil’s advice here is consistent with everything he’s ever said about new ad platforms: get in early, because that’s where the arbitrage is.

When a platform with a large user base releases advertising, early entrants face less competition. That means lower CPCs, lower CPAs, and learnings that compound over time. When 50 or 100 competitors show up later, costs go up and the advantage narrows.

The math works differently for large and small companies, but the logic applies to both. Larger brands have the budget to compete when things get expensive, but they’re leaving money on the table by waiting. Smaller brands have the most to gain from early entry: lower acquisition costs mean more cash flow to invest and grow.

Adthena has now tracked over a thousand advertisers inside ChatGPT ads. The channel is real, the inventory is growing, and the brands showing up consistently are building an early-mover advantage that will be hard to replicate later. AdBridge makes it straightforward to get your existing Google Ads campaigns live on ChatGPT in minutes.

Phillip’s final question to Neil: by end of 2026, is ChatGPT ads still a CPM and CPC world, or does it consolidate toward something else, maybe CPA?

The honest answer: nobody knows yet

Neil’s answer is characteristically honest: it’s too early to call. The platform is two and a half months into running ads. The buying models are still being tested. CPA would be the logical evolution, it’s what advertisers ultimately want, but the infrastructure to support it at scale takes time to build.

What’s more predictable is the direction: toward performance, toward automation, toward the kind of consolidated buying interface that makes it easier for advertisers to spend and easier for OpenAI to grow revenue. Watch how ChatGPT ad placements evolve over the next quarter for the clearest signal of where it’s going.

Three episodes in, the picture is clear enough to act on.

Google and Microsoft will win the agent race. Not because they have better products, but because they have better distribution and deeper corporate relationships. Gemini will overtake ChatGPT, just not by the end of this year. Google’s paid products will consolidate across AI surfaces, so your existing campaigns are likely already future-proofed. And ChatGPT ads are open, accessible, and still in the arbitrage window.

The paid search predictions for 2026 aren’t all that complicated. The brands that get in early, stay curious, and keep humans in the loop will be the ones writing the playbooks in 2027.

Watch episode three of The Great Search Shift or catch up on previous episodes.