ChatGPT Ads: Your Brand's New Battleground

2007: “Mobile ads will never work. Screens are too small.”

2015: Mobile ad spend passed desktop.

2025: Ads arrived in ChatGPT.

We’ve been here before. And the marketers who moved early won. I watched the mobile version of this play out in real time, and the pattern’s repeating, only faster. This piece is part of how I think about AI in advertising: where the budget moves next, and who gets there before the price spikes.

(Video: ChatGPT Ads are Coming. Here’s What To Expect, MarketingAlec on YouTube)

Deja vu: the mobile playbook

Remember the shift from desktop to mobile? Here’s how it actually went:

A new platform shows up.
Users drift over slowly, then all at once.
Ad formats chase the new behavior.
Early movers grab market share cheap.
Latecomers pay a premium for the same ground.

That’s the script. We’re running it again, and the cast list hasn’t changed. The only thing that’s different is the speed.

What’s actually happening

The facts, as of now:

ChatGPT ads have gone from quiet test to real platform. OpenAI runs a self-serve Ads Manager (still in beta), CPC bidding, and placements that are open to advertisers in the US.
Around 800 million weekly users. That’s already bigger than most platforms you currently buy on.
The ads are conversational, woven into the answer instead of bolted to the side as banners.

The mobile parallel is the part that should make you sit up. When the iPhone launched, it took about 500 days to reach a million users. ChatGPT did it in 5. This transition is running on a different clock than mobile did, and the budgets are following.

The three phases (and where we are)

Phase 1: Denial (2023 to 2024)

“It’s just a fad.”

Marketers waved off large language models and ChatGPT.
Budgets stayed parked on Google and Meta.
Mobile equivalent: 2007 to 2009.

Phase 2: Experimentation (the recent past, and for some teams, still now)

“Maybe we should test this so we’re ready.”

Sharp brands started building real AI presence.
AI ad placements began rolling out.
This is the 2010 to 2012 moment, when Instagram and mobile ads showed up.

Phase 3: Domination (2026 and on)

“How did we miss this?”

AI-native brands own their categories.
Costs climb fast for the teams showing up late.
Mobile’s version was 2013 on, when it was mobile-first or nothing.

We’re stepping into Phase 3 right now. The teams that did the Phase 2 homework are about to look very smart.

The mobile playbook, updated for AI

What worked then maps almost cleanly onto what works now:

Mobile (2010): build a mobile site. AI (2025 and 2026): get cited in ChatGPT’s answers.
Mobile (2011): buy early Facebook mobile ads. AI (2026): test ChatGPT ads while they’re still cheap.
Mobile (2012): ship an app. AI (2026): build AI-optimized content hubs.
Mobile (2014): go mobile-first. AI (2027): go AI-first.

5 actions for early movers

Action 1: Audit your AI brand presence (this week)

The same way you Googled your own brand back in 2008, you now need to see what ChatGPT says about you. Open ChatGPT (or Claude, or Gemini) and ask it to check the web for recent reviews, discussions, and comparisons of your brand. Then read the answer the way a customer would, and ask:

How am I positioned against my competitors?
What are my top three strengths in its answer?
Which criticisms come up most often?
Do I show up in the “best of” comparisons?

Reality check: if ChatGPT doesn’t know you exist, you’re invisible to the roughly 800 million people who now ask it before they ask Google.

Action 2: Build your “mobile-friendly” content (the 2026 version)

Remember what mobile forced us to do? Simplify navigation for thumbs. Speed up load times. Cut content down to something you could read on a phone.

The AI version of that work looks like this:

Write detailed guides and video. Those are the formats AI pulls from most.
Use clear structured data and clean layouts. That’s how AI reads you.
Publish unique, data-rich case studies. AI loves to cite specific numbers.
Distribute on Reddit, YouTube, and social. That’s where AI does a lot of its learning.

Action 3: Reserve your test budget

The early mobile advertisers carved out 5 to 10 percent of their digital budget for mobile testing, ran A/B tests of desktop versus mobile, and started with high-intent, proven keywords.

Run the same play now:

Set aside 5 to 10 percent of your Google Search budget for AI ad testing.
Run parallel campaigns: Google versus ChatGPT.
Start with bottom-funnel queries (product comparisons, “best X for Y”).

Here’s a short version you can send up the chain to your CFO, CEO, or CMO:

We missed the early mobile ad wave back in 2010. Our mobile CPCs are now several times what the early adopters paid. ChatGPT has around 800 million weekly users and live, self-serve ad placements. I want 7 percent of our search budget to test before the costs spike. This is the mobile moment again, only faster.

Action 4: Rethink creative for conversation

A few things killed brands during the mobile shift: desktop sites that never adapted, ads built for a mouse instead of a thumb, and a flat refusal to deal with the UX change at all.

The AI version of those mistakes:

AI slop dropped into a conversational space.
Keyword-stuffed copy where natural language should be.
Ignoring how people actually talk to AI when they research you.

Watch how the format moved:

Desktop-era ad: “Award-Winning CRM Software. Trusted by 10,000+ Businesses.”

Mobile-era ad: “CRM for Small Teams. Free Trial.”

AI-era response: “Unlike Salesforce, we’re built for teams under 50. You can set up in 10 minutes without IT, and our customers say onboarding is much faster than HubSpot.”

See the shift? Conversational. Specific. Comparative. That’s the creative job now.

Action 5: Build attribution now (before it turns into chaos)

The mobile attribution mess is worth remembering. Customers browsing on their phone, buying on a laptop, and cross-device tracking that took years to untangle. Don’t repeat it.

Set this up now:

UTM parameters specific to AI-assistant traffic.
A post-purchase survey question: “How did you find us?” with “AI assistant” as a real option.
Multi-touch attribution. ChatGPT will rarely be the last click, the same way mobile browsing rarely was.

A survey question you can drop in today:

How did you first learn about us?

Google Search

Social media

AI assistant (ChatGPT, Gemini, Claude, etc.)

Friend or colleague

Other

The mistakes that cost millions

The excuses that sank brands during the mobile transition still sound familiar:

“We’ll wait and see.” They went under while they waited.
“Our customers aren’t on mobile.” They were, and they were buying from a competitor.
“Mobile ads don’t convert.” They did. The desktop ads run on mobile screens didn’t.
“Too expensive to rebuild.” Losing the market cost a lot more.

Don’t run that play again with AI:

Test with 5 to 10 percent of budget now.
Build AI-native content today. It can take months to get picked up.
Train your team on conversational formats this quarter.
Track your competitors’ AI moves monthly. They’re already testing.

The window

The mobile transition took roughly eight years, from 2007 to 2015. The AI version is running on a tighter clock, and we’re well into it. You’ve got two real options:

Move now, with a test budget and the early learnings that come with it.
Pay several times more in a couple of years, after the window closes.

Most CMOs missed mobile. They said the screens were too small, their customers weren’t there, they’d wait for the data. That call was expensive. You don’t have to make the same one with AI.

If you want the bigger picture of where AI fits across your whole marketing function, that’s the AI marketing map. This piece is the advertising slice of it.

This week’s challenge

Send your paid media lead this article (two minutes).
Sketch your plan.
Get a test budget approved by your CFO, CEO, or CMO.

That’s it. A real head start.

The marketers who win in 2027 are the ones moving on AI right now. Be one of them. I’m rooting for you.

Cheers, Alec

Don’t overpay for the AI ad rush

ChatGPT ads are live and the early-mover window is open right now. The teams testing this week are the ones who won’t be paying premium CPCs in 2028. Every Friday I send the real test results, the ad formats that are actually landing, and the moves worth copying before the costs climb. Want the playbook while it still buys you a head start?

Subscribe free and get the next issue.