Somewhere right now, ChatGPT is pitching your product to a potential customer. You didn’t write the pitch. You don’t know what it’s saying. And there’s a good chance it’s wrong.
Hi, I’m Jeff Payne. You’re listening to The Jeff Payne Show, Episode #39: AI Is Your #1 Product Marketer.
Here’s what’s changed in the last few months. Profound, one of the firms tracking AI referral data, found that ChatGPT traffic to company websites jumped more than 60% starting in early May — and it’s held. Even more telling: homepages went from about 3.5% of that referral traffic to nearly a quarter of it.
Translate that. People are asking AI who to buy from, AI is sending them somewhere, and increasingly, that somewhere is your homepage. Not a review site. Not a comparison blog. You.
Most of you have already figured out the first part of this — you’re tracking whether AI mentions you at all. That’s table stakes now. The real game, the one almost nobody’s playing yet, is different: it’s not “does AI talk about me.” It’s “Does AI talk about me correctly, and in the way I’d want?”
Because right now, for most companies, the answer is: AI is guessing.
Here’s the uncomfortable part. AI doesn’t read your whole website and form a balanced opinion. It skims. It weighs certain real estate more heavily than others — your headers, your navigation, your footer — the same way it treats a bold subhead as more important than a buried paragraph three pages deep. What you put in the prominent structural spots on your site is disproportionately what AI repeats back to prospects.
That means the companies actually winning this aren’t the ones with the most content. They’re the ones who’ve figured out where AI is looking and put the right things there on purpose.
Three examples worth studying.
First — Rippling. When the industry scandal broke between Rippling and a competitor over an alleged corporate espionage attempt, Rippling didn’t wait for TechCrunch or Reuters to break the story. They published it themselves, first, on their own blog. Full statement, full framing, their words.
Result: Google indexed Rippling’s version as the original. Every outlet that covered it afterward linked back to Rippling’s page. Every future AI answer about that story pulls from a page Rippling controls. They’ve repeated this move across funding rounds and product launches—they’ve essentially built their own newsroom, which means that when AI tells that story, it’s telling Rippling’s version.
Second — Clay, the sales intelligence company. They put their best customer logos and case studies directly in the website footer — the one spot that shows up on every single page. Someone tested this by asking ChatGPT, “What notable brands use Clay?” Every source the AI cited traced back to that footer. Then they ran the identical question for competitors with just as many customer stories — but buried deeper in the site. Those companies got cited far less consistently. Same proof. Different placement. Different result.
Third — Lovable, the AI coding platform. Their pricing model is genuinely confusing—it’s usage-based and credit-driven. Instead of explaining it in a wall of paragraphs, they rebuilt their FAQ with screenshots paired with simple tables: here’s the request, here’s the work, here’s the cost. The screenshots are for human readers. The tables are for the AI quoting it — because language models overwhelmingly prefer to cite something structured over something buried in prose.
Three different companies. Three different moves. Same underlying principle: they didn’t leave it to chance where AI would look, or what it would find when it got there.
You don’t need a content overhaul to start. Pick one thing this week. Look at your own footer — is your best proof there, or is it three clicks deep? Look at your FAQ — if you have anything complicated to explain, could a table do it better than a paragraph? And if something newsworthy happens in your business, publish it on your own site first, in your own words, before anyone else gets to frame it for you.
There’s a longer list of moves worth testing, too — comparison pages that position you honestly against AI tools your prospects might build themselves instead of buying, glossary pages that map out the terms your industry lives on, even free tools that expand what AI associates with your brand. Different lift, same idea: stop hoping AI finds the right story. Build the story where AI is already looking.
Your website used to be for people. Now it’s briefing material for a machine that’s going to speak on your behalf, whether you’ve prepped it or not. The companies winning this aren’t smarter. They just decided that mattered before everyone else did.
Somewhere right now, ChatGPT is pitching your product to a potential customer. You didn’t write the pitch. You don’t know what it’s saying. And there’s a good chance it’s wrong.
That’s not a hypothetical. Profound, one of the firms tracking AI referral data, found that ChatGPT traffic to company websites jumped more than 60% starting in early May 2026 — and it’s held. Even more telling: homepages went from about 3.5% of that referral traffic to nearly a quarter of it.
People are asking AI who to buy from, AI is sending them somewhere, and increasingly, that somewhere is your homepage. Not a review site. Not a comparison blog. You.
Most companies have already figured out the first part of this shift — they’re tracking whether AI mentions them at all. That’s table stakes now. The real game, the one almost nobody’s playing yet, is different: it’s not “Does AI talk about me?” It’s “Does AI talk about me correctly, and in the way I’d want?”
Because right now, for most companies, the answer is: AI is guessing.
AI doesn’t read your site. It skims it.
Rippling built its own newsroom. When an industry scandal broke involving an alleged corporate espionage attempt against Rippling, the company didn’t wait for TechCrunch or Reuters to break the story. They published it themselves, first, on their own blog — full statement, full framing, their words. Google indexed Rippling’s version as the original. Every outlet that covered it afterward linked back to Rippling’s page. Every future AI answer about that story pulls from a page Rippling controls. They’ve repeated the same move for funding rounds and product launches.
Clay put its best proof in the one place every visitor sees. The sales intelligence company links its top customer case studies directly in the website footer — the one spot that appears on every page of the site. When someone tested this by asking ChatGPT, “What notable brands use Clay?”, every cited source traced back to that footer. The same question, asked about competitors with just as many customer stories — but buried deeper on their sites — returned far fewer consistent citations.
Lovable rebuilt its FAQ for two audiences at once. Their pricing model is genuinely confusing — usage-based, credit-driven. Instead of explaining it in a wall of paragraphs, they paired screenshots with simple structured tables:
| User Prompt | Work Done | Credit Cost |
|---|---|---|
| “Build a login page” | Component generations + styling | 4 credits |
| “Fix responsive layout bug” | Debug + patch across breakpoints | 2 credits |
| “Add Stripe integration” | API wiring + testing | 8 credits |
The screenshots are for human readers. The tables are for the AI quoting it — language models overwhelmingly prefer to cite something structured over something buried in prose.
They didn’t leave it to chance where AI would look or what it would find when it got there.
The Move
You don’t need a content overhaul to start. Pick one thing this week:
There’s a longer list of moves worth testing, too — comparison pages that position you honestly against AI tools your prospects might build themselves instead of buying, glossary pages that map out the terms your industry lives on, even free tools that expand what AI associates with your brand. Different lift, same idea: Stop hoping AI finds the right story. Build the story where AI is already looking.
The companies winning this aren’t smarter. They just decided that mattered before everyone else did.
Your website used to be for people. Now it’s briefing material for a machine that’s going to speak on your behalf, whether you’ve prepped it or not.
COMPLETE THE FORM TO
BOOK A STRATEGY CALL
"*" indicates required fields
CHAPTERS:
COMPLETE THE FORM TO
BOOK A STRATEGY CALL
"*" indicates required fields
Subscribe and Share – WE APPRECIATE YOUR SUPPORT
