OpenAI is quietly turning ChatGPT into a new ad surface—and the next step isn’t “better targeting.” It’s a new format: ads that behave like the product itself.

On April 1, OpenAI signed Smartly as its first creative adtech partner, with the explicit goal of evolving ChatGPT ads from basic sponsored placements into interactive, conversational ad units users can engage with inside the assistant. Business Insider reported that Smartly will initially help brands optimize ChatGPT ads in real time, but the longer-term ambition is “ads that can chat.”

This matters because LLM interfaces collapse the funnel: users aren’t browsing; they’re asking, comparing, deciding. And OpenAI says its early ad pilot is already material—on track for $100M in annualized recurring ad revenue and working with 600+ advertisers. Business Insider

1) Conversational ads change the “unit of marketing” from a click to a dialogue

In classic search, the unit is a keyword and a click. In social, it’s a thumb-stopping impression. In ChatGPT, the unit is a conversation state: what the user has already said, what constraints they’ve given, and what they’re trying to accomplish.

Smartly’s CEO, Laura Desmond, framed the advantage clearly: “The opportunity with conversational advertising is you can do more follow-ups, and you can ask again.” Business Insider

For marketers, that implies a structural shift:

  • Your ad is no longer a message. It’s a mini-experience with branching questions and answers.
  • Your “landing page” moves inside the assistant. The first 30 seconds of interaction may happen before the user ever hits your site.
  • Your conversion rate will depend on conversation design. Not just copy, but the logic of how you guide discovery.

2) Measurement will move upstream: “did the ad help the user decide?”

One reason OpenAI has been conservative is trust. ChatGPT has limited space for ads, and OpenAI has said ads will be separate, clearly labeled, and won’t influence organic answers. Business Insider

But once ads become interactive, attribution changes. Expect performance reporting to evolve beyond clicks into:

  • Conversation-start rate (how often users choose to engage the ad experience)
  • Qualified intent signals (budget range, use case, timeline, constraints)
  • Completion rate (how often users finish the guided flow)
  • Assist value (whether the ad reduced time-to-decision even if the purchase happens later)

There’s a real business incentive here: Criteo—OpenAI’s first adtech partner—said users referred from LLM platforms like ChatGPT convert at ~1.5x the rate of other referral channels (based on a sample of 500 U.S. retailers in February 2026). Criteo

That’s not a promise that ChatGPT ads will outperform Google Search—yet. But it is a strong clue that assistant-driven traffic can be higher-intent when the experience is designed for discovery and consideration, not interruption.

3) Creative strategy: build “answer assets,” not campaigns

Conversational ads will reward brands that already have strong “answer infrastructure”—clear product data, clean differentiators, and proof that can be summarized without distortion.

Here’s the operational shift we recommend for teams preparing now:

  • Create decision trees for your top 5 buyer scenarios. (If X industry + Y budget + Z constraint → recommend A vs B.)
  • Turn proof into compact artifacts. Case study bullets, benchmark metrics, pricing ranges, integration lists—structured so an assistant can use them.
  • Write objection-handling responses. Not long FAQs—short, evidence-backed answers to “why not you?” questions.
  • Design the ad as a guided consult. The goal is to diagnose and recommend, not pitch.

Smartly’s own example is instructive: it cited conversational ads for UK retailer Boots on Meta platforms that were nearly five times as effective at driving sales as Meta’s basic ads. Business Insider The point isn’t that the same lift will happen in ChatGPT—the point is that interactive formats can outperform static creative when the user is in exploration mode.

4) Brand safety and compliance will become product requirements

If your ad can “chat,” your ad can also go off-script. That introduces a new category of risk: regulated claims, competitive positioning, privacy expectations, and hallucination-adjacent misinterpretations.

OpenAI has emphasized guardrails: users under 18 won’t see ads, and ads won’t appear near certain topics like politics and health. Business Insider Still, marketers should plan for governance like you would for a customer-facing chatbot:

  • Approved-claims library (what you can and cannot say)
  • Citations and proof links for performance statements
  • Escalation paths for sensitive requests
  • Conversation QA (testing prompts that try to break your policies)

What to do this week (CEO-level action plan)

  • Inventory your “conversion-critical answers.” What are the 20 questions that decide deals in your category?
  • Build a structured proof pack. Metrics, case studies, pricing logic, differentiators—ready to be summarized cleanly.
  • Design a conversational funnel. A simple 6–10 turn guided flow beats a “learn more” CTA in an assistant interface.
  • Upgrade measurement. Track qualified intent signals (not just sessions) so you can judge assist value.

Bottom line: ChatGPT ads are evolving into a new kind of marketing surface—one where the best-performing brands will look less like advertisers and more like expert guides.

If you want help building citation-ready content, structured proof assets, and conversational funnels that perform in AI search and assistant interfaces, Real Internet Sales can help. Call 803-708-5514 or visit realinternetsales.com.