I use ChatGPT all the time, and not just for work. I’ve used it to find the best cheap-to-good restaurant ratio, troubleshoot my 20-year-old Bronco when, not if, it breaks, compare products, plan trips and understand topics I knew absolutely nothing about five minutes earlier.

It’s amazing. And as someone who has spent 20 years in media, it also makes me look at the experience through a different lens.

People are researching, comparing, asking questions and making decisions inside ChatGPT. That sounds like a marketer’s dream. But finding where people spend time has never been enough. Our job is figuring out how advertising fits the way they actually use that place.

Stone Ward recently received access to test advertising within ChatGPT. Before we start adding another logo to every media plan, here are five things I’m watching.

Every great advertising platform eventually connects user behavior to a natural ad opportunity.

 

Search is easy. I type “best running shoes,” and ads sit at the top of the results. Social figured out how to use identity, interests and behavior to place advertising directly into the feed.

 

ChatGPT is different. I can explain exactly what I want, add context, ask follow-up questions and narrow my choices. That is a lot more information than a two-word Google search.

 

Coach CMO Joon Silverstein recently described the shift perfectly: “It’s not just based on keywords. The searches are extraordinarily specific.”

 

That specificity is what makes the advertising opportunity so intriguing.

 

The problem is that today the ad still sits separately below the answer. OpenAI has been adding new formats, buying options and measurement tools quickly, but even OpenAI describes the platform as being early in its development.

 

There is a lot of context above the ad. We are still figuring out how well the advertising connects to it.

Imagine the difference between searching “restaurants Chicago” and telling ChatGPT:

 

“I’m in Chicago for one night, staying near River North, want somewhere locally known but not touristy, around $50 a person, good cocktails and a reservation tonight.”

 

As a marketer, it is hard not to get excited about that.

 

But understanding a conversation and translating it into the right commercial opportunity are two different things.

 

Recent tracking from Seer Interactive found ChatGPT ads appearing across a rapidly growing share of the conversations it monitors, but the average alignment between the ad and the response was only 3.29 out of 5. More than half of the ads it analyzed were categorized as “category-adjacent” rather than highly aligned.

 

That feels like an important early lesson: great data does not automatically create great targeting.

This is the part I keep coming back to.

 

If I ask ChatGPT how to fix something on my Bronco and it gives me the answer, what role does the ad underneath play? Maybe it is perfect because now I need a part, a tool or a mechanic.

 

Or maybe I’m done.

 

Search advertising often gets the first word. Social advertising is woven into the experience. ChatGPT advertising currently gets something closer to the last word.

 

That could be incredibly valuable during comparison and decision-making. It could also be completely forgettable during an informational conversation.

 

The conversation matters as much as the audience.

Every new media channel gets compared with whatever came before it. We like benchmarks because benchmarks make new things feel less new.

 

But I’m not sure “Can ChatGPT Ads beat paid search?” is the right first question.

 

Maybe conversational advertising becomes a great conversion channel. Maybe it is better during consideration. Maybe certain categories thrive while others never make sense.

 

A better test is: Can this reach someone at a moment our existing media cannot reach as effectively?

 

That tells us something useful about its role in the plan instead of forcing a brand-new behavior into an old media box.

There is always pressure in advertising to be first.

 

Gary Vaynerchuk recently summed up the broader technology shift as “Extreme AI, extreme analog. Don’t get caught in the middle.”

 

I agree with the spirit of that. But being early does not mean throwing money at something because everyone at the conference is talking about it.

 

For ChatGPT Ads, start with a hypothesis, not a budget.

 

What kind of conversation do we believe creates an opportunity for the brand? What does success look like beyond a click? Are the leads any good? Did the advertising add something to the customer journey that search, social or another channel did not?

 

And if the answer is “not yet,” that is useful too.

 

OpenAI is moving fast. CPC buying, conversion measurement, CPA options, new creative formats and additional targeting capabilities have all arrived within months. What ChatGPT advertising looks like today probably is not what it will look like a year from now.

 

That is exactly why media planners should be paying attention.

 

Not because ChatGPT needs to be on every media plan.

 

Because the opportunity is never simply being where the consumer is. It is understanding what they are doing there, and finding a useful way for the brand to enter the conversation.