Well, here we are.
The chat window is turning into media inventory.
I don’t say that dramatically. I say it like a founder who has watched every useful interface eventually discover monetization.
OpenAI’s update on testing ads in ChatGPT says the pilot is moving into its next phase after early signals showed no impact on consumer trust metrics and low dismissal rates. OpenAI’s advertiser page now frames ChatGPT as a place where people move from exploring to deciding through conversation. And Reuters reported last week that ads will be shown to all free and Go users in the U.S. in the coming weeks.
That’s not a feature tweak.
That’s a new performance channel being born in public.
The intent quality could be absurdly good
This is why every serious GTM operator should pay attention.
Search ads work because intent leaks through the query.
Chat ads may work even better in some cases because intent leaks through the conversation.
That is a much richer signal.
A user is not just typing:
“best CRM”
They may be saying:
“I run a 15-person agency, I’m drowning in follow-ups, I hate clunky reporting, and I need something my sales lead will actually use.”
That is not just a keyword.
That is buying context.
If platforms can place relevant offers around that moment without wrecking trust, the economics could get very interesting.
But trust is the whole game
This is where the opportunity gets fragile.
Chat is more intimate than search.
People ask messy questions here. They think out loud. They compare options. They ask for help before they’ve fully decided what they believe.
That means the ad model cannot feel predatory.
OpenAI clearly knows that. Their ads update leans hard on trust, usefulness, and user control. They have also published ad placement rules and say sensitive contexts are off-limits.
That is not just safety language. It is commercial logic.
If users start feeling like every useful answer is surrounded by opportunistic nonsense, the interface gets worse fast — and the underlying product loses its magic.
This will change performance marketing strategy
A lot of advertisers will initially approach chat the wrong way.
They’ll try to cram search-era habits into a conversational surface:
brute-force targeting,
generic ad copy,
landing pages written for traffic instead of decisions,
and poaching tactics without any real user fit.
Some of that will work briefly because ad markets are messy at first.
But the long-term winners will probably be the brands that do three things well:
1. Show up at the right moment
Not just “in market,” but mid-decision.
2. Match the conversation
If the ad sounds disconnected from the user’s actual context, it will feel like spam with better timing.
3. Deliver after the click
Conversational intent is usually more specific than classic traffic. That means the landing page and offer need to pick up exactly where the conversation left off.
The new media buyer will need a new brain
This is the part I find fun.
Performance teams used to live in a world of keywords, placements, audiences, and creative tests.
Now they need to think about:
conversational context,
decision-stage relevance,
trust preservation,
richer intent interpretation,
and how sponsored options fit naturally inside a research flow.
That is a different skill set.
Closer to product thinking. Closer to answer design. Closer to demand capture with restraint.
Which, honestly, is probably healthy.
My founder take
I don’t think chat ads are automatically bad.
I think bad chat ads will be bad very quickly.
If done well, sponsored results inside decision-oriented conversations could be genuinely useful. If done poorly, they’ll feel like someone interrupting your best thinking with a flyer.
So yes — this is the birth of a new ad surface.
But it’s also a test.
Not of whether chat can be monetized.
Of whether monetization can coexist with trust in a place where users are trying to think clearly.
That answer will matter a lot more than the CPM.
