Every few years, a new ad platform arrives with enough hype to make performance marketers question everything. Should CPG and DTC brands advertise on ChatGPT? Not yet. Not without applying the same disciplined evaluation you’d use for any emerging channel. The platform is new, measurement is promised but unproven, and “shiny object” budget decisions have a long, expensive history of underperforming.
That doesn’t mean ignore it. It means be honest about what you’re buying.
What ChatGPT Ads Actually Represent (And What They Don’t)
ChatGPT’s move into self-serve advertising is genuinely significant. A platform with hundreds of millions of active users, removing its minimum spend barrier and opening to smaller brands, is a real structural shift. Not just a press release.
But significant doesn’t mean proven. There’s a difference between a platform that can carry advertising and one that should carry yours.
Right now, ChatGPT ads exist in a space where user intent is high, but purchase behavior is largely unknown. We don’t yet have independent data showing CPAs, ROAS benchmarks, or conversion rates for CPG or DTC categories. Third-party measurement is promised. Promised is not the same as delivered.
The Questions Every Performance Marketer Should Be Asking Right Now
Before a single dollar moves, here’s the framework we’d apply at Junction 37:
- What does intent actually look like here? Someone asking ChatGPT “what’s a good magnesium supplement” is not the same buyer signal as someone searching the same phrase on Google Shopping. The context is different. The proximity to purchase is different. That matters enormously for performance budgets.
- Who controls measurement, and how independent is it? CPA bidding is only useful if the CPA data is trustworthy. Self-reported metrics from a platform selling you inventory is not a measurement strategy. Until third-party measurement partners are live, verifiable, and integrated — treat reported numbers with healthy skepticism.
- What are you actually displacing? This is the question nobody asks loudly enough. If you’re moving $20K from a proven Meta or paid search campaign to test ChatGPT ads, you’re not just experimenting — you’re making a trade. Know what you’re trading and what the opportunity cost is before you make that call.
- Is your creative built for this environment? Ad creative that performs on Instagram Reels will not automatically translate to a conversational AI interface. This is a fundamentally different context: more editorial, more considered, less interruptive. If you’re not thinking about format-specific creative, you’re setting the test up to fail.
- What does success look like at 90 days? Define it before you launch. Not vanity metrics. Real ones: cost per acquisition, contribution margin impact, new customer rate. If you can’t define what winning looks like, you’re not running a test. You’re running an experiment with no hypothesis.
The Brands Most Likely to Win Early on ChatGPT Ads
If we had to place a bet on who benefits first, it’s not mass-market CPG. It’s:
- DTC brands with high consideration products: supplements, skincare, personal finance — where ChatGPT’s research-oriented users are already in the right mindset
- Brands with strong editorial content: if your brand already communicates with depth and substance, you’ll adapt faster to conversational ad formats
- Performance-mature brands: teams with clean attribution infrastructure and the discipline to read results without bias
If you’re still cleaning up your Meta account structure or haven’t nailed incrementality testing, ChatGPT ads are not your next move. Fix the foundation first.
What We’d Actually Do Right Now
We’d allocate no more than 5-10% of an experimental budget to ChatGPT ads in Q3/Q4 of this year. Enough to generate real data, not enough to hurt you if the results disappoint.
We’d run it alongside, not instead of, your proven performance media channels. We’d build creative specifically for the format. And we’d measure it with a third-party tool the moment that capability is live, not before.
Curiosity is smart. Chasing every new platform without a disciplined process is how brands burn budget and lose confidence in media altogether.
The platforms that last earn their place in the mix. Give ChatGPT ads the chance to do that — on your terms, not theirs.
Ready to build a performance media strategy that evaluates new channels without abandoning what’s working? Talk to the Junction 37 team about our approach to performance media strategy.
FAQ: ChatGPT Ads for CPG and DTC Brands
What are ChatGPT ads and how do they work?
ChatGPT ads are sponsored placements within OpenAI’s ChatGPT platform, served to users as they interact with the AI interface. Advertisers can now access them through a self-serve ads manager in the U.S., with CPA bidding options and third-party measurement in development. Think of them as contextual ads inside a conversational search experience — not display, not social, something new.
Are ChatGPT ads worth testing for DTC brands?
Potentially, but only with a disciplined test structure. DTC brands selling high-consideration products may find genuine intent-match with ChatGPT’s user base. However, until independent measurement is established and benchmark data exists by category, treat it as a learning investment — not a primary performance channel.
How is advertising on ChatGPT different from Google or Meta ads?
Google captures active search intent. Meta targets based on behavioral and demographic signals. ChatGPT operates in a research and decision-support context — users are often in an exploratory, pre-purchase mindset. That creates different creative requirements, different funnel placement, and currently, far less performance data to rely on. Google’s own research on search intent illustrates how intent context shapes conversion behavior.
How much should I budget for a ChatGPT ads test?
Start small. 5-10% of your experimental media budget at most. Enough to gather statistically meaningful data across a 60-90 day window, but not so much that a disappointing result materially impacts your overall performance targets. Define your success metrics before you launch.
Chris Pyne, Founder, Junction 37 – 30+ Years in Performance Media.