ChatGPT Ads: First Impressions After Real Campaigns – Is It Worth Testing?

ChatGPT Ads has been live for a few weeks now in Europe (in US it has more time already). The hype was immediate. So were the first campaigns.

We have been watching from an interesting position. As the team behind CPV Lab and CPV One, we see what happens when media buyers and agencies connect a new traffic source to their tracker and start spending real money on it. We see the setup questions, the early data, the surprises, and the frustrations.

And conversations have started in the broader community too. Agencies are sharing early numbers on LinkedIn. Media buyers are posting results on forums like Afflift. The picture that is emerging is more nuanced than the initial hype suggested – and more interesting.

This article is our honest attempt to capture what the first real campaigns look like. Not a product pitch. Not a setup guide – we have the full documentation for that. Just the clearest picture we can give you of what ChatGPT Ads actually delivers right now, based on early data from people who are already spending.

If you want the broader picture of how AI advertising is changing across ChatGPT, Google, Meta, and Microsoft, read our full guide: AI Advertising in 2026: Everything Marketers Need to Know

We will update this article as more data comes in.

What ChatGPT Ads Actually Is – The Quick Version

Before we get into results, a few things worth knowing that are easy to miss in the launch coverage.

ChatGPT Ads only reaches Free and Go plan users.

Pro and Team subscribers do not see ads. This is an important detail for audience planning – you are reaching a specific segment of ChatGPT’s user base, not all of it.

Targeting is minimal by design.

Geographic targeting yes. Age and gender targeting no. The main qualitative targeting option is “context hints” – a list of keywords and phrases that tell OpenAI which conversations to show your ad in. This is contextual targeting, not intent-based targeting like Google Search. The system decides where your ad appears within those constraints.

There are two ad formats.

Standard ads (image plus text) and product feed campaigns for ecommerce. The feed format has implementation challenges in the current beta – more on that later.

The bidding landscape has changed fast.

Initially ChatGPT Ads only offered CPC and CPM with high minimum bids. CPA bidding has since been added. This matters for performance – the CPA model with automatic bidding is behaving very differently from the early manual CPC campaigns.

The Early Numbers – What People Are Reporting

We want to be transparent about how we gathered this data.

These are early results got from one of our agency users and also shared publicly by agency teams and media buyers who started testing ChatGPT Ads in the first weeks after launch – from LinkedIn posts, from community discussions on Afflift, and from support questions we got. Numbers come from different markets, different industries, and different campaign setups. They are early signals, not statistically significant conclusions.

That said, they are real. And they paint a consistent enough picture to be useful.

B2B Lead Gen – Digital Marketing Agency

One agency team results after the first 7 days of running ChatGPT Ads for lead gen. Their setup was a standard lead gen campaign for digital marketing services in a country in Europe.

Results after a week:

Budget spent: ~€90
Impressions: 6,075
Clicks: 74
Average CPC: ~€1.12
Qualified leads: 2

For their specific niche, agencies looking for digital marketing services, volume was limited.

ChatGPT Ads shows in relevant conversations, and there are only so many conversations about finding a digital marketing agency happening at any given moment.

Tracking worked well, especially with CPV Lab already configured.

B2B Services – Agency, Romania

A second agency team ran a campaign promoting their own services, targeting locally with context hints for their niche. This one came with more specific performance data.

Results from early campaign:

• Click-through rate: 1.83%
• Average CPC: €0.75
• Market: Romania, B2B services niche

Their positioning assessment was one of the most useful things we read about ChatGPT Ads:

“Somewhere between Meta Ads and LinkedIn Ads in terms of CPC and CTR. Higher cost per click than Meta, lower than LinkedIn. Volume is limited so do not expect a major impact – think +5% on budget for +2-3% additional traffic.”

So, ChatGPT Ads sits in the consideration phase of the customer journey. Not a direct response channel where you expect immediate conversions. More of a presence channel where you reach users at the moment they are researching and asking questions.

B2B Lead Gen – US Market, Independent Affiliate Media Buyer

From our affiliate users we didn’t get much data yet about their CPA campaigns and how they work on ChatGPT Ads. From support we learned that they consider ChatGPT Ads still too expensive to test running their CPA campaigns.

But, on Afflift, a media buyer running US B2B lead gen campaigns shared a detailed performance breakdowns:

“It is expensive, but it has been improving a lot. Initially they only had CPC and CPM campaigns and the bids were crazy. Now they added CPA bidding, and while the manual recommendation barely gets you any traffic, if you set it on auto you actually get decent clicks and conversions. I am paying around $3 per click, which is a bit less than Meta and Google. US, B2B industry, lead gen. Still optimizing but I am just shy of -20% ROI, and the last campaign I launched before optimizing is close to breaking even. I have a good feeling about it becoming positive this month.”

A few things stand out here.

The $3 CPC in the US B2B market is lower than what many media buyers pay on Google Search or LinkedIn for the same audience. The shift from manual CPC to automatic CPA bidding appears to be a meaningful change in how the channel performs – something worth testing if you are still on manual CPC.

And the trajectory matters.

Going from “crazy bids” to close-to-break-even in a few weeks of optimization is a positive signal for a brand new channel still in beta.

What the Numbers Tell Us When Read Together

If you check these 3 data points (who come from different markets, different niches, and different campaign setups) we notice that they point in a consistent direction.

Volume is real but limited.

Every data point shows that ChatGPT Ads does not deliver the scale of Meta or Google right now. This is partly by design – ads only reach Free and Go plan users, and they only show in contextually relevant conversations. For niche B2B categories with low conversation volume, that is a genuine constraint.

CPC is competitive with major channels.

The range we are seeing – €0.75 to $3.00 depending on market and niche – puts ChatGPT Ads in a similar bracket to Meta Ads and below LinkedIn for B2B. That is not the “impossibly expensive new channel” narrative some early coverage suggested.

CPA auto bidding changes the equation.

The Afflift experience is an important signal. Manual CPC on a new channel with no conversion history produces poor results. Switching to automatic CPA bidding – and giving the algorithm conversion data to learn from – appears to shift performance meaningfully. This is the same pattern we saw with Facebook in its early days.

The channel is improving fast.

Every person we heard from noted the channel has changed significantly even in its first weeks. CPA bidding added. Delivery improving. Costs stabilizing. This is not a channel that launched finished – it is a channel that launched early and is iterating quickly.

	Agency – B2B	Agency – Romania Services	Media Buyer – US B2B
Market	Germany	Romania	United States
Niche	Digital marketing services	B2B services	Lead generation
CPC	~€1.12	€0.75	~$3.00
CTR	~1.2%	1.83%	Not reported
Budget tested	~€90 / 7 days	Not reported	Ongoing
Volume	Limited	Limited	Improving with CPA auto
Bidding model	Not reported	Not reported	CPA auto (recommended)
Conversions	2 qualified leads	Not reported	Trending to break-even
Verdict	Niche-dependent	Consideration phase	Positive trajectory

Where ChatGPT Ads Fits in the Media Mix

Based on what we are seeing, here is our honest read on where ChatGPT Ads belongs right now.

Best fit:

• B2B lead gen where audiences are actively using ChatGPT to research solutions
• Direct response advertisers comfortable with higher CPC in exchange for high-intent context
• Brands wanting to be present in a new channel early before competition increases costs
• Performance marketers who have proper tracking in place and can read the data correctly

Not the right fit yet:

• Ecommerce at scale – the feed format has current implementation challenges
• Campaigns needing high volume quickly – the reach is genuinely limited right now
• Performance marketers who cannot afford to test and optimize – this channel needs patience
• Any campaign running without proper conversion tracking – the algorithm needs conversion signals to improve

Where it sits in the funnel: The agencies we heard from consistently placed ChatGPT Ads in the consideration phase. Users who see your ad are in the middle of a research conversation – they are interested but not necessarily ready to buy immediately. Think of it as a presence channel that builds familiarity and captures users mid-research, rather than a direct conversion channel like Google Search.

The Tracking Question

We cannot write about a new ad channel without talking about tracking – it is what we do.

The pattern we see with new channels is consistent: people launch fast, tracking is an afterthought, data is messy, optimization is impossible. ChatGPT Ads has one thing going for it here that not every new channel does – OpenAI built it with server-side tracking in mind from the start.

The oppref click ID parameter works reliably. ChatGPT CAPI functions the same way as Facebook CAPI and TikTok Events API. Cost data pulls via the OpenAI Ads API. The infrastructure is clean.

But it still requires setup. And the users who are getting the most useful data from their early ChatGPT Ads campaigns are the ones who had proper tracking configured before they spent their first dollar – not the ones who added it later.

If you are going to test ChatGPT Ads, set up your tracking first. Full setup guide here: How to Track ChatGPT Ads with CPV Lab

Is ChatGPT Ads Worth Testing?

Our honest answer: yes, if you approach it correctly.

Not as a replacement for your existing channels. Not with the expectation of Google Search volume or Facebook reach. But as a new source of high-intent traffic from users who are actively researching solutions – tracked properly, optimized patiently, and evaluated against real data.

The media buyers who will benefit most from ChatGPT Ads in the next 6-12 months are the ones who start now, build tracking infrastructure from day one, and treat the early campaigns as data collection rather than immediate profit.

The ones who will be frustrated are those who spend $500 with no tracking, see messy numbers, and write the channel off before giving the algorithm enough conversion data to do anything useful.

The practical starting point:

• Budget: start small – $200-500 is ok to get some initial data
• Bidding: test manual CPC first to understand your baseline, then switch to automatic CPA once you have conversion history
• Tracking: set up CPV Lab or CPV One before you launch – not after
• Expectations: treat month one as research, not profit

What Comes Next

This article will be updated as more campaign data becomes available. We are watching results across CPV Lab and CPV One users running ChatGPT Ads and will add aggregate insights as the data becomes meaningful.

If you are running ChatGPT Ads and want to share your experience – results, observations, what is working and what is not – we would love to hear from you. The more real data this community shares, the better everyone’s campaigns get.

We are also published an article specifically on the setup mistakes people make with ChatGPT Ads tracking, the specific errors that cause data to break and how to fix each one.

Frequently Asked Questions

❓ Is ChatGPT Ads worth it for affiliate marketers? Based on early results, ChatGPT Ads shows the most promise for B2B lead gen and direct response advertisers in categories where users actively research solutions in ChatGPT. Volume is currently limited compared to Google or Meta, but CPC is competitive and the channel is improving fast. The key is having proper conversion tracking set up from day one so the algorithm has data to optimize from. For context on how ChatGPT Ads fits into the wider AI advertising landscape alongside Google, Meta, and Microsoft, see: AI Advertising in 2026

❓ What is the average CPC for ChatGPT Ads? Early data from campaigns running in September 2026 shows CPC ranging from around €0.75 for niche B2B services in European markets to approximately $3.00 for US B2B lead gen. This puts ChatGPT Ads in a similar range to Meta Ads and below LinkedIn for B2B audiences. These are early numbers and will shift as the platform matures and competition increases.

❓ How do ChatGPT Ads compare to Meta Ads and Google Ads? Based on early campaign data, ChatGPT Ads sits between Meta and LinkedIn in terms of CPC and CTR for B2B audiences. Volume is lower than both. The channel is more contextual than intent-based – users are in research mode, not search mode. It is best treated as a consideration-phase channel rather than a direct conversion channel, at least in its current state.

❓ Who sees ChatGPT Ads? ChatGPT Ads only shows to users on the Free and Go subscription plans. Users on Pro, Team, or Enterprise plans do not see ads. This limits total reach but concentrates delivery toward a specific audience segment.

❓ Does CPA bidding work better than CPC on ChatGPT Ads? Early evidence suggests yes. Manual CPC bidding on a new channel with no conversion history tends to produce limited traffic and poor optimization. Switching to automatic CPA bidding – once you have some conversion data flowing – appears to improve both volume and performance. The pattern mirrors what happened with Facebook’s bidding evolution in its early days.

❓ How do I track ChatGPT Ads conversions properly? ChatGPT Ads uses a unique click ID parameter called oppref for attribution. CPV Lab and CPV One support full ChatGPT Ads tracking including click capture via oppref, server-side conversion reporting via ChatGPT CAPI, and automatic cost synchronization via the OpenAI Ads API. Full setup guide: How to Track ChatGPT Ads with CPV Lab

❓ What verticals work best for ChatGPT Ads right now? B2B lead gen, digital marketing services, and direct response offers in categories where users actively ask ChatGPT for recommendations appear to be the strongest early performers. Ecommerce via product feed campaigns has current implementation challenges. High-volume consumer categories may see limited reach due to low conversation volume in those specific topics.