ChatGPT Ads: What early adopters are saying.
Updated: Sep 24
A new bombshell has entered the villa.
ChatGPT Ads had its first launch in Feb 2026, asking for a $200,000 minimum spend, which was then reduced to $50,000, and now we are in the era of OpenAI’s self-service.
The marketing community had mixed feelings when ChatGPT Ads launched; half the timeline called it the next big channel, the other half called it vapourware. I made an account the day it went live in the UK. Not because I had a client ready to spend, but because I wanted to see it before the takes did.
That's the thing about a new channel. Everyone has an opinion. Not everyone has spend behind it.
So I went looking for people who actually do. Marketers who've put real budget into ChatGPT Ads, across lead gen, ecommerce, and B2B. I asked them the same handful of questions: what surprised you, what would you tell someone testing this for the first time, and is this hype or is this real?
Here's what they told me.
Daniel Marshall, Head of Demand Generation at MoneyPenny
What's one result or insight from ChatGPT Ads that surprised you, good or bad?
The biggest surprise has been the quality of the leads. I expected impressions and clicks, but I didn't expect ChatGPT Ads to generate genuinely qualified leads that fit our ideal customer profile.
The cost per lead isn't always as competitive as Google Ads, but the quality has been really encouraging. In particular, our US campaigns have delivered strong CPLs, while the UK has generated solid lead volume with relevant prospects. Overall, it's exceeded my expectations as a lead generation channel.
What's one thing you'd tell someone who wants to test their first campaign?
The biggest challenge today is the lack of visibility into where your ads are appearing, so measurement is essential.
I'd recommend two things: first, use robust UTM tracking so you can measure on-site behaviour and lead quality. Second, use a third-party tool, whether that's a prompt visibility platform like Peek or an ad monitoring tool like Athena, to track the prompts you care about and make sure your ads are showing in the right places.
On a scale of hype versus real, where's ChatGPT Ads at right now for you?
For me, it's much closer to real than hype.
We're seeing qualified leads, good engagement, and evidence that businesses are discovering us through the platform. We're also seeing wider benefits, including improved organic visibility in ChatGPT and prospects telling us they've seen our ads.
That said, I don't see ChatGPT Ads replacing channels like Google Search or Meta. It's a different channel that works best as an incremental part of your marketing mix.
Kat Sale, Co-Founder, House of Performance
What's one result or insight from ChatGPT Ads that surprised you (good or bad)?
How easy to use the interface; they've clearly invested in making it simple to set up campaigns.
What's one thing you'd tell someone who wants to test their first campaign?
Think really carefully about your campaign plan - do not just copy and paste from your search campaigns; this is not a search platform, and people use it very differently. You need to think about relevant conversations to target, then structure your campaign in a way that will allow you to see data between prompt themes.
On a scale of hype vs. real, where's ChatGPT Ads at right now for you?
Loving the hype, seeing mixed performance, but it's developing fast, so I'm all for the HYPE!
Matt Beswick, Paid Marketing Manager at Quality Bearings Online
What's the biggest difference between running ChatGPT Ads and running a platform like Meta or Google Ads?
The biggest difference really is the complete lack of any type of targeting. Your targeting is done entirely through context hints, rather than targeting settings. Whilst you can exclude certain countries, you can't target/exclude locations within that country, and there is no real way of knowing within the platform if your contextual hints are being adhered to. In fact, when I've told it not to serve ads on mobiles, it's completely ignored this contextual hint! As such, we are treating the platform as more of an awareness tool rather than for direct response.
If you had double the budget on ChatGPT Ads tomorrow, would you spend it there, or hold off? Why?
Whilst we have not really had any meaningful conversions, just a couple of phone calls and a couple of add-to-carts, I probably would double the budget if given the opportunity, just so I can test on a bigger scale. If I can fail faster, I can learn faster and find out what changes actually move the needle within a couple of days rather than waiting a week or so. We are trying to get first-mover advantage on the platform, so having a bigger budget would help us learn faster than everyone else. However, if it was just from a purely commercial viewpoint, at the moment it would have to be a no.
What's one metric or signal you're watching closely to decide if this channel is working?
We have added UTMs to each ad so we can analyse within GA4 which ads are resonating most and which ads are bringing in quality clicks. So far, we have been monitoring average engagement time, as when we first launched, this was really low, suggesting click quality was poor. We are optimising to try and get this engagement time up but also looking at softer conversions like Add to Carts and telephone calls.
Jeremie Roblin, Paid Search Lead at DataCamp
What's the biggest difference between running ChatGPT Ads and running a platform like Meta or Google Ads?
I would say the most interesting difference is targeting. Paid search runs on keywords, narrow terms like "data science certification." ChatGPT Ads uses context hints, which let us describe a persona and intent instead, for example, "someone about to switch careers who might want a data science and AI certification", even if they never typed those words. We're still testing how reliably that persona and demographic detail actually carries through to who gets served the ad.
The platform is also new, so several controls standard on Google or Meta are still missing. That's closing fast; conversion optimization and audience exclusions both shipped in the last few weeks.
If you had double the budget on ChatGPT Ads tomorrow, would you spend it there, or hold off? Why?
I'd hold off. We're testing several campaign variations to find what drives activation, not just clicks. Costs run a bit high, but it's not too dissimilar to search for us. We haven't put serious budget behind it yet, so it's hard to tell if activation is limited by tools still catching up, bid optimization just rolled out, or by traffic quality. I want that clearer before adding budget.
What's one metric or signal you're watching closely to decide if this channel is working?
User activation, someone actually starting a course on their DataCamp account, rather than subscriptions. At these budget levels, subscriptions directly from the channel aren't realistic yet, so activation is the earliest signal of whether the traffic is worth scaling.
Céline Silva, Senior Marketing Specialist at TaskRabbit
What's the biggest difference between running ChatGPT Ads and running a platform like Meta or Google Ads?
Google is increasingly acting like ChatGPT with more natural-language experiences, while ChatGPT is increasingly acting like Google with features such as product feeds. The distinction is therefore less about functionality and more about how people use and trust each platform. ChatGPT is particularly interesting as a consideration-led channel, where users research and compare solutions before going directly to a website to take action. It also gives us incremental reach and a new type of contextual targeting, making it an opportunity to get ahead while the channel is still relatively new. Google, however, remains significantly stronger in reach, data availability, audience targeting, placements, campaign types, and optimisation capabilities.
If you had double the budget on ChatGPT Ads tomorrow, would you spend it there, or hold off? Why?
There are several capabilities I'd want to see develop before scaling significantly, such as search-term visibility and control, postcode-level targeting, live audience integrations, 3rd-party connections, and better reporting. For example, we currently can't see spend by individual keyword or "context hints," and some data still needs to be imported manually. I'd continue testing new features as they roll out, such as remarketing and product feeds, and increase investment as we gain more transparency and confidence in performance.
What's one metric or signal you're watching closely to decide if this channel is working?
The most important signal is internal backend conversions attributed to ChatGPT, but at this stage I'm also watching ad engagement closely. As a top-of-funnel, consideration-led channel, strong CTR is encouraging, particularly given we currently have limited control over the search terms triggering our ads. Ultimately, though, we need to see that engagement translate into incremental conversions at an efficient cost.
Brock Anderson, Marketing Practice Manager at Graybox.co
What's the biggest difference between running ChatGPT Ads and running a platform like Meta or Google Ads?
The biggest difference for me has been how quickly the platform seems to make delivery decisions. With Google and Meta, I'm used to much longer learning periods and needing a lot more conversion data before I trust what the platform is doing. With ChatGPT Ads, it picked a clear winning message very quickly and shifted most of the spend there. The targeting setup also feels different since you're working more from conversational intent and context rather than traditional keyword or audience structures.
If you had double the budget on ChatGPT Ads tomorrow, would you spend it there, or hold off? Why?
I would increase the budget, but I wouldn't double it blindly. Our first test gave us enough positive signals to keep investing, but I still see it as a testing channel rather than something I'd move a large chunk of Google or Meta budget into yet. Now that conversion optimization is available, I'd rather scale gradually while giving the platform stronger lead data to optimize toward.
What's one metric or signal you're watching closely to decide if this channel is working?
Leads and what users do after the click. CPC and CTR have been encouraging, but I'm much more interested in whether that traffic is actually engaging with the site, clicking through to Contact Us, and submitting forms. We set up pretty detailed GTM and GA4 tracking so we can follow that behavior instead of judging the channel on clicks alone.
What it comes down to
Every channel has its challenges, and it will get worse before it gets better, then worse again. Nobody here is saying ChatGPT Ads is printing money.
Dan's seeing qualified leads. Kat's seeing mixed performance but loves how easy the interface is to use. Matt's basically running it as brand awareness because targeting barely exists. Jeremine is holding budget back until activation data gets clearer. Céline's holding too, waiting on search-term visibility and better reporting before she scales.
Five people, five different answers to "is this working?" That's the honest state of the channel right now.
But underneath the mixed results, there's a pattern. Nobody's treating this like search. Kat said it outright: don't copy-paste your search campaigns in here. Matt, Jeremine, and Céline all flagged the same thing from different angles: targeting isn't a setting you configure; it's a context hint you write and hope holds. And almost everyone, without being asked, brought up UTMs and manual tracking workarounds. If the platform won't tell you what's working, you have to build that visibility yourself.
ChatGPT Ads isn't mature enough to trust its own reporting. The people getting something out of it are the ones doing their own measurement on top.
I just booked an intro call with Raquel Corradi, ChatGPT Ads Activation partner, looking to explore how this channel diversifies media planning going forward and changes buying strategy.
Comments