ChatGPT Ads: why budgets remain in the testing phase

A client spends $10 million a month on Google and less than $100,000 on ChatGPT — not because they don't want to, but because they can't tell if the money is working.

Seven months of advertising on ChatGPT — and it's still unclear what works

OpenAI started selling ad space inside ChatGPT in early 2026. Seven months later, the main issue isn't the cost of the ads or the quality of the audience: it's that nobody can reliably measure the results. This is reported by Digiday after speaking with several agencies managing active campaigns on the platform.

This doesn't mean ChatGPT Ads don't work. It means, today, you can't know for sure. And without certainty, scaling the budget is a leap in the dark.

What's wrong with the numbers the platform shows you

The most striking example: a lead generation campaign on ChatGPT generated completed forms visible in real time, but after two weeks the platform still showed zero recorded conversions. Same scenario, opposite direction: the platform reported about 100 clicks on a campaign, while the external tracking system had detected 20.

Even when the data finally shows up, it is late. Some agencies see conversions pop up 24-36 hours later, a far cry from the few minutes Google has us used to. And the reports do not include any demographic info — like age, gender, or interests — or the topics of the chats where your ad popped up.

For those managing campaigns across multiple channels, this creates a real problem: you can't compare ChatGPT with Google or Meta on a level playing field, and therefore you can't decide where to shift your budget. If you want to dive deeper into how attribution models can mess with your spending decisions, check out the article on advertising attribution and the models that deceive the budget is a good starting point.

Why many companies don't even enable tracking

To receive conversion data, you need to accept OpenAI's tracking terms. Many legal departments refuse, worried about liability for customer data and uncertainty over how OpenAI will use it. The paradox is that these same companies have been sharing similar data with Google and Meta for years—but with those platforms, there's a familiarity built over time. With OpenAI there isn't, and the company hasn't shown a willingness to negotiate custom clauses.

There is also an API (a direct technical connection between systems) that could improve data quality, but large companies are slow to implement it and smaller ones often lack the technical resources to do so. The result is a vicious cycle: OpenAI needs more data to improve tracking, but those with the most data are also the least willing to share it.

Here at SHM Studio, we believe technology should serve decision-making, not create new blind spots. Adopting an advertising tool without an independent measurement system isn't innovation: it's giving up control over your own budget.

Optimizing for clicks and impressions doesn't tell you if you're selling

Without reliable conversion data, most ChatGPT campaigns end up optimizing for clicks and impressions — meaning how many people saw or clicked the ad. Those are visibility metrics, not results. They do not tell you if anyone actually bought something, filled out a form, or asked for a quote.

The platform is introducing the ability to optimize toward cost per acquisition (how much you spend to get a customer or lead), but it only works well if OpenAI has enough verified conversion data to calibrate bids. Today, that data is scarce. Meanwhile, some agencies are receiving automated alerts asking them to raise their bids without explaining by how much—alerts that sometimes disappear on their own after a few days, with no explanation.

Ad delivery is equally unpredictable: thousands of impressions one week, less than a hundred the next, for no obvious reason. This makes it tough to learn from one campaign and apply those lessons to the next. To figure out how we got here, it is worth rereading the article on how OpenAI opened up to advertising in 2026 and the one on how OpenAI is monetizing ChatGPT .

OpenAI has partnered with some measurement providers for mobile apps, but agencies report that the available integrations only confirm that an ad was seen — not that it produced a result. The format click-to-chat , which keeps the user within ChatGPT instead of taking them to the site, further complicates traditional tracking.

For those who want a broader picture on how to evaluate AI channels in advertising, the article on AI and advertising: why results matter more than hype offers a useful perspective. You can find more insights on the topic in the section Marketing and ADV performance .

What to check before increasing your budget on ChatGPT Ads

• Do you have an external tracking system in place (Google Analytics, a CRM, a third-party pixel) that measures conversions independently of the platform?
• Do your click and conversion data on ChatGPT align with your external system, or are there significant discrepancies?
• Has your legal department reviewed OpenAI's tracking terms and given the green light?
• Are you optimizing on a result metric (leads, purchases, quotes) or just on clicks and impressions?
• Have you defined your success indicators (KPIs) in advance, without relying solely on the platform's reports?
• Is the current budget small and sustainable as a test phase, without pressure to scale before having solid data?

Frequently Asked Questions

A channel you can't measure isn't an opportunity: it's a cost you don't know if you're incurring for a reason.