Who Should Not Buy ChatGPT Ads Yet

Last verified 2 October 2026. Version 1.2. Next scheduled review 2 November 2026.

Every agency page in this category answers the question "should you advertise on ChatGPT" with some version of yes. That is not because the answer is yes. It is because the pages are written by people who sell the service.

We do not run campaigns, which means we have nothing to lose by writing the disqualification list. Six conditions rule a business out of this channel today. Five cost nothing to check.

1. Your buyers sit on plans that do not show ads

This is the one that quietly kills most B2B cases, and it is structural rather than temporary.

OpenAI's documentation states that ads are not shown to users on Plus, Pro or any Business plan, or to users it predicts are under 18. Ads run on the Free and Go tiers.

Now describe the enterprise software buyer. A director of engineering at a company that bought seats. A procurement lead evaluating vendors. A marketing operations manager at a Series C startup. These people are on Business plans, and their employers pay for them precisely because the paid tiers are better for work.

So a B2B SaaS company advertising in ChatGPT is buying an audience that systematically excludes the job titles it sells to. Not partially. By design.

The counter-argument worth taking seriously: those buyers have personal accounts too, and early-stage research often happens on a personal device outside procurement hours. That is real, and it is unmeasurable with what OpenAI currently reports. If your case rests on it, you are buying an audience you cannot verify you reached.

2. You sell in a restricted or excluded category

Ad policies v1.6, updated 10 September 2026, limits advertising primarily to consumer categories: household and consumer goods, local services, travel and experiences, digital products and education.

Disallowed outright: gambling, alcohol, tobacco, dating and adult content, recreational drugs, counterfeit goods, scams, and political or contested social issue content.

Two traps in the shorthand version.

Finance, health and legal services are US-only carve-outs. OpenAI may approve advertisers in these categories case by case with proof of licensure. Legal advertisers must be licensed in the jurisdiction where the ad is shown. A licensed insurer or law firm outside the US does not get the case-by-case route.

Category eligibility is not the same as inventory. Ads do not run beside conversations about personal health, mental health or politics, or in what the policy calls sensitive user journeys. If your product gets bought during exactly those conversations, your available inventory is thinner than your category suggests and bidding does not fix it.

3. Your buyers are in a market that does not serve ads

Opening an advertiser account and reaching a particular buyer location are separate checks. OpenAI now lists 63 self-service advertiser countries, including India, the UAE, Singapore and the whole of the EU, so this disqualifier is narrower than it was in August. It still applies if your buyers are in China, sub-Saharan Africa, Latin America outside Brazil and Mexico, or anywhere else not on the country list.

There is also a timing version of it. OpenAI says some new self-serve accounts can advertise only in their home country until verification is approved and they reach a required spend there. If your buyers are abroad and your entity is not, your first weeks of spend may not reach them. Confirm the buyer location in the live campaign picker before treating it as reachable.

4. You have no organic baseline to compare against

This one is not a platform restriction. It is a measurement problem that makes the spend uninterpretable.

If a brand starts buying ChatGPT ads without knowing how often assistants already mention it, name it in a category answer, or cite it as a source, there is no way to attribute what happens next. Traffic rises, and the cause could be the ads, the organic position moving, seasonality, or a competitor's site going stale.

OpenAI states that it has not published performance benchmarks across advertisers, industries or campaign types, so there is no external number to compare your results against either. Without a baseline you have no internal one and no external one.

Whether paid placement lifts or cannibalises organic citation has no published answer at all. A brand with no baseline cannot even answer it for itself.

5. You cannot fund the floor for long enough to learn anything

The claim that ChatGPT ads have no minimum spend is widespread and out of date. The 200,000 USD entry commitment was removed on 5 May 2026. A campaign floor is a different thing and it still exists: 25 USD per day per campaign in the United States, with local equivalents elsewhere.

That is not expensive. The problem is what it buys. At 25 USD a day, a two-week test is roughly 350 USD, and at the reported clearing range for this channel that is a few thousand impressions against a conversion event most businesses see a handful of times a month.

A test too small to reach statistical significance is not a cheap test. It is a purchased opinion. If the budget you can defend internally is 500 USD, the honest read is that you will finish the test knowing roughly what you knew going in.

6. You are brand-safety sensitive in a way this surface cannot yet satisfy

Ads appear inside a conversational answer. Advertisers receive aggregated performance data and do not see the conversations their ads appeared beside.

For most consumer advertisers this is fine and comparable to other channels. For a regulated brand, a listed company, or anyone whose legal team requires context-level placement reporting, the current disclosure level is thinner than display or search, and there is no adjacency report to hand over.

Separately, an Ipsos finding reports that 63% of US adults say ads in AI search results reduce their trust in the results. That is survey data about attitudes rather than measured brand damage, and it is the kind of finding that gets over-read in both directions. It belongs in the decision, not at the centre of it.

Who should buy

The symmetric list, because a disqualification page that disqualifies everyone is just pessimism.

You are a good fit today if your business is in a supported country, the intended buyer location is available in Ads Manager, you sell in an approved category, your buyers are likely on an ad-supported plan, you have a measured organic baseline, and you can fund a test large enough to produce an actionable signal.

Household goods, local services, travel, consumer digital products and education. That is the shape of the fit, and it is a real market.

What we cannot tell you

Whether the Business plan exclusion changes. OpenAI's documentation names the excluded plans today. It has revised that documentation twice in the last month.

How much B2B research happens on personal accounts. This is the strongest counter to disqualifier one and there is no published data on it.

What a statistically adequate test costs in your category. It depends on your conversion rate and your event volume. Nobody has published a channel-wide answer and we are not going to invent one.

Whether the Ipsos trust finding translates into measurable brand damage. Stated intent and behaviour diverge routinely. No study we have found measures the second.

Quick answers

Are ChatGPT ads good for B2B? Structurally weak today. OpenAI does not show ads to users on Plus, Pro or any Business plan, which excludes most enterprise buyers on work-provisioned accounts. B2B cases rest on buyers researching from personal accounts, which is currently unmeasurable.

Who does not see ads in ChatGPT? Users on Plus, Pro or any Business plan, and users OpenAI predicts are under 18. Ads run on the Free and Go tiers.

Can financial services advertise on ChatGPT? In the United States, case by case with proof of licensure. Outside the United States, ads for financial services are generally prohibited. The same applies to health services.

Is there a minimum spend for ChatGPT ads? The 200,000 USD entry commitment was removed on 5 May 2026. A campaign floor remains: 25 USD per day per campaign in the United States, with a published floor for each billing currency elsewhere.

Should I run ChatGPT ads if I have never measured my organic AI visibility? The spend will be hard to interpret. There is no published benchmark to compare against and no baseline to attribute against, so any change after launch has several equally plausible causes.

Sources

Related reading

• Should You Buy ChatGPT Ads Yet? — the six gates as an interactive scorecard.

• ChatGPT Ads by country — if disqualifier three is yours

• What has to be true before you can run a ChatGPT ad — the eleven conditions if none of the six above apply to you

Changelog

2 October 2026. v1.2 updated disqualifier three for the 63-country list and added the home-country rule for new accounts.

13 September 2026. v1.1 corrected legal services from globally prohibited to US-only, case-by-case eligibility for licensed advertisers under policy v1.6.

15 August 2026. v1.0 published.

Field kit

Tools

• Eligibility Checker
• Budget Calculator
• OAI-AdsBot Checker
• Context Hints Generator
• Buy Scorecard

Site

• ChatGPT Ads
• Google AI Mode Ads
• Methodology
• Glossary
• Contact

OpenAI (primary)

• ChatGPT Ads Help Center
• Ads in ChatGPT: The Basics
• Measure Results
• Ad Policies
• Ads Manager
• OAI-AdsBot verification