ChatGPT Ads Now Reach 89 Countries: How to Sell a Setup Service Without Hype

Key takeaways

• OpenAI said on 21 October 2026 that ChatGPT Ads expanded to 26 additional countries, taking availability to 89 countries worldwide.
• More than 100,000 advertisers have run campaigns on ChatGPT since ads launched just over eight months before 21 October 2026, and more than 1.2 billion people use ChatGPT each week, according to OpenAI.
• Ads appear only on the Free and Go plans. Plus, Pro, and Enterprise subscriptions remain ad-free, according to OpenAI.
• The service opportunity is setup, testing, and reporting, not selling guaranteed sales. You can charge for work and time, but you cannot promise a cost per click or a return that OpenAI has not published.
• New coverage includes 14 more Latin American countries, including Argentina, Chile, Colombia, and Peru, plus Nigeria, South Africa, Serbia, Ukraine, Cambodia, Bangladesh, and Pakistan.
• Start with one niche, one clear landing page, and one small test. Do not sell a large monthly package before you have run a controlled test for a real client.

What changed on 21 October 2026?

OpenAI announced that ChatGPT Ads expanded to 26 new countries, taking total availability to 89 countries worldwide. That is the news that makes this a service opportunity now, rather than a general idea. A business in a newly covered country, or a business that wants to reach users in those countries, may now be able to access ChatGPT Ads through technology partners or through self-service access in Ads Manager, with more information at ads.openai.com.

The scale behind the expansion is also worth noting. OpenAI says more than 1.2 billion people use ChatGPT each week to explore ideas, compare options, and make decisions. It also says over 100,000 advertisers have run campaigns on ChatGPT since ads launched just over eight months before 21 October 2026. Those numbers do not mean every advertiser will succeed. They mean the audience is large, and the platform is no longer a tiny experiment. That is a fair, honest way to describe the opportunity to a client.

The expansion builds on earlier launches across Europe and Southeast Asia and Taiwan. New coverage includes 14 more Latin American countries, following launches in Mexico and Brazil, plus markets across Europe, the Middle East, Africa, and Asia, including Nigeria, South Africa, Serbia, Ukraine, Cambodia, Bangladesh, and Pakistan. If you work with clients in those markets, check whether their country is covered before you pitch anything. Do not assume coverage from a headline alone.

Who actually sees ChatGPT Ads?

This is the detail that prevents a bad pitch. Ads appear only on the Free and Go plans. Plus, Pro, and Enterprise subscriptions remain ad-free, according to OpenAI. That means ChatGPT Ads are not reaching every ChatGPT user. They reach users on the free and lower cost tiers, while paid subscribers do not see ads. For a client, that changes the audience picture. It may still be a large audience, but it is not the same as all 1.2 billion weekly users.

OpenAI also states that it keeps conversations private from advertisers, does not sell customer data, keeps ads clearly labeled and separate from ChatGPT's answers, and says advertising does not influence the answers ChatGPT provides. Those points matter for client trust, especially for businesses that worry about privacy. Repeat them accurately, and do not add stronger privacy claims than OpenAI has made. If a client asks a detailed privacy question, point them to OpenAI's own post and Ads Manager information, rather than guessing.

A useful way to explain this to a client is simple. ChatGPT Ads put a clearly labeled ad in front of people who are already asking questions and comparing options. That is different from a social feed ad that interrupts browsing, and different from a search ad that appears after a short keyword. It is closer to a conversation moment, where the user has already stated a need. That can be valuable, but only if the landing page answers that need quickly.

What service should you actually sell?

Sell setup and testing, not magic. A clean ChatGPT Ads setup service has four parts. First, check eligibility and access, including country coverage and whether the client can use Ads Manager or needs a technology partner. Second, build or fix one landing page that matches the ad promise, with a clear next step, such as a booking, a quote request, or a purchase. Third, set up tracking so the client can see leads, sales, or bookings, not just clicks. Fourth, run a small, controlled test, then report what happened and what to change next.

Do not sell a vague promise like more sales, cheaper leads, or instant results. You do not control the auction, the audience, or the client's offer. You control the setup, the message, the landing page, and the test structure. Price the work you control. That protects you and the client, and it makes renewals easier, because the client can see what was done and what was learned.

A good first offer is a two week or four week test, not a long contract. The test should have a written goal, a fixed ad budget paid by the client, a start and end date, and a simple report at the end. If the test works, the client can continue. If it does not, you both learn why, and you can fix the offer, the page, or the targeting before spending more. That discipline is the real service.

How much can you charge? Planning examples only

The fees below are planning examples, not market rates and not OpenAI pricing. OpenAI has not published a cost per click or a minimum spend in the post used for this article, so this article does not invent one. Use the examples to plan your own pricing, then adjust for your experience, your country, and the time each job takes.

| Service package, example only | What it includes | Example fee, planning only | Client pays separately | | One week setup test | Access check, one campaign, one landing page review, basic tracking, short report | $300 to $500 one time | Ad spend, chosen by the client | | Four week test and report | Setup, two message tests, landing page fixes, weekly check, final report | $600 to $1,200 one time | Ad spend, chosen by the client | | Monthly management, after a test | Ongoing tests, page fixes, monthly report | $300 to $800 per month, example only | Ad spend, chosen by the client |

Now work through a realistic example. Suppose you sell a four week test for $800, and the client sets an ad budget of $500 for that period. Your revenue from that client in that period is $800, before your own costs and taxes. The client spends $1,300 in total, including ad spend. That does not mean the client makes a profit. If the client's product margin is low, or the landing page converts poorly, the test can lose money even if the ads run correctly. Always show the client the full cost, including ad spend, before you start.

Your own costs are mainly time. A setup test can take 5 to 10 hours for a simple business, and more if the landing page needs real fixes, tracking needs work, or the client needs many revisions. If you charge $500 for 10 hours of work, that is $50 per hour before costs, taxes, and unpaid sales time. If the same job takes 20 hours, it is $25 per hour. That math is why small, clear packages beat large, vague promises. Track your hours on the first three clients, then set prices that reflect real time.

How do you run a first test without wasting the budget?

Start with one offer and one audience. Do not test five products, five messages, and five pages at once, because you will not know what worked. Pick the client's best selling or easiest to explain offer, and write one clear promise for it. The landing page should repeat that promise, show proof, and ask for one action. Remove extra links, popups, and distractions for the test period.

Next, set a daily or total budget cap with the client, and agree on a stop rule. For example, stop and review if the test reaches the agreed spend without a single lead, booking, or sale, or if the landing page shows a clear problem. A stop rule is not a failure. It is how you avoid spending a full budget on a broken page. Also, check the results daily for the first few days, not just at the end. Early checks catch tracking errors, wrong links, and mismatched messages.

Finally, report in plain language. Show spend, leads or sales, cost per lead or sale if tracking allows, and what you changed. Then recommend one next step, such as fixing the page headline, changing the offer, testing a new message, or pausing the channel. Clients renew when they understand the result, even when the result is that this channel is not right for them yet. Honest reporting builds the next sale better than a polished chart that hides a weak result.

Which clients are a good fit, and which are not?

Good first clients are businesses with a clear offer, a working website, and a way to handle enquiries quickly. Local services, software tools, education providers, and ecommerce brands with a simple product can all be good test candidates, if they have a landing page that matches the ad. Businesses that already answer customer questions online, or that sell a product people research before buying, may fit well with a conversation based ad format.

Poor fits include businesses with no clear offer, no landing page, no tracking, and no time to reply to leads. Also be careful with regulated claims, such as health, finance, or legal promises, where advertising rules are stricter and mistakes are costly. Do not take a client whose main goal is to mislead users, hide fees, or make claims you cannot verify. A short test with a bad offer will not fix a bad offer, and it will damage your reputation faster than it builds income.

Ask three questions before you accept a client. What is the offer, and what does it cost? What happens after a click, and who replies? How will we know if the test worked? If the client cannot answer those questions, help them answer them first, or decline the test. That filter saves time and protects both sides.

What should you not promise?

Do not promise a cost per click, a cost per lead, a sales number, or a return on ad spend. OpenAI's expansion post does not publish those figures, and they vary by country, niche, offer, and landing page. Do not promise that ChatGPT Ads will beat Google or social ads, because you have not tested that for the client yet. Do not promise that Plus or Pro users will see the ads, because OpenAI says those subscriptions remain ad-free. Do not promise privacy outcomes beyond OpenAI's own statements.

Also, do not copy ad copy from competitors without checking it, and do not use client data in ways the client has not approved. Keep ad accounts, passwords, and billing details in the client's own accounts where possible, rather than in your personal accounts. That is safer for the client, and it makes handover simple if the relationship ends. Write down who owns the account, who pays for ads, and what happens to the campaign data at the end of the test.

The bottom line

ChatGPT Ads reaching 89 countries is a real opening, but the money is in careful setup and testing, not in hype. Offer a small, clear test, charge for work you control, show the full cost including ad spend, and report honestly. If the test works, you have a repeatable service. If it does not, you have saved the client from a larger spend, and you have a reason to fix the offer or try a different channel.

Explore the Make Money with AI hub, read How to Make Money With AI: 12 Real Ways for Beginners (2026), and see AI Affiliate Marketing for Beginners: A Simple Start (2026).

Sources: OpenAI, ChatGPT Ads expands to 26 new countries as over 100,000 advertisers have used the platform, 21 October 2026, at openai.com, and Ads Manager information at ads.openai.com.

Disclosure: Links in this article are informational. No affiliate commission is claimed for OpenAI or Ads Manager links. If you later add affiliate links to a related tool, disclose them clearly near the link.