ChatGPT Ads Reportedly Start at $200,000. Most Clinics Still Don't Need Them.

ChatGPT started selling ads in February. The reported minimum buy is $200,000. If you run a medical clinic, that number probably answers the question before you finish asking it. According to Axios and other trade reporting, OpenAI began testing ads in ChatGPT for certain users in February 2026. ADWEEK reported a $200,000 minimum commitment for select advertisers in the early beta. That number tells you a lot about who this beta is designed for. Probably not your practice, and probably not most medical practices in Tijuana operating on normal budgets.

I say that as someone who runs marketing for VIDA Wellness & Beauty Center, a major medical tourism practice in Tijuana. When the ad announcement hit, my phone lit up. Surgeons, practice owners, even a couple of marketing directors at competing operations. Everyone wanted to know the same thing: Do we need to be on this? Is this the next Google Ads? Are we already behind?

For most practices, the answer right now is no. But the reason it's no is more interesting than the price tag. The doctors who are already showing up when patients ask ChatGPT "Who is the best bariatric surgeon in Tijuana?" or "best deep plane facelift near San Diego" didn't pay for ad placement to get those mentions. Many got there through stronger entity signals and better digital legibility. And based on current restrictions and tier coverage, ads are unlikely to reach many of the highest-intent patients doing that kind of research.

The Timeline

On January 16, 2026, OpenAI announced they would begin testing ads in ChatGPT. On February 9, the first ads went live for adult, logged-in users on the Free and Go subscription tiers. ADWEEK reported that OpenAI was asking select advertisers to commit at least $200,000 to participate in the beta. Reported launch advertisers included Best Buy, AT&T, Expedia, Adobe, Ford, Target, HelloFresh, and Shopify. Enterprise brands with eight-figure ad budgets.

OpenAI's own testing page describes how the system works: ads are matched to the topic of the conversation, appear at the bottom of ChatGPT's answers, and are clearly labeled as sponsored content. OpenAI states that ads don't influence the AI's actual answer. The Associated Press reported Altman's framing: "It is clear to us that a lot of people want to use a lot of AI and don't want to pay."

By March, Criteo became the first ad-tech partner integrating with the ChatGPT pilot. The ad infrastructure is clearly moving beyond rumor.

The headlines missed the actual economic driver behind this pivot. Roughly 18 months before this reported ad test, Altman told a Harvard audience that "ads-plus-AI is sort of uniquely unsettling to me" and called advertising "a last resort." The pivot from "last resort" to "let's test this" happened fast. It likely reflects pressure to monetize at scale as OpenAI reportedly projects significant losses despite rapid revenue growth. The economics likely made ad testing more attractive.

None of this is editorial. It's just the math of a company that needs to monetize hundreds of millions of free users. But a surgeon in Tijuana spending $4,500 a month on Meta isn't part of that math.

Five Reasons This Doesn't Apply to Medical Practices (Yet)

I keep saying "yet" because I don't think ads on AI platforms are permanently irrelevant. But right now, the structural barriers for healthcare providers are severe enough that you can safely ignore them while focusing on what actually works.

A reported $200K minimum locks out every clinic I've personally worked with. The largest medical tourism operations in Tijuana likely spend $15,000 to $50,000 per month on total digital marketing. Everything. Meta, Google, SEO, content, the whole stack. A $200,000 single ad buy represents four to sixteen months of total marketing budget for even the biggest players. Mid-tier practices with two to five surgeons spend $3,000 to $8,000 a month. For most of you, that's not a stretch budget. It's a different category entirely.

Ads don't show on paid tiers. OpenAI's testing materials indicated that certain paid tiers, including Plus, Pro, Business, Enterprise, and Education, would remain ad-free during the pilot. Only Free and Go users see them. The patients willing to pay $20 or $200 a month for better AI, the ones with disposable income, the ones making $5,000 to $15,000 elective surgery decisions, never see your ad.

Health topics are restricted. OpenAI's own policy states that ads are "not eligible to appear near sensitive or regulated topics like health, mental health or politics." In my view, this is less a product-timing issue and more a legal and compliance constraint informed by the FTC's Health Products Compliance Guidance (updated 2023) and the FDA's evolving stance on AI-generated health claims. That suggests it may not loosen quickly.

Most ChatGPT usage appears to be informational rather than directly transactional. OpenAI hasn't publicly broken out query intent in detail, but the observable pattern is clear. Medical tourism patients aren't typing "buy gastric sleeve." They're asking "Is bariatric surgery in Tijuana safe?" or "What should I look for in a facelift surgeon?" Those queries trigger organic entity citations, not ad placements.

Based on current beta documentation, there's no conversion tracking beyond impressions and clicks. No conversion attribution, no retargeting, no way to know if the person who saw your ad became a patient. For a $5,000 to $15,000 purchase decision with a weeks-long consideration window, that's functionally useless measurement.

Many of the Highest-Value Patients May Never See an Ad

This is the part I think matters most for your practice, and it's the single most important thing I can tell you.

Paid-tier users may skew more affluent and more digitally engaged, which could overlap significantly with the profile of elective-surgery researchers. Higher household income, higher tech literacy, higher willingness to pay for quality tools. Those traits likely overlap with many people researching an $8,000 gastric sleeve or a $12,000 deep plane facelift. They're doing serious comparative research. They're asking detailed questions. They're the exact patients you want.

And in the current pilot structure, those users are largely outside the ad-supported experience. By design.

Organic AI visibility doesn't have this limitation. When a patient on ChatGPT Pro asks "Who is the best bariatric surgeon in Tijuana for someone with a BMI of 40?" and the model cites a surgeon by name, referencing their procedure volume, credentials, and training lineage, it doesn't matter what tier they're on. Organic mentions can show up across both free and paid AI products, though not uniformly across every platform or query. One important caveat: AI outputs are non-deterministic and can vary by model version, prompt wording, account state, geography, and date. Still, the reach advantage over paid ads is substantial.

And it's not just ChatGPT. Organic entity visibility works across Gemini, Perplexity, Claude, and Copilot. Paid ChatGPT ads work on one platform, on two tiers, with health restrictions. The asymmetry is pretty clear at this point.

What We Saw at VIDA

I'm not going to rehash our entire GEO strategy here. We've written about the methodology in detail elsewhere. But the results are relevant because they show what organic AI visibility produces at costs that are a rounding error compared to $200,000.

Disclosure: VIDA offers GEO and AI visibility services. The results below come from our own practice and client work. I'm sharing them because they illustrate what's possible, not because every practice will see identical outcomes. Multiple factors contributed to each result, and none can be attributed to any single channel.

Dr. Alejandro Quiroz went from zero AI mentions to being cited by name when patients ask about deep plane facelifts in Tijuana. We didn't pay for ad placement to get those mentions. We built the entity signal: structured content about his training lineage (he studied under Dr. Andrew Jacono, per his published professional profile), physician schema markup, review enrichment with procedure-specific language, consistent NAP data across every platform where his name appears.

Over approximately 12 months during a broader repositioning effort that included AI visibility work, Dr. Carlos Castaneda's monthly surgical volume increased from roughly 5 to the mid-20s (full case study). The work centered on positioning, structure, and discoverability rather than paid media. Entity optimization across the board. These results reflect the combined effect of multiple channels working together.

In our plastic surgery campaigns during the same period, measured acquisition cost decreased from about $1,200 to $655. That's roughly a 45% reduction. For context, many healthcare marketers have seen acquisition costs rise materially since 2021, though the exact rate varies by specialty and source. We went the other direction. At 25 procedures per month with a $5,000 average, that $545 CAC reduction means roughly $163,000 in annual savings. Nearly enough to buy the reported ChatGPT ad minimum. Except the organic approach keeps producing patients next month too, and the month after that.

In Tijuana, patient coordinators run everything through WhatsApp. A typical coordinator handles 50 to 200 inbound leads per week, and a huge chunk of their time goes to qualifying people who were never going to convert. Price shoppers. People not medically eligible. People who won't cross the border. When a patient arrives because ChatGPT named their doctor, based on coordinator feedback, that lead is pre-qualified in ways no ad can replicate. They've already decided on the procedure, decided on Tijuana, and received what feels like a third-party recommendation. The coordinator conversation shifts from "Let me tell you about our practice" to "Let me help you book your dates." Our coordinators report that close rates on these leads tend to be meaningfully better, though we're still building enough volume for rigorous comparison.

The Economics, Side by Side

Look at the unit economics. $597 per month for 12 months is $7,164 per year. That's 3.6% of the reported ChatGPT ad minimum. If organic AI visibility produces one additional surgery per month at a $5,000 average, the annual return is $60,000 on $7,164 in spend. Even if it takes three months to see results, the modeled return would still be attractive.

And there's an economic dimension that goes beyond CAC. Medical tourism in Tijuana runs partly on a facilitator and broker economy. Online aggregators and San Diego-based referral networks charge 15 to 30 percent commissions on referred patients. When a patient finds your surgeon through an AI answer, they contact your practice directly. No facilitator. No commission. The effective revenue per patient increases even if you don't change your pricing. That's not a marketing win. That's a margin win.

Why the Window Is Actually Wider Than You Think

Many major platforms that initially resisted ads eventually embraced them. Gmail held out for three years. Netflix said "never" for fifteen years before launching an ad tier in 2022. Spotify waited until 2009. The pattern is often similar: organic-first participants who build presence before ad saturation enjoy years of uncontested visibility.

Google launched AdWords in 2000 with managed sales and a minimum buy. Self-serve came in 2002. It took another three to four years before small businesses meaningfully adopted it. The total cycle from launch to SMB saturation was roughly six years. Meta followed a similar trajectory from 2007 to 2013.

OpenAI is compressing this timeline. Things move faster now. But even in an accelerated scenario, self-serve ChatGPT ads probably arrive late 2026 or 2027. Meaningful adoption by medical practices? 2028 or 2029. Health ad restrictions loosening enough to matter? That depends on FTC and FDA rulemaking, which could take until 2027 to 2030.

The point isn't that ads will never matter. They will. But there's a two-to-four year window, minimum, where organic AI visibility is the only realistic option for medical practices. And organic assets built during that window can compound. Some patients acquired through AI discovery later leave reviews mentioning the doctor's name, procedure, and outcome. That review becomes retrievable content for future model responses. The next patient asking a similar question may see an even stronger entity signal. Each citation can reinforce the next one.

Ads don't have that same compounding dynamic. An impression consumed is gone. Organic visibility can compound over time, though the mechanism isn't as simple or predictable as a traditional ranking algorithm. It's directionally consistent with how retrieval-based AI systems tend to reinforce prominent entities, but it's not a guaranteed loop.

When ChatGPT Ads Will Matter

To keep this practical, here's what I'm watching next.

Self-serve opens (probably late 2026 or 2027). This is when smaller budgets can participate. Criteo's integration suggests the ad-tech plumbing is being built now. When self-serve launches, I expect CPMs to drop and minimums to disappear.

Health restrictions loosen (unclear timeline). This is the bigger gate. It's not a product decision. It's a legal and regulatory one. The FTC's health advertising guidance (updated 2023), the FDA's evolving AI position, and the uniquely blurred line between "ad" and "AI health advice" in a conversational interface all create liability that OpenAI has strong incentives to avoid. I wouldn't plan around this changing before 2028.

Measurement improves. Like any ad platform, OpenAI will likely face pressure to improve attribution. When they can offer conversion tracking comparable to Meta or Google, the channel becomes evaluable. Right now, spending $200K on impressions and clicks for a high-consideration surgery decision is hard to defend.

Commerce features mature. OpenAI has tested in-chat checkout with Shopify and Walmart. If they build a medical booking flow (a big if), the ad-to-booking pipeline could eventually work for practices.

We're preparing for that scenario, but today's economics still favor organic visibility.

The smart move is to build organic visibility now, so that when ads become accessible, you're already the entity the AI knows. You'll have a head start that paid placement can amplify rather than replace.

From 'Uniquely Unsettling' to 'Maybe They Don't Suck'

Altman's public comments on ads are worth tracking.

At Harvard in 2024: "Ads-plus-AI is sort of uniquely unsettling to me." He called it a last resort.

In January 2026, on X: "It is clear to us that a lot of people want to use a lot of AI and don't want to pay."

After the Super Bowl, when Anthropic ran ads mocking ChatGPT's ad model, Altman called them "clearly dishonest."

The shift from "uniquely unsettling" to "maybe they don't suck" to "stop criticizing us for running them" took about eighteen months. It's a notably fast pivot for a platform leader who had publicly framed ads as a last resort. It tells you something about the financial pressure OpenAI is under. And it tells you that the ad product is being built to serve OpenAI's revenue needs, not your patient acquisition needs. The advertisers writing $200K checks are subsidizing free users. You're not the customer. You're not even the target market.

The patients you want are paying $20 a month precisely to avoid being the product. Right now, the more durable way to reach them is to become part of the answer, not to hope this ad format eventually reaches them.

Your Move

Five things you can do this week that cost nothing.

Most clinics in Tijuana haven't tested their AI visibility. Most haven't checked whether their surgeons appear in a single AI answer. Most are still spending 100% of their budget on Google and Meta, channels where costs rise every quarter and where you rent attention instead of building it.

The reported $200,000 ad buy is a headline. It's designed for Fortune 500 brands with media agencies. It has nothing to do with a surgeon in Zona Rio trying to fill next month's OR schedule.

Open ChatGPT. Type "Best [your specialty] in Tijuana." See what happens. That result is the visibility layer that matters most right now.