ChatGPT ads are here. Any eligible business can now buy them right inside the app, and most of the coverage I’m seeing is pushing brands to move fast before the window closes.
My take, for most of the local businesses we work with, is more measured. Doctors have a phrase for this kind of situation: watchful waiting. You don’t ignore the thing, you monitor it closely and act the moment acting is warranted. That’s where we are with ChatGPT ads. For some clients, that moment is coming soon. For others, it’s further off. The right answer depends on your industry, your budget, and how much you need to measure every dollar. Here’s the full picture, what these ads cost, how they work, and how we’re thinking about who moves when.
A quick note: ChatGPT’s ad platform is changing fast, and prices, features, and eligibility rules have shifted repeatedly through 2026. The figures and policies below reflect what was publicly reported as of mid-2026 and are meant as general information, not advice for a specific campaign. Check OpenAI’s current advertising policies, or talk to us, before making decisions for your business.
What ChatGPT Ads Actually Are
OpenAI announced ads in ChatGPT in January 2026 and began showing them to US users on the Free and Go tiers on February 9. The rollout moved fast from there. What started as an invite-only pilot with a $200,000 minimum became a self-serve Ads Manager, and on May 5, 2026, OpenAI opened that platform to all US advertisers with no minimum spend at all.
A few things worth knowing about how they work:
- Who sees them. Ads appear only for users on the Free and Go plans. Plus, Pro, Business, Enterprise, and Education accounts stay ad-free, and OpenAI says ads are shown only to users 18 and older.
- Where they sit. In a labeled “Sponsored” block beneath ChatGPT’s answer, visually separated, not woven into the response itself. OpenAI says ads don’t change the actual answer you get.
- How targeting works. This is the big departure from Google Ads. There are no keywords. ChatGPT matches ads to the context of the conversation, using what you’re discussing in the current thread, basic context like general location and language, and, if the user has personalized ads enabled, their broader chat history and past ad interactions. You can layer on geographic targeting, but you can’t target by individual identity, and advertisers never see the actual conversations.
The pitch is genuinely compelling on paper. Someone chatting with ChatGPT about what to do after a basement flood, or how to plan a week in Colorado, is deeper into a decision than someone idly scrolling a feed. The intent is high.
Why We’re Cautious for Most Clients: The Reporting Isn’t There Yet
This is the heart of it, and it isn’t a matter of taste. It’s a concrete limitation, and it hits small budgets hardest.
The native dashboard is bare. For this early test, OpenAI gives advertisers only aggregated numbers: impressions, clicks, spend, click-through rate, average cost per click, average CPM, and conversions. There’s no search term report, no data on the actual prompts people typed, no demographic breakdown, and no detail on which conversations your ad appeared in. If you’ve ever optimized a Google Search campaign, you know that removes most of your levers. You can’t see what people actually asked, you can’t add negative keywords, and you can’t fully understand why your ad showed up.
Then there’s attribution, which is the part that would quietly cost a small business real money. ChatGPT is a mostly closed environment, so standard web tracking doesn’t carry over cleanly. OpenAI’s own guidance is that advertisers should add UTM parameters to their landing page URLs to track ChatGPT traffic in their existing analytics, because the platform won’t hand you that attribution automatically. Skip that step, and a chunk of your ChatGPT traffic can land in Google Analytics as Direct, and you’ll see conversions with no idea they came from ChatGPT.
It’s fixable. It takes UTM tagging on every landing page, a dedicated GA4 setup to isolate ChatGPT traffic, and conversion tracking through OpenAI’s measurement tools rather than eyeballing the dashboard. We know how to do all of that. But here’s the honest math for a small local advertiser: you’d be paying us to build measurement infrastructure so you can spend into a channel that still can’t tell you which searches triggered your ads or let you optimize against them. For a business with a bigger budget and real volume, that trade can be worth it. For a local business watching every dollar, it often isn’t. Not yet.
What ChatGPT Ads Cost
Pricing has moved faster in a few months than most platforms move in years, so treat these as a mid-2026 snapshot:
- Cost per click: commonly reported around $3 to $5, with bids under about $3 often failing to deliver in competitive contexts.
- CPM (cost per thousand impressions): reported roughly in the $25 to $60 range, softening toward the lower end as more inventory has come online.
- Minimum spend: none, as of the May 2026 self-serve opening. The old $200,000 pilot minimum is gone, so you can start with a modest daily budget.
Costs are relatively soft right now because competition is still building, and they’re more likely to rise than fall as advertisers pile in. That “get in while it’s cheap” argument is real, and it’s the main case for testing early. I just don’t think cheap clicks you can’t properly measure are a good deal for a business that can’t absorb the learning cost.
Our Recommendation Depends on Your Industry and Budget
Here’s the nuance that a blanket “wait” misses. This isn’t one-size-fits-all, and I don’t want to pretend it is.
Take travel and experiences, which is one of the consumer categories OpenAI explicitly supports right now. Trip planning is one of the most natural fits imaginable for this format: people work through destinations, dates, budgets, and lodging in long, detailed conversations, exactly the high-intent moments where a well-placed ad can earn its keep. For a travel or tourism client with a larger budget and the volume to make attribution setup worthwhile, I expect we’ll test ChatGPT ads relatively soon. The category fit is strong and the measurement work pays for itself faster when there’s real spend behind it.
For a smaller local business on a tighter budget, the calculus flips. The same reporting gaps that a larger advertiser can absorb become a real problem when every dollar has to be accounted for, and the cost of building proper tracking is harder to justify against a small test budget. For most of these clients, watchful waiting is the right call, at least until the reporting matures.
So our posture is deliberately split: move earlier where the category fit is strong and the budget supports proper measurement, wait where it doesn’t. Same platform, different timing, based on what actually serves the client.
How We’re Approaching It: Watch, Test, Roll Out
Watchful waiting isn’t passive. Here’s what it actually looks like day to day, and how we expect it to move.
Right now, we’re in the watching phase, and it’s active work. Two things are happening in parallel. First, we’re reviewing what other advertisers are reporting: who’s seeing results, in which categories, at what cost, and where the platform is falling short. A lot is being published as brands run early campaigns, and there’s no reason to pay for lessons someone else has already paid for. Second, we’re doing our own hands-on testing and monitoring of the platform, so that when we recommend it to a client, it’s from direct experience, not a press release.
In the next month or two, we expect to move into fuller testing for the right clients. As the platform’s reporting matures and we’ve validated our measurement setup, we’ll start running real campaigns, beginning where the fit is strongest. That’s likely to be travel and experiences clients with the budget to test properly, since the category is a natural match and the spend justifies the tracking work.
From there, we roll out more broadly as it proves out, and the timing will differ by industry. A strong-fit, well-funded category moves first. A small local business on a tight budget waits longer, until the reporting is good enough that a modest test can actually be measured. We’re not going to put a client’s budget into a channel we can’t yet prove is working for them, and we’re not going to make every client wait just because some should. Same platform, different timing, based on what each business actually needs.
Who’s Even Allowed to Advertise
Category eligibility is its own filter, and it matters for a lot of local businesses.
During this initial test period, OpenAI’s ad policy limits ads primarily to consumer verticals: lifestyle and household goods, local services, travel and experiences, and digital products or education. That’s good news for many local businesses, and for travel and tourism in particular. Home services trades like plumbing, HVAC, roofing, electrical, and landscaping generally fall under local services.
The regulated categories are more restricted. Health, mental health, financial services, and legal services all face additional limits. Based on OpenAI’s published policies as of mid-2026, these categories are disallowed by default, though OpenAI says it may approve advertisers within financial services, healthcare, and legal services on a case-by-case basis through manual review. Eligibility here has been changing over time, so if you operate in one of these areas, it’s worth confirming your current status directly against OpenAI’s advertising policies before planning any spend. For many law firms, financial advisors, and healthcare providers, paid ads remain limited today, which is one reason we point these clients toward AI search visibility as the more dependable route.
Watchful Waiting Doesn’t Mean Sitting Still
Here’s the part people miss: choosing not to run ads doesn’t mean sitting out ChatGPT. The more durable opportunity is making sure your business shows up when people ask ChatGPT, Claude, and Google’s AI Overviews for a recommendation in the first place. That’s AI search optimization, and unlike a paid placement, an earned mention doesn’t disappear the moment your budget stops. It compounds.
For most of our clients, that’s simply the better use of the same money today. It builds visibility in the exact AI tools everyone’s worried about, it works across every engine at once rather than just ChatGPT’s ad slot, and it doesn’t require pouring spend into a channel you can’t yet fully measure. If you want a starting point, we walk through the fundamentals in our guide to making your small business visible on ChatGPT, Claude, and Google AI Overviews.
Where to Go From Here
ChatGPT ads are real, they’re live, and for the right advertiser with the right budget and category, there’s a genuine first-mover argument. For a travel client with the budget to test properly, that moment may be close. For most small and local businesses on tighter budgets, the lack of query-level reporting makes early adoption a poor trade right now, and the smarter money goes into AI search visibility.
That’s watchful waiting in practice. We have the platform figured out, we’re tracking every policy and feature change, and the day the reporting matures enough to justify your spend, we’ll be ready to move, with the measurement already built to prove it’s working.
If you’re wondering how your business should show up in ChatGPT and other AI tools, that’s exactly the kind of work we do. Schedule a discovery call to talk it through.
Frequently Asked Questions
Can businesses advertise on ChatGPT in 2026?
Yes. As of 2026, eligible US businesses can advertise on ChatGPT through OpenAI’s self-serve Ads Manager, which opened with no minimum spend requirement in May 2026. Ads appear as labeled “Sponsored” blocks beneath ChatGPT’s responses for users on the Free and Go plans. Eligibility depends on your category: consumer verticals like local services, travel and experiences, retail, and education can advertise now, while legal, financial, and health services are restricted.
How much do ChatGPT ads cost?
As of mid-2026, ChatGPT ads have generally been reported in the range of roughly $3 to $5 per click, or about $25 to $60 per thousand impressions. There is no minimum spend, so a small business can start with a modest daily budget. Pricing is set by auction and varies by category, and because competition is still building, costs may rise as more advertisers join.
Should a small local business run ChatGPT ads right now?
For most small local businesses on tight budgets, we recommend waiting. The platform’s reporting is still limited, with no search-term data and clicks that can misattribute in Google Analytics, which makes results hard to measure and optimize, exactly what a small budget needs most. Businesses in strong-fit categories like travel, with larger budgets that support proper measurement setup, may be in a better position to test sooner. In the meantime, AI search optimization helps your business appear in ChatGPT’s organic recommendations without paying per click.
How is ChatGPT advertising different from Google Ads?
The core difference is targeting. Google Ads match to the keywords someone types, while ChatGPT ads match to the context of an ongoing conversation, the topic and, where enabled, chat history, rather than keywords. ChatGPT also offers far less reporting detail than Google Ads right now, with no search-term report and only aggregated performance metrics during its test phase.
Can law firms advertise on ChatGPT?
For legal services, availability is limited as of mid-2026. Based on OpenAI’s published ad policies, legal services is a restricted category that is disallowed by default, though OpenAI says it may approve some advertisers in regulated categories through case-by-case manual review. These rules have been changing, so law firms should confirm their current status against OpenAI’s advertising policies before planning any spend. In the meantime, the more dependable path to ChatGPT visibility for most firms is answer engine optimization.
What is answer engine optimization (AEO) and why does it matter for ChatGPT?
Answer engine optimization is the practice of structuring your content and online presence so AI tools like ChatGPT, Claude, and Google’s AI Overviews recommend your business when people ask for options. It matters because a growing share of buyers now research through AI chat instead of a search bar, and unlike a paid ad, an earned AI recommendation doesn’t disappear when your budget stops. For businesses that can’t or shouldn’t run ChatGPT ads yet, AEO is the most reliable way to show up in those conversations.
Sources: Figures and policy details in this post draw on OpenAI’s published advertising policies and reporting from Search Engine Journal, Search Engine Land, CNBC, and Reuters, as of mid-2026.
This post reflects information available as of its publication date. ChatGPT’s advertising platform, pricing, and eligibility policies change frequently, and details may be out of date by the time you read this. Nothing here is legal, financial, or professional advice. Confirm current terms against OpenAI’s official advertising policies before making decisions for your business.