Direct answer: OpenAI Ads campaign types are the two live objectives inside the self-serve ChatGPT Ads Manager: Reach, bought on a CPM basis for awareness, and Clicks, bought on a CPC basis for direct response. A Conversions bidding objective is reported as in development. Each campaign holds ad groups and ads, takes a daily or lifetime budget, runs on a max CPM or max CPC bid, and schedules by date and time.
Key takeaways
- Two live campaign types: Reach (CPM) for brand and launches; Clicks (CPC) for lead gen, trials, e-commerce, and demos. A Conversions/CPA bidding objective is reported as in development, per PPC Land.
- Structure is three tiers: Campaign → Ad Group → Ad, where the objective and budget live at the campaign level and context hints live at the ad-group level.
- Set CPC bids at $3-$5. OpenAI recommends that range; bids under $3 reportedly returned zero impressions due to delivery thresholds.
- Budget by campaign shape: daily caps for always-on programs, lifetime caps for fixed flights. Daily budgets are only available on newly created campaigns.
- Use the break-even CTR rule (around 1.05% at ~$42 CPM / ~$4 CPC) to decide when CPM beats CPC, instead of guessing.
- Abmatic AI manages every lever above in one console alongside LinkedIn and Google Ads, two-way synced with OpenAI Ads.
What are the OpenAI Ads campaign types?
OpenAI Ads campaign types are the buying models you choose when you create a campaign, and each is tied to one objective. If you have run Google or LinkedIn campaigns the structure will feel familiar, but the types map to a different matching model. There are two ways to buy today, plus a third reported in development. The table below is the fast orientation; the sections that follow go deep on each lever.
| Campaign type | Objective | Bid model | Best for | You pay for |
|---|---|---|---|---|
| Reach | Awareness | CPM (per 1,000 impressions) | Brand, launches, conquest | Impressions |
| Clicks | Direct response | CPC (per click) | Lead gen, trials, demos, e-commerce | Clicks only |
| Conversions | Reported in development | CPA (reported, later 2026) | Pipeline / purchase optimization | Not yet a bidding option |
Takeaway: pick Reach to pay for visibility and Clicks to pay only for action; a Conversions bidding objective is not yet live.
How is an OpenAI Ads campaign structured?
An OpenAI Ads campaign uses a three-tier structure, Campaign → Ad Group → Ad, confirmed by PPC Land, which describes the campaign as defining the objective, budget, and dates, the ad group as a theme-based collection, and the ad as the individual creative unit. What sits inside each tier is what differs from legacy platforms.
The campaign sets the objective (Reach or Clicks) and the budget. The ad group holds your geo and your context hints, the topical signals that tell OpenAI which ChatGPT conversations are relevant. The ad is the sponsored chat card itself: title, body, landing URL, and image.
The deeper difference is the matching model. A context hint is a topical signal, not a keyword or an audience list: OpenAI matches on the topics in the user's live ChatGPT thread when that thread shows commercial intent, per the OpenAI Help Center. So your campaign type is a decision about which commercial conversations you want to appear in, not which people you want to chase. We cover that matching layer in the OpenAI Ads targeting guide; this playbook stays on objectives, budget, bidding, and scheduling.
Takeaway: the objective and budget live on the campaign, the context hints on the ad group, and the chat-card creative on the ad.
Reach vs Clicks: which OpenAI Ads objective should you pick?
OpenAI launched CPC bidding in May 2026 alongside the existing CPM model, per OpenAI's own announcement and corroborated by MediaPost. That gives you two campaign types, mapped to two objectives.
When should I use the Reach (CPM) objective?
The Reach objective is the CPM-based campaign type that charges per thousand impressions, regardless of clicks. Choose Reach when the goal is visibility: brand awareness, a product launch, or competitive conquest where you want to appear in a category conversation even if the user does not click today. OpenAI charged a $60 CPM at the February 2026 launch, and clearing CPMs had eroded to as low as roughly $25 within about ten weeks, per The Next Web.
When should I use the Clicks (CPC) objective?
The Clicks objective is the CPC-based campaign type that charges only when a user clicks. Choose Clicks when you want measurable action: lead gen, free trials, e-commerce, or demo requests, because delivery without engagement costs nothing. Search Engine Land framed the CPC turn-on as OpenAI's move from impression-based to performance-driven advertising, tying spend to user actions as CPMs eroded.
What is the Conversions objective and when does it arrive?
A Conversions bidding objective is a reported third campaign type, with CPA bidding described as in development for later 2026; OpenAI has already shipped conversion measurement tooling separately. Until a CPA bidding objective ships, the Clicks objective is the closest performance proxy. The auction underneath all of these is a relevance-weighted, second-price mechanism, so a more relevant ad can win at a lower effective price.
When does CPM beat CPC? Use the break-even CTR
Most guides stop at "bid $3 to $5." That tells you the number but not which campaign type to run. The honest answer is a single line of math.
CPM charges you per impression; CPC charges you per click. Your click-through rate (CTR) is what converts one into the other. The break-even CTR is the click-through rate at which a thousand impressions on CPM cost the same as the clicks those impressions would generate on CPC.
At a working CPM around $42 and a CPC around $4, the break-even lands near 1.05% CTR (because $42 buys 1,000 impressions, and at $4 per click that same $42 buys about 10.5 clicks, which is 1.05% of 1,000). Below that CTR, CPC is cheaper per click. Above it, CPM wins, because you are paying a flat impression rate while harvesting more clicks for free. Plug your own clearing CPM and CPC into the same formula; the verified reference points for those inputs are the OpenAI bid figures reported by PPC Land and The Next Web.
The practical rule: if you have no CTR history, start on CPC. You cannot overpay for impressions that do not earn clicks. Switch to CPM only after about two consecutive weeks of CTR data confirms you clear the break-even.
Takeaway: start on CPC by default; only move to CPM once two weeks of data shows your CTR is above ~1.05%.
How much should you bid on OpenAI Ads?
OpenAI recommends a starting maximum CPC bid of $3 to $5 per click, per PPC Land. Advertisers who bid under $3 reported zero impressions because the bid fell below OpenAI's delivery thresholds. On the CPM side, the launch default was $60 while clearing rates eroded toward roughly $25, per The Next Web. Because the auction is second-price and relevance-weighted, your max bid is a ceiling, not what you pay; a tight, relevant ad can clear well below it.
Tier your starting bids by what the campaign is trying to do. The table below is a working starting grid, not a guarantee. Adjust on delivery and CTR data.
| Goal | Objective | Bid model | Starting max bid |
|---|---|---|---|
| Lead generation | Clicks | CPC | $4-$5 |
| Free trials | Clicks | CPC | $3-$4 |
| E-commerce | Clicks | CPC | $3-$5 |
| Demo requests | Clicks | CPC | $4-$5 |
| Brand awareness | Reach | CPM | $30-$45 |
| Product launch | Reach | CPM | $35-$50 |
One context note: the self-serve Ads Manager's minimum spend was cut to $50,000, down from $250,000 at launch, per PPC Land. That lowers the entry bar but still rewards disciplined bidding over a blunt budget dump.
Takeaway: open CPC at $3-$5, treat the bid as a relevance-weighted ceiling, and never bid under $3 or delivery stalls.
Skip the manual work
Abmatic AI runs targets, sequences, ads, meetings, and attribution autonomously. One platform replaces 9 tools.
See the demo →Daily vs lifetime budgets: which should you choose?
OpenAI added daily budgets alongside lifetime budgets in its May 22, 2026 update, per PPC Land. The choice is not cosmetic; it should follow the shape of the campaign type you picked.
A daily budget is a steady day-to-day spending ceiling that fits always-on programs: evergreen lead gen, category awareness, retargeting. It keeps spend from spiking. Important operational detail: daily budgets are only available on newly created campaigns, so you cannot retrofit one onto an existing campaign.
A lifetime budget is a single total the platform paces across a fixed run window, which fits fixed flights: a product launch, an event push, a quarter-end test. You give the platform a total and a run window, and it paces spend across the flight, stopping when the budget is exhausted. Pair a lifetime budget with a hard end date so spend self-terminates without manual babysitting.
| Dimension | Daily budget | Lifetime budget |
|---|---|---|
| Best for | Always-on (lead gen, retargeting, category awareness) | Fixed flights (launch, event, time-boxed test) |
| Pacing | Steady daily ceiling | Spread across the run window |
| Stops when | Each day's cap is hit | Total is exhausted or end date passes |
| Availability | New campaigns only | All campaigns |
| Pair with | Open-ended schedule | Hard start/end date and time |
Takeaway: daily budgets suit open-ended always-on programs (new campaigns only); lifetime budgets suit time-boxed flights with a hard end date.
How does OpenAI Ads scheduling work?
OpenAI Ads scheduling supports start and end date and time, not just dates. That precision matters for B2B, where commercial-intent conversations cluster in working hours and weekdays. Align flights with your buying windows, and pair any lifetime budget with a hard end so spend cannot run past the flight.
A simple pattern: launch and event campaigns get a lifetime budget plus a fixed start and end time. Evergreen programs get a daily budget and an open-ended schedule you review weekly. That keeps the two budget models doing what each does best without manual cleanup.
Takeaway: schedule to the hour, and bind every lifetime flight to a hard end time so it self-terminates.
How should you structure ad groups?
Build each ad group around a single coherent set of commercial conversations. Because OpenAI matches on the topics of the live thread, a sprawling ad group with mixed context hints dilutes relevance, and relevance is half of a second-price auction. Split distinct buyer topics into separate ad groups so each can carry its own context hints and its own creative.
Run a faster optimization cadence than legacy channels: weekly pacing reviews, monthly reallocation on performance data, and quarterly strategy resets. Quarterly-only budgeting is too slow for a platform that shipped CPC, daily budgets, geo-targeting, and dynamic CTAs in a single quarter. Dynamic CTAs auto-select from Shop Now, Book Now, Sign Up, and Learn More, per PPC Land; creative detail lives in the chat-card creative guide.
On budget allocation while you are still learning the channel, a common starting point is 10-15% of total paid spend during testing, scaling on performance, with roughly a 70/30 split between proven topic clusters and experimentation.
Takeaway: one tight topic cluster per ad group, then review weekly, reallocate monthly, and reset strategy quarterly.
Why structure this now? The audience is already arriving
The reason to stand up a campaign at all: ChatGPT is already sending B2B referral traffic organically, and it overtook LinkedIn in 2026, see why advertise on ChatGPT for the evidence.
The practical read: structuring a Clicks or Reach campaign now buys into a channel where your buyers are already showing up on their own.
How does Abmatic AI manage campaign structure?
Abmatic AI ships a first-class, two-way, live-synced OpenAI Ads integration that runs in the same console as LinkedIn Ads and Google Ads. We built it directly against the live OpenAI Ads API, and you connect by pasting an OpenAI Ads API key, validated live. From there you manage the full hierarchy: campaigns, ad groups, ads, objective (reach vs clicks), geo, context hints, budget (daily or lifetime), max-bid control (max CPM or max CPC), and start/end scheduling by date and time. Per-day impressions, clicks, conversions, and spend flow back into one performance view.
Every change is idempotent: re-saving updates a campaign rather than duplicating it, and one-click import pulls in campaigns built natively in the OpenAI portal, deduped 1:1, with OpenAI staying the source of truth. Pause or activate in Abmatic AI and it flips live on OpenAI, and the reverse holds too. Abmatic AI replaces point tools like 6sense, Demandbase, Mutiny, and Qualified, and pipes results into your CRM and marketing automation such as Salesforce, HubSpot, and Marketo.
There are three ways to run it: self-service in the platform, managed service where our paid-media team runs strategy and pacing, or fully AI-generated and AI-managed campaigns. The autonomous path is the structural edge: Abmatic AI's optimization engine queries the AI engines to see where you do and do not appear, cross-references your organic rankings to find topics where you have demand but are not surfacing, and builds campaign structure to buy ads exactly where you do not show up organically. Because that audience is already arriving purely organically, this is where we aim to deliver better ROI than Google or LinkedIn over time, a forward-looking promise tied to our expertise rather than a measured result yet. That logic is detailed in letting AI run your ChatGPT ads. Want to see your own gaps mapped to campaign structure? Book a demo.
For the broader shift this sits inside, see ChatGPT Ads explained and the step-by-step ChatGPT Ads setup guide.
FAQ
What campaign types does OpenAI Ads support in 2026?
Two campaign types are live in the self-serve Ads Manager: Reach, bought on a CPM basis for awareness, and Clicks, bought on a CPC basis for direct response. A Conversions/CPA bidding objective is reported as in development for later in 2026, per PPC Land.
Should I use CPM or CPC for OpenAI Ads?
Use CPC (the Clicks objective) for direct response such as lead gen, trials, and demos, and CPM (the Reach objective) for awareness and launches. The break-even is roughly 1.05% CTR at ~$42 CPM and ~$4 CPC: below it CPC is cheaper, above it CPM wins. Start on CPC if you have no CTR history, then reassess after about two weeks of data.
How much should I bid on ChatGPT Ads?
OpenAI recommends a starting maximum CPC bid of $3 to $5 per click. Bids under $3 reportedly received zero impressions because they fell below delivery thresholds. On CPM, the launch default was $60 while clearing rates eroded toward roughly $25, per The Next Web. The auction is second-price and relevance-weighted, so you often pay less than your ceiling.
What is the difference between daily and lifetime budgets?
A daily budget sets a steady day-to-day ceiling and suits always-on programs; it is available only on newly created campaigns. A lifetime budget gives the platform a total to pace across a fixed run window and suits launches, events, and time-boxed tests. Pair lifetime budgets with a hard end date so spend self-terminates.
Who actually sees OpenAI ads?
Ads appear as a clearly labeled sponsored placement at the bottom of a ChatGPT response on commercial-intent conversations, shown to logged-in adult users on the Free and Go tiers only. Plus, Pro, Business, Enterprise, and Education tiers do not show ads, per OpenAI and TechCrunch.
Can I target outside the US with OpenAI Ads?
The self-serve Ads Manager's geo-targeting was United States only at this stage, by state, DMA, or ZIP, per PPC Land. Where ads are shown to users is expanding separately. Abmatic AI's OpenAI Ads integration supports advertiser geo-targeting in the US, Canada, Australia, and New Zealand today; see the targeting guide for detail.
