ChatGPT Ads Manager: The Complete Guide to OpenAI's Ads Manager Beta (2026)

Short answer: The ChatGPT Ads Manager — officially OpenAI's Ads Manager Beta, reached by signing in at ads.openai.com — is OpenAI's self-serve platform for creating, launching, monitoring, and measuring the ads that appear below relevant ChatGPT conversations. You sign in with an OpenAI account, work a campaigns then ad groups then ads hierarchy, write plain-language context hints at the ad-group level, and read performance through table views, charts, and CSV exports. OpenAI describes it as a beta where some capabilities are limited today.

This page is the platform explainer: a navigable map of the dashboard and what each surface does. If you want the step-by-step signup, that lives in our separate5-step account setup guide. Here we stay inside the product — campaign building, reporting, and account settings — and stay strictly to what OpenAI's own documentation confirms, flagging everything else as a GPT Ads AI observation.

What is the ChatGPT Ads Manager (OpenAI Ads Manager Beta)?

The Ads Manager Beta is the dashboard where advertisers build and run ChatGPT Ads — the clearly labeled ad units that appear below relevant ChatGPT conversations. OpenAI frames the ad experience around three trust pillars: ads are "clearly labeled," kept "separate from answers," and give users "choice and control." The Ads Manager is the operator side of that promise — the tooling that turns a business offer into an eligible, measurable campaign.

Functionally it sits in the same family as Google Ads or Meta Ads Manager: a self-serve console with a campaign hierarchy, an auction, bidding, budgets, reporting, and account settings. What is different is the targeting model. Instead of exact-match keywords or interest audiences, ChatGPT Ads are selected by relevance to the context and intent of the conversation, guided by your context hints. OpenAI's selection system considers your context hints, landing page, ad title, and ad copy, then makes the final relevance call. We unpack the targeting model inChatGPT Ads explained.

How do I log in to the OpenAI Ads Manager?

You reach the dashboard at ads.openai.com and sign in with an OpenAI account. A work email is recommended so the account is tied to the business, not a personal address. That is the whole login mechanism — there is no separate "ads manager" password or consumer login page. Your OpenAI identity is your key.

The catch is that signing in only reaches the working dashboard once your business already has a verified, approved advertiser account. Per OpenAI's setup documentation, access requires:

• An OpenAI account — created during sign-up if you do not have one.
• An approved advertiser account — one per business, set up by a single account owner.
• Identity verification through Persona, OpenAI's verification partner, completed during onboarding.
• Approval from the rolling review queue — OpenAI emails the account once platform access is granted.

In short, "chatgpt ads manager login" is just signing into ads.openai.com with your OpenAI account — but the door only opens after approval. OpenAI publishes no review SLA; as a GPT Ads AI observation, plan for days to weeks rather than minutes. The full five-step setup (sign-up, onboarding plus verification, account info, billing plus payment, invite team) is documented in ouraccount setup guide. We keep this page on the platform itself.

Who can use it, and who sees the ads?

Two different questions matter here, and most write-ups conflate them. Keep "who can advertise" separate from "who sees the ads."

Market expansion: ChatGPT Ads began rolling out to select countries in Europe, India, the Middle East and North Africa in September 2026, the largest expansion since Brazil and Mexico. Treat the availability of any specific feature in a new market as something to confirm in your own account rather than to assume — the beta has consistently rolled capabilities out per-market rather than all at once. We log confirmed launches in the news timeline.

The platform hierarchy: campaigns then ad groups then ads

Everything in the Ads Manager hangs off a three-level tree. Knowing exactly what each level controls is the single most useful mental model for the dashboard.

An ad in ChatGPT Ads is built from six components, verbatim from OpenAI's spec: advertiser name, favicon (your logo), title (headline), copy (description), landing page URL, and image asset (the creative). Get those six right and the unit renders cleanly. The advertiser name and favicon come from your account info, so set them carefully during onboarding. For dissected example units, see ourguide to what ChatGPT Ads look like.

Where do context hints live in the Ads Manager?

Context hints live at the ad-group level. They are plain-language descriptions of the conversations, topics, or keywords where your offer is relevant — not exact-match keywords, and not a delivery guarantee. The system weighs them alongside your landing page, ad title, and ad copy, then makes the final relevance decision about whether to show your ad in any given conversation.

Because hints sit at the ad-group level, your ad-group structure is your targeting structure: one ad group per coherent theme, each with its own hints. This is the field that most rewards craft. We have a full field guide incontext hints for ChatGPT Ads, and a freecontext hint generator to draft them.

What does a clean ad-group build look like in practice?

A useful way to internalize the hierarchy is to walk one worked build. Say you sell a project-management tool for agencies. Inside a single Clicks-objective campaign, you might structure three ad groups, each a distinct theme with its own context hints and its own ads:

Three things make this build clean. First, each ad group owns one buyer situation, so the context hints stay coherent rather than blurring across themes. Second, the ad copy and landing page in each group match the hint — relevance is judged across hints, title, copy, and landing page together, so a mismatch between them dilutes every input. Third, keeping themes separate means reporting is legible: when you read average CPC, CTR, and conversions per ad group, you can tell which buyer situation actually pays, instead of averaging three different intents into one muddy number.

Creating campaigns: guided UI, bulk upload, or the plugin

OpenAI documents two paths to the same hierarchy inside Ads Manager — a guided UI flow and bulk upload — and, since September 2026, a third path that does not involve the dashboard at all: the ChatGPT Ads Manager plugin. Pick based on how many ad groups you are launching and how much of the build you want to describe rather than click.

Bulk upload is the unlock for scale. Because the template mirrors the same nested hierarchy as the guided flow, one file can define a whole matrix of ad groups and ads.

The ChatGPT Ads Manager plugin: running the account from a conversation

Launched in September 2026, the ChatGPT Ads Manager plugin puts the account inside ChatGPT itself — and inside Codex, which matters if your team already lives in a terminal. You describe what you want in plain language and it does the mechanical work: turning a website or a brief into ready-to-launch ads, generating variants, updating live campaigns, troubleshooting delivery, reviewing performance, and returning prioritised recommendations.

The design decision that matters most is the one about consent: the plugin previews recommended changes and asks for confirmation before applying them. It is an assistant with a handbrake, not an autopilot — nothing lands in a live account because a model decided it should.

How to install and connect it

1. Add ChatGPT Ads Manager from Plugins.
2. Connect your ad account when prompted — the plugin acts on the accounts your OpenAI login already has access to, with the same permissions your seat carries.
3. Start a conversation. "Set up Ads Manager for my business," "help me create a new ad," or a paste of your homepage URL are all valid opening moves.

What it can actually do

Custom audiences: matching, exclusions, and size

Custom audiences let you match your own first-party list against ChatGPT users, then target or exclude them. September 2026 rebuilt most of the ergonomics, and the changes matter more than they sound — they turn audiences from a build-once artefact into something you can maintain.

Performance monitoring and reporting

The reporting surface gives you several ways to read the same numbers: a table view for row-level detail across campaigns, ad groups, and ads; charts to see a metric move over time; and CSV export to pull the data into your own stack. The metrics OpenAI's documentation lists are these seven:

How do I export reports from Ads Manager Beta?

Use CSV export from the reporting view. Export pulls the same campaign, ad group, and ad rows you see in the table so you can feed them into a spreadsheet, BI tool, or your own attribution model. For ongoing measurement, set up conversion tracking inside Ads Manager Beta and let your UTM parameters carry the rest into your analytics.

Dynamic URL parameters and macros

Ads Manager Beta supports custom landing page query parameters built from your own text or from supported macros, populated automatically at delivery time. This is the feature that lets you join ChatGPT Ads data to your own analytics without string-matching campaign names, and most accounts we audit are not yet using it.

Configure them from the three-dot menu on the Campaigns, Ad Groups or Ads page — choose Edit campaign, Edit ad group or Edit ad, then use Landing page query parameters in the edit modal.

Conversion measurement and tracking

Conversion measurement is configured inside Ads Manager Beta. The portable tracking method OpenAI documents is parameter-based: UTM or static tracking parameters appended to your landing-page URLs persist through ad clicks, so your own analytics can attribute sessions and outcomes to ChatGPT Ads. Build a consistent convention before launch — our free UTM builder keeps it clean.

Pixel validation diagnostics

Ads Manager surfaces Pixel validation detail under Tools → Conversions, in the diagnostics panel on the right-hand side. It reports why events were dropped before being sent, which field was affected, the error type, and a recommended fix — which is the only practical way to catch the silent failures, like a badly normalised hash or an amount sent in major rather than minor units, that never throw a browser error. Put it in the weekly account review rather than saving it for when a number looks wrong.

One further change with a date attached: automatic advanced matching is now the default for new Web pixels, and OpenAI enables it for existing Web pixels on 17 August 2026. If you need to opt out or want more time for a privacy review, do it before then under Tools → Conversions → Data Source → Edit pixel. Full treatment in the conversion tracking guide.

Account settings tour: members, permissions, API keys, billing, change logs

Settings is where multi-seat teams and agencies actually live. OpenAI's documentation lists these surfaces, each doing one job.

A practical seat design we use on client accounts: a media buyer with campaign edit rights, an analyst with read-only reporting, and a finance contact with billing visibility. The account owner should sit inside the client business; the agency joins as a member. We detail this seat structure in theaccount setup guide.

Budgeting and objectives: how pricing and bids work

ChatGPT Ads run on a relevance-weighted, second-price auction. There is no rate card; your final cost depends on competition, the intent in the conversation, and the quality of your creative. You choose one of two objectives per campaign:

Practical guidance from our desk, framed as a GPT Ads AI observation: most direct-response advertisers start on the Clicks objective near OpenAI's recommended $3-5 max bid, then let observed average CPC and conversions guide adjustments. Because the auction is relevance-weighted, a sharper creative and tighter context hints can win impressions at a lower effective cost than simply bidding higher.

Should I pick the Reach or Clicks objective?

The objective decision is the first real fork in the Ads Manager, and it sets your pricing model for the life of the campaign. Use this as a decision rule:

• Choose Clicks (CPC) when you are measuring response — landing-page visits, sign-ups, purchases, demo requests. You pay per click, OpenAI recommends a $3-5 starting max bid, and every dollar maps to a session you can attribute with UTM parameters. This is where most performance advertisers should start.
• Choose Reach (CPM) when the goal is presence and awareness across relevant conversations rather than immediate clicks. You pay per thousand impressions, and OpenAI's default max bid is $60 CPM. Reach makes sense for brand launches, category-defining offers, or when you want share of voice in a topic regardless of click volume.

A GPT Ads AI observation on sequencing: because the auction rewards relevance, it is usually worth proving an offer can earn clicks profitably under the Clicks objective before scaling spend into Reach. A creative that cannot win clicks cheaply rarely becomes efficient just by switching to a CPM model. Note that whichever objective you choose, the underlying selection is the same — your context hints, landing page, ad title, and ad copy decide whether you are eligible for a given conversation at all.

Bid strategy: Maximize results vs manual bidding

The objective decides how you are priced. The bid strategy decides who sets the bid — and as of August 2026 the answer is "the platform, unless you say otherwise." Maximize results lets ChatGPT Ads set and adjust bids automatically, optimising toward your selected campaign goal to get as many results as possible from your budget. It is selected by default for eligible new ad groups, which means a large number of advertisers are now running automated bidding without having explicitly chosen it.

How do I set or change the bid strategy?

1. In Ads Manager Beta, create a campaign as normal.
2. Continue through the campaign setup flow to the ad group step — bid strategy is an ad-group setting, not a campaign one.
3. Review the Bid strategy section. Maximize results is pre-selected for eligible new ad groups.
4. To take manual control instead, select Manual: Max bid and set your ceiling.
5. Complete your ad group and ad setup, then launch. Revisit the setting per ad group as performance data arrives — it is not an account-wide policy and does not have to be consistent across ad groups.

A GPT Ads AI observation: we tend to open a new ad group on Maximize results to gather conversion signal quickly, then move the ad groups carrying our tightest efficiency targets onto manual ceilings once we can see what a result actually costs. Automated bidding works considerably better once it has data to work from, which is an argument for sequencing rather than for picking a side. More on how we structure this in our campaign management service.

Platform targeting: which ChatGPT surfaces you run on

Campaign targeting used to mean location and nothing else. It now includes platform targeting — a control over which ChatGPT surfaces your campaign is eligible to deliver on. In the campaign dialog, the Platforms multi-select dropdown lets you include one or more of:

• iOS App — the ChatGPT native app on iPhone and iPad.
• Android App — the ChatGPT native app on Android devices.
• Web — both desktop web and mobile web, grouped together as one selectable platform.

Platform targeting is optional; leaving it unset makes the campaign eligible across supported platforms. The cases where it earns its keep are concrete rather than theoretical: an app-install offer that only makes sense on the matching mobile OS, a desktop-only B2B workflow, or a landing experience that genuinely underperforms on small screens. Splitting by platform purely to "see the data" is worth less than it sounds while reporting stays grouped.

Brand safety and ad eligibility

One reason the Ads Manager feels more constrained than a mature network is intentional. OpenAI places ads only near chats it considers safe, appropriate, and aligned with user trust and brand safety, and a published set of Ads Policies governs which advertiser categories are eligible to run at all. For operators, this has two practical effects.

First, eligibility is a gate, not just a setting. If your category falls under the Ads Policies' restrictions, no amount of bidding or context-hint tuning will force delivery — the policy layer sits in front of the auction. Second, brand-safety placement is part of why context hints never guarantee delivery: even when your hints describe a conversation perfectly, the system can withhold an ad if the surrounding chat is not a safe, trust-aligned context. Plan creative and category positioning with the Ads Policies in mind before you build, rather than discovering a restriction after onboarding.

Beta limitations: what OpenAI confirms today

Honesty about the beta is the point. OpenAI states plainly that the Ads Manager Beta is an evolving product and that some capabilities are limited today. Here is a clean separation of what its documentation confirms versus what we treat as unconfirmed, as of June 2026.

The hard limit that will not change with a feature flag: context hints never guarantee delivery in any specific conversation. The system makes the final relevance decision. Plan around influence, not control.

Why isn't my ChatGPT Ads campaign delivering?

The most common cause is incomplete setup, not the auction. Walk this checklist:

1. Account info incomplete — advertiser name and logo must be set; ads cannot render without them.
2. Billing not finished — no payment method, no delivery.
3. Bids too low — under a relevance-weighted, second-price auction, an uncompetitive max bid wins few impressions.
4. Context hints too narrow — overly specific hints shrink the pool of eligible conversations.
5. Policy or category review — eligibility runs against OpenAI's published Ads Policies, which govern which categories can advertise.

How the ChatGPT Ads Manager compares to Google Ads and Meta Ads Manager

If you have run Google Ads or Meta Ads Manager, the scaffolding will feel familiar — and one thing will feel genuinely new.

The key reframe: there are no keywords, negative keywords, or match types in ChatGPT Ads. Targeting is conversational relevance expressed through context hints. Importing a keyword mindset wholesale is the most common strategic error we see from paid-search veterans.

Managing multiple advertisers and agency accounts

OpenAI's model is one advertiser account per business. To manage several advertisers, you set up separate accounts. The operational consequences:

• Set up a separate Ads Manager Beta account for each advertiser you manage.
• Each account has its own billing profile and payment method.
• You manage multiple advertisers through separate accounts, each with its own members, permissions, and change logs.

If running this across several brands sounds like a lot of overhead — it is, today. That is part of why a done-for-you ChatGPT Ads agency exists: to absorb the multi-account setup, the context-hint craft, and the day-to-day optimization while the platform matures. If you are already spending on paid acquisition and want this run for you,book a discovery call.

Platform facts in this guide are sourced from OpenAI's public documentation as of June 2026 — "Ads in ChatGPT: The Basics," "Create Ad Groups for ChatGPT," "Ads Manager Beta Account Setup," and "Ads Manager Beta Overview," together with ads.openai.com. Where this guide goes beyond those documents — review timing, file-format details, the maturity of measurement and bidding features, and CPC ranges — it is labeled a GPT Ads AI observation. CPC ranges are GPT Ads AI managed-account observations: anonymized aggregates from our roster, not OpenAI figures and not guarantees. GPT Ads AI is an independent agency, not affiliated with, endorsed by, or owned by OpenAI; "ChatGPT" is a trademark of OpenAI, used descriptively.