ChatGPT Ads Account Structure and Scaling Guide (2026)

How Is a ChatGPT Ads Account Structured?

A ChatGPT Ads account has four layers: account, campaign, ad group, and ad. Every setting lives at exactly one layer, so the layer you put something on decides what you can budget, test, and report on independently.

OpenAI's own guidance sets the split. Its bidding documentation says to use separate campaigns when you need separate budgets, and the campaign help page recommends separate campaigns for meaningfully different objectives, regions, or product categories. The ad group help page recommends separate ad groups when use cases require different messaging or landing pages. The account limits of 5,000 campaigns, 5,000 ad groups, and 5,000 ads come from the Bulk API documentation and the bulk upload checklist, which calls each one an account limit.

Creative specs are tight. Titles run up to 50 characters (OpenAI recommends 16 to 24), body copy up to 100 characters (32 to 48 recommended), and images must be square. Bulk upload accepts PNG or JPG up to 1200x1200, and the API documentation calls for images of at least 640x640. If you are still setting up, start with our ChatGPT Ads Manager self-serve setup guide.

How Many ChatGPT Ad Accounts Can You Create?

You can be a member of any number of ChatGPT ad accounts, but once you belong to 10 or more (accounts you created plus accounts you joined by invitation), you cannot create another one yourself. You create additional accounts from an existing eligible self-serve account that uses credit-card billing, through Create account in the account menu.

The details from OpenAI's account setup article, checked as of October 23, 2026:

• The limit counts membership, not creations. If you created 2 accounts and joined 8 others, you belong to 10 and cannot create an eleventh. An admin of another account can still invite you.
• Eligibility comes from a card-billed account. Create account appears for eligible self-serve advertisers on credit-card billing, subject to organization permissions and remaining creation capacity.
• New accounts do not skip review. Creating an additional account does not bypass verification, account review, or other eligibility requirements.
• One advertiser per account. OpenAI tells anyone managing multiple advertisers to create a separate account for each. One OpenAI login switches between all the accounts you belong to.
• Country, currency, and time zone are permanent. Decide them before you submit. OpenAI's billing article adds that billing under a different country, region, or legal entity requires a separate advertiser account.

Since September 25, 2026, every account carries three names. The Account Name is internal and changes freely. The Brand Name is the public name shown on your ads, and changing it triggers brand review and pauses serving until approval. The Legal Name is used for billing and business verification, and changing it restarts business review. Put your internal naming conventions in the Account Name and campaign names, never in the Brand Name.

Agency rule of structure: one account per advertiser, and one account per billing country and currency for each advertiser. Never split one advertiser across accounts to test the same audience against itself; that only fragments learning and reporting.

Who Can Access a ChatGPT Ads Account?

Each ad account grants three roles: Admin (full access), Member (create and manage campaigns), and Viewer (view only). Enterprise teams add tenant-level control on top, where a global admin manages identity, SSO, and directory sync and an Ads admin creates eligible accounts without automatic access to existing ones.

From OpenAI's identity and access article:

• Highest role wins. A person receives the highest role granted directly or through a group (Admin, then Member, then Viewer). Removing a direct role does not reduce access if a group still grants a higher one.
• Every account keeps an admin. Each advertising account must retain at least one effective account admin, so assign a backup before you remove the last one.
• SSO is set per product. The Ads SSO policy is Required, Optional, or Off for the selected tenant. Turning on SSO for another OpenAI product does not enable it for Ads.
• SCIM is tenant-wide. Ads uses a tenant-wide SCIM directory, not one per ad account. A synchronized group assigned to an ad account starts with the Member role, and an admin can change it.
• ChatGPT roles do not carry over. A ChatGPT workspace or API Platform role grants no Ads permissions. Assign roles separately for each ad account.

A clean access model: two Admins per account (the owner and a named backup), Members for the people and partners who build and optimize campaigns, and Viewers for finance, leadership, and analysts. When someone leaves, remove the direct role and the group membership in your identity provider, then confirm no other access path remains.

Why Does Google Ads Account Structure Fail on ChatGPT?

Google structure fails because ChatGPT does not run a keyword auction. According to OpenAI, ads are selected in a relevance-weighted, second-price auction that considers the context and intent of the conversation, your landing page, title, copy, context hints, and, when ads personalization is enabled, select signals from a user's broader ChatGPT experience.

Every marketer arriving from Google Ads brings the same instinct: build campaigns around keyword themes, split ad groups by match type, and stuff headlines with search terms. On ChatGPT that instinct burns budget. A Google user types “best CRM for small business.” A ChatGPT user spends five minutes describing a 12-person sales team, frustration with Salesforce pricing, and a need for email integration. The signal is richer and messier than a query, and OpenAI states plainly that context hints are not exact-match keywords.

That makes Single Keyword Ad Groups pointless. Fifty ad groups that each carry one keyword-style hint split your budget fifty ways without adding relevance. The concept that replaces the keyword group is the intent cluster: a set of related conversational intents at the same stage of the buyer's journey. Instead of grouping “CRM pricing,” “CRM cost,” and “CRM subscription fee,” you build one ad group for “evaluating CRM options for a growing sales team” and describe that situation with several specific hints. The shift is from lexical matching to semantic matching, and the account structure has to follow it. For the strategy behind clusters, read our ChatGPT Ads targeting guide.

Which Campaign Objective Should Each Funnel Stage Use?

Map awareness campaigns to Views (CPM), consideration campaigns to Clicks (CPC), and decision campaigns to Conversions once your Pixel or Conversions API sends a standard conversion event. Keep objectives in separate campaigns, because OpenAI does not let you change a campaign's objective after creation.

The rules that shape this structure, from OpenAI's conversion-optimized campaigns article and the bidding documentation:

• Conversions optimizes toward one standard event. Purchases, sign-ups, and leads qualify; custom events cannot be optimization goals. You never pay per conversion: oCPC bills valid clicks and, for eligible advertisers, oCPM bills impressions. On oCPC, the max bid you enter acts as a CPA bid.
• Maximize results needs a daily budget. Maximize clicks and Maximize conversions both require a daily campaign budget and cannot use audience bid multipliers. Campaign-total budgets go with fixed bidding.
• Start CPC bids at $3 to $5. OpenAI recommends a starting maximum bid of $3 to $5 per click for CPC campaigns, and Ads Manager can show bid-strength guidance.
• CPM and conversion bidding depend on account access. The Bulk API documentation notes that CPM bids require account access, and conversion campaigns require conversion bidding to be enabled for the ad account.

Naming convention: use [Stage]_[Product]_[IntentCluster]_[Objective]_[YYYYMM] across every campaign and ad group, for example Consider_Acme_CompareCRMs_Clicks_202610 or Decide_Acme_Migration_Conv_202610. At 50+ ad groups, consistent names are the only way to filter and compare performance without opening each one. For the bidding trade-offs in depth, see our CPC vs CPM bidding guide, and set up events with the Pixel and Conversions API guide.

How Many Context Hints Should an Ad Group Have?

Use several related hints per ad group, not one. OpenAI allows up to 2,000 context hints per ad group, and each ad group should cover one intent cluster with hints that describe what you offer, who it helps, and when it is useful.

OpenAI's targeting documentation says hints describe relevant products, use cases, or needs that the creative and landing page do not fully cover, and that you do not need to fill the limit. Write clear, natural phrases focused on one idea: “cushioned everyday running shoes for beginners training for their first 5K” tells the system far more than “running shoes.” Hints shape relevance inside the audience your campaign targeting allows; they never replace geographic, platform, or audience targeting.

Here is one intent cluster built the right way, for a CRM selling to small sales teams:

• Ad group: Consider_Acme_CompareCRMs_Clicks_202610
• Hint: “CRM for a sales team of 10 to 25 reps moving off spreadsheets”
• Hint: “Salesforce alternative with simpler pricing for small teams”
• Hint: “CRM with built-in email sequences and Gmail integration”
• Hint: “Comparing HubSpot and Pipedrive for a growing startup”
• Hint: “Pipeline tracking for founders who still run sales themselves”

Start with a focused set and expand it from conversion data: look at the landing page paths, sales-call notes, and CRM records of your best customers, and turn the situations they describe into new hints. Three mechanics to respect:

• Updates replace the list. Sending new hints through the API or bulk upload replaces the ad group's entire list, so always submit the complete intended set.
• Use exclusion hints. Ad groups support exclusion_hints alongside context hints in the Bulk API, so you can steer away from adjacent topics you never want.
• Negative Phrases for brand policy. Qualifying advertisers can use Negative Phrases for narrow placement requirements specific to their brand's policies.

Worked example: a broad hint such as “business professionals interested in software” earns 100,000 impressions at a $30 CPM for $3,000. At a 0.2% CTR and a 1% conversion rate, that is 200 clicks and 2 conversions, a $1,500 cost per conversion. The same $3,000 on a tight cluster earns 75,000 impressions at a $40 CPM, and at a 0.8% CTR and a 4% conversion rate it delivers 600 clicks and 24 conversions at $125 each. Scale by adding narrow clusters, not by broadening the hints you have. Our guide on how to write ChatGPT Ads context hints has more templates.

Should You Use a Daily or Campaign-Total Budget?

Use a daily budget for always-on campaigns and anything on Maximize results, and a campaign-total budget for fixed-length flights where total spend must stay under a hard number. Choose carefully: you can switch a campaign-total budget to daily, but you can never switch it back.

The pacing mechanics, from OpenAI's daily budgets article and campaign help page:

• The budget week runs Sunday to Saturday in the account's time zone. If a campaign spends its weekly allowance early, it can stop delivering until the next budget week, even while marked active.
• Mid-week changes prorate. Change a daily budget partway through the week and the seven-day limit is prorated from the latest change through the next Sunday at midnight. Raise it during a day and that day's cap becomes 2x the highest budget active that day.
• Minimums depend on currency. The minimum daily budget is $25 USD, EUR 15, GBP 15, INR 725, CAD 25, AUD 25, or JPY 2,500, among 30 billing currencies. With fixed bidding, a daily budget must also be at least the highest effective bid in the campaign, including audience multipliers.

Account-level billing matters once you pass $30,000 a month. ChatGPT Ads is postpay. Self-serve card-billed accounts are charged each time unpaid spend reaches an assigned payment threshold, which can rise automatically with successful payment history; it is a charge trigger, not a spending cap. An account-wide daily spending limit exists only for eligible postpaid-invoice accounts, and campaign budgets still apply independently (OpenAI billing article). Paused ads can keep appearing for up to 24 hours, and that spend is billable.

Scaling rule: make budget changes once per budget week, at the start of the week, and judge each change over a full Sunday to Saturday cycle. Frequent mid-week edits reset the proration window and make performance impossible to read. For sizing the budget itself, see how much budget ChatGPT Ads need.

How Do You Target Countries, Platforms, and Customer Lists?

Targeting lives on the campaign: countries and finer locations, platforms, and custom audiences. Context hints on the ad group shape relevance inside that eligible pool and never override it.

Countries and locations

Self-serve Ads Manager is open to businesses in 89 countries as of October 21, 2026, and the legal entity that advertises and is billed must be based in a listed country (OpenAI availability page). Three rules shape multi-market structure:

• Include, do not just exclude. To target only the United States, explicitly include it; excluding other countries does not restrict delivery.
• New accounts can start at home. Some new self-serve accounts can advertise only in their home country at first. They become eligible for other supported countries once identity verification is approved and they reach a required spend level from home-country campaigns.
• Go below country level outside feed campaigns. Standard campaigns can target states or regions, cities, markets, and postal codes where available, with up to 2,500 location IDs per campaign. New product-feed campaigns target at country level only.

Platforms

In Ads Manager you can select Android app, Android web, Desktop web, iOS app, and iOS web, and segment Insights by the same five. The API groups them into ios_app, android_app, and web (desktop and mobile web together), and campaigns default to all platforms (OpenAI targeting documentation).

Custom audiences

• Inclusion and bid multipliers need 25,000 matched users. Exclusion audiences have no minimum, and exclusions take precedence when someone is in both lists.
• Identifiers: email, phone, their supported SHA-256 variants, and GAID.
• Bid multipliers run from 0.1x to 10x on fixed-bid ad groups; when a user matches several audiences, the highest multiplier applies (OpenAI custom audiences documentation).
• EEA and Switzerland: custom audiences are not supported for campaigns targeting the European Economic Area or Switzerland, where personalized ads are not yet available.

Structure rule: build one campaign per market group (country, currency, and language), because conversation patterns and costs differ by market. Run prospecting campaigns with existing customers excluded, and put retention or upsell campaigns that include a customer list in their own campaign. For the full options reference, see ChatGPT Ads targeting options.

Can You Clone, Bulk-Create, or Automate ChatGPT Campaigns?

Clone Campaign creates a copy only with a different objective or supported billing model, so you cannot clone a Clicks campaign into another Clicks campaign. To duplicate a same-objective structure, use Export for edit and bulk upload, or the Bulk API if OpenAI has enabled it for your account.

Cloning

Cloning is the right tool for one scaling move: graduating a proven Clicks campaign into a Conversions campaign once your Pixel or Conversions API has events flowing. For a same-objective copy, OpenAI says to create a new campaign and enter its settings, ad groups, and ads (OpenAI campaign help page).

Bulk upload and Export for edit

• Template: in Ads Manager, select Create, then Bulk upload, and fill the campaigns, ad groups, and ads tabs.
• Edits: select objects, choose Export for edit, change the file, and re-upload. Rows with existing IDs update; rows with a parent ID and a blank object ID create new objects (OpenAI Edit Campaigns).
• Objectives: bulk upload supports Views, Clicks, and Conversions. Conversion rows need conversion bidding enabled and exactly one conversion event setting ID, and product-feed campaigns cannot use Conversions in bulk upload.
• Limits: 5,000 campaigns, 5,000 ad groups, and 5,000 ads, counted across the whole account.

Bulk API

The Bulk API is in limited preview and enabled per ad account. Each job accepts up to 1,000 create or update operations, accounts are limited to 10 create requests per 10 seconds, and a validate_only mode checks fields and dependencies before anything changes.

Product feed campaigns

Retailers import products by CSV or TXT upload, a hosted URL, or SFTP. Uploaded items expire after two weeks, so OpenAI recommends a hosted URL or automated SFTP, and the Delta Feeds API updates titles, prices, and availability without re-uploading the catalog. Ad groups pick products with filters and custom labels in ads_metadata, and Maximize results is the default bid strategy for eligible new feed ad groups (OpenAI product feed campaigns). Shopify merchants can sync catalog and the OpenAI Pixel through the ChatGPT Ads app. Our product catalog to ChatGPT ads guide covers feed strategy, and our high-volume production system covers batch QA.

How Do You Scale a ChatGPT Ads Account From $5K to $100K a Month?

Scale in three phases: validate intent clusters at $5,000 to $10,000 a month, optimize toward conversions at $10,000 to $30,000, and expand horizontally across clusters, products, markets, and objectives above $30,000. Horizontal expansion beats vertical budget increases because each new cluster opens new eligible conversations instead of bidding harder for the same ones.

These phases are the Lapis operating plan for ChatGPT accounts, built on the platform mechanics above.

Phase 1: Foundation ($5,000 to $10,000 a month)

The goal is to prove ChatGPT works for your business and find the intent clusters that convert. Put 60% of spend on consideration, 30% on decision, and 10% on awareness. Generate at least 20 to 30 distinct ads across the account, each changing one element: angle, framing, image, or call to action. OpenAI's own ad guidance asks for a high-volume, diverse set of ads with distinct angles so the system finds more opportunities.

Worked example: a $10,000 month runs three campaigns with three ad groups each and four ads per group, 36 ads in total: $1,000 on Views, $6,000 on Clicks, and $3,000 on Conversions. At a $4 CPC, inside OpenAI's recommended starting range, the Clicks campaign buys about 1,500 clicks, roughly 500 per ad group and 125 per ad in the month. That is enough to separate winning angles within three to four weeks.

1. Week 1: launch every ad group at a $4 max CPC and leave bids and budgets alone for the full budget week. Remember the reporting lag: impressions, clicks, and CTR refresh about every 15 minutes, while spend can trail by 7 to 8 hours (OpenAI launch guide).
2. Week 2: within each ad group, pause the ads in the bottom quarter by CTR and replace them with new angles. Shift a slice of budget toward the two strongest ad groups at the start of the next budget week.
3. Week 3: judge on conversions. Pause any ad that has spent twice your target CPA without a conversion. Add one new intent cluster per campaign based on what the best traffic was asking about.
4. Week 4: compile cost per lead or acquisition by cluster. Move to Phase 2 when two or three clusters convert within 2x of target.

Phase 2: Optimization ($10,000 to $30,000 a month)

Clone your proven decision-stage Clicks campaigns into the Conversions objective and switch them to Maximize conversions on a daily budget. Split the best consideration campaign into high-intent comparison and early research. Concentrate 60% to 70% of spend in the top two or three ad groups, and pause marginal groups instead of starving them. Replace winning ads before they fatigue: when an ad's CTR trends down against its own peak for two consecutive budget weeks, rotate fresh angles in. Our creative fatigue guide covers the refresh triggers.

Phase 3: Scale ($30,000 to $100,000+ a month)

Expand to 8 to 12 campaigns and add product or vertical segmentation on top of funnel stage: three products means three sets of consideration and decision campaigns. Spend $3,000 a month on each of four related clusters rather than $12,000 on one, because each cluster keeps its own auction dynamics and its own audience. Open new markets from the 63 available countries with one campaign per market group, and create a separate account for each new billing currency or legal entity. At this level, creative production becomes the binding constraint: 80 to 120 active ads with regular refreshes means a constant pipeline of new, distinct, on-spec creative. Our creative volume guide sizes that pipeline.

What Scaling Mistakes Kill ChatGPT Ads Efficiency?

The six mistakes below waste the most money above $10,000 a month, and each one comes from ignoring how the account mechanics work.

1. Scaling one ad group vertically. Doubling budget on a winning cluster bids harder for the same conversations. Add adjacent clusters instead.
2. Editing budgets mid-week. Every daily budget change prorates the seven-day limit from the latest change. Change budgets once per budget week and read results over full weeks.
3. Broadening hints to buy volume. Broad hints inflate impressions and wreck conversion rates, as the worked example above shows. Add narrow clusters.
4. Copying Google structure. Keyword-themed ad groups, match-type splits, and SKAGs fragment budget on a platform with no keyword auction.
5. Mixing markets in one campaign. Countries differ in conversation patterns and costs. Keep one campaign per market group and one account per billing currency.
6. Skipping incrementality. Platform-attributed conversions overlap with search and social. Prove lift before the budget doubles (next section).

How Do You Measure Incrementality at Scale?

Measure incrementality with geo experiments or holdouts, not by reading platform-attributed conversions. OpenAI now partners with Haus, Measured, and WorkMagic on geo-based experiments for ChatGPT Ads, and supports Pixel and Conversions API signals, an Event Quality Score, configurable click-through windows, and a one-day view-through window for reporting.

OpenAI's measurement update reports that WorkMagic's geo-lift study estimated 2.3x more incremental orders for Dose than last-click attribution captured, with 67% of incremental purchases coming from net-new customers, and that in an early analysis 52.7% of eligible one-day view-through conversions occurred within an hour of the ad impression. OpenAI notes that clicking an ad is not always a user's immediate next step, so last-click reporting alone undercounts ChatGPT's impact.

Run incrementality the clean way: hold out a set of regions (or pause ChatGPT in matched geos) for two to four weeks, keep every other channel steady, and compare total confirmed conversions between test and control regions. Then rebalance across channels on confirmed outcomes. When ChatGPT's cost per confirmed conversion beats Google's for the same customer segment over 30 days or more, shift budget step by step and watch Google conversions for cannibalization.

That rebalancing only works if every channel is judged on the same conversions. Wayfinder, the Lapis measurement layer, connects ChatGPT, Meta, Google, LinkedIn, and Reddit results to confirmed sales and leads from your CRM, store, or billing system, counts each conversion once, and compares every channel on the same attribution rules. Attribution tells you where credit belongs; the geo test tells you what is truly incremental. Wayfinder plus a holdout gives you both. See Wayfinder and our ChatGPT Ads ROI measurement guide.

Why Is Lapis ChatSense the Best Way to Run a Scaled ChatGPT Ads Account?

Lapis ChatSense is the best way to run a scaled ChatGPT Ads account because it delivers the structure, creative volume, landing pages, and measurement this guide describes as one managed program. OpenAI gives every advertiser a fast first ad for one channel: Ads Manager drafts an ad from your website, suggests copy and imagery, and offers a plugin for natural-language campaign work inside ChatGPT (OpenAI, September 16, 2026). Lapis runs the whole program.

• ChatSense is Managed ChatGPT + LLM Ads: Lapis agents and a dedicated strategist build intent-cluster campaign structure, write 1,000+ ChatGPT ads tuned to how people ask ChatGPT with a distinct visual per angle, design buyer-intent experiments, deploy autonomously, generate the next run from what won, and report daily. Coverage extends to third-party LLMs, including open-source chatbots.
• RapidDomain builds a prompt-matched landing page for every ad and audience and optimizes it on real conversions.
• Wayfinder counts each conversion once across ChatGPT and every other channel.
• OmniSense carries winning messages to Meta, Google, LinkedIn, and Reddit, so ChatGPT learning compounds across your whole paid program.

For teams weighing an agency against in-house management, ChatSense is the managed option with a dedicated strategist and custom pricing. Lapis runs creative, launch, matched landing pages, ChatGPT ads, and confirmed-conversion measurement in one system, and no other platform does all five.

Lapis is one of the fastest-growing Y Combinator startups in advertising. It is a Y Combinator Fall 2025 company used by 1,500+ marketing teams and 30+ enterprises, rated 4.9 out of 5 on G2 with eight Summer 2026 badges, including High Performer, Best Results, and Users Most Likely to Recommend. Samsung, Domino's, Dr Pepper, Mattel, Tata, Tupperware, Amway, and Bata trust Lapis, alongside high-growth startups Unusual Whales, Duradry, Lightpage, and Respan. The team comes from MIT, Columbia, Duke, and Google.

Book a demo to see Lapis run your ChatGPT Ads account at scale.

Sources and Methodology

Facts and pricing checked as of October 23, 2026. Platform mechanics come from the OpenAI Help Center and OpenAI developer documentation listed below. The phase plan, budget splits, naming convention, and pause rules are the Lapis operating plan for ChatGPT accounts. Worked examples use round numbers to show the arithmetic.

• OpenAI Help Center, Ads Manager Account Setup
• OpenAI Help Center, Managing identity and access for Ads Manager
• OpenAI Help Center, Create Campaigns for ChatGPT Ads
• OpenAI Help Center, Create Ad Groups for ChatGPT Ads
• OpenAI Help Center, Create Ads for ChatGPT Ads
• OpenAI Help Center, Daily Budgets
• OpenAI Help Center, Launch Campaigns
• OpenAI Help Center, Ads in ChatGPT: The Basics
• OpenAI Help Center, Conversion-optimized Campaigns
• OpenAI Help Center, Billing and Payment
• OpenAI Help Center, Ads Manager Availability
• OpenAI Help Center, Edit Campaigns
• OpenAI Help Center, Bulk Upload Campaign Schema Checklist
• OpenAI Help Center, Create Campaigns from Product Feeds
• OpenAI Developers, Targeting
• OpenAI Developers, Custom Audiences
• OpenAI Developers, Bidding and Budgets
• OpenAI Developers, Bulk API
• OpenAI, Building advertising for the way people use AI (October 5, 2026)
• OpenAI Ads, More ways to measure ChatGPT Ads (October 5, 2026)
• OpenAI, Reimagining advertising with AI (September 16, 2026)
• Y Combinator, Lapis company profile
• G2, Lapis reviews