---
title: "ChatGPT Ads Glossary: 20 Terms to Check Before Your AI Spends"
description: "ChatGPT Ads borrows familiar advertising terms but changes their meaning. Learn 20 terms and the questions to ask before approving campaign changes through groas."
url: "https://groas.com/post/chatgpt-ads-glossary-18-terms-to-know-before-your-ai"
image: "https://pub-87da24ecbbfc4c3bad6875f3aa013712.r2.dev/generated-images/b2ae0138-abda-4900-9338-d0b5576c133d.png"
published: "2026-11-20T15:34:08.663Z"
modified: "2026-11-20T15:34:08.735Z"
---

November 20, 2026 · 11 min read

# ChatGPT Ads Glossary: 20 Terms to Check Before Your AI Spends

[](https://groas.com/author/alexander-perelman)

[Alexander Perelman](https://groas.com/author/alexander-perelman)Personally Tested

Head Of Product @ groas

[LinkedIn](https://www.linkedin.com/in/alexander-433793253/)

![A weekly pill organiser filled with coins, Tuesday's compartment overflowing with double the money, hinting that a daily ad budget is really a weekly average.](https://pub-87da24ecbbfc4c3bad6875f3aa013712.r2.dev/generated-images/b2ae0138-abda-4900-9338-d0b5576c133d.png)

In this article

1. [ChatGPT Ads placement](#chatgpt-ads-placement)
2. [Campaign](#campaign)
3. [Ad group](#ad-group)
4. [Ad](#ad)
5. [Context hints](#context-hints)
6. [Viewer eligibility](#viewer-eligibility)
7. [Advertiser eligibility](#advertiser-eligibility)
8. [Custom audience](#custom-audience)
9. [Daily budget](#daily-budget)
10. [Total budget and pacing](#total-budget-and-pacing)
11. [Intelligent bidding (Maximize Results)](#intelligent-bidding-maximize-results)
12. [Target CPA](#target-cpa)
13. [Target CPC](#target-cpc)
14. [Enhancements](#enhancements)
15. [Product-feed campaign](#product-feed-campaign)
16. [Visual ad](#visual-ad)
17. [Sponsored Agent and Business Agent destination](#sponsored-agent-and-business-agent-destination)
18. [OpenAI Pixel and Conversions API](#openai-pixel-and-conversions-api)
19. [Attribution window and reporting freshness](#attribution-window-and-reporting-freshness)
20. [Keywords](#keywords)

Your AI assistant says it adjusted your ChatGPT Ads targeting and budget. Before you approve, check what those words mean: the targeting is not a keyword list, and the daily budget is not a daily cap.

That is the trap. ChatGPT Ads borrows the vocabulary of search and social advertising without bringing all the old mechanics along. Familiar words can make an unfamiliar decision sound routine.

I spent years mining search terms at 1am. That habit is useful background, not a setup template. I want an assistant to explain what it changed, why it changed it, and which numbers support the decision. Fluent reporting is not the same as sound account management.

I ordered these terms the way you meet them: placement, structure, audience, money, creative, measurement. Read through once, then look up the term your assistant uses before approving a change through groas. **Same words, different mechanics. Verify before you approve.**

![Ink cartoon of an AI assistant mislabeling ad terms while a PPC manager checks its work](https://pub-87da24ecbbfc4c3bad6875f3aa013712.r2.dev/generated-images/9bd21e06-2e8f-433a-b823-4fef5b88d9e2.png)

## ChatGPT Ads placement

**Placement is the sponsored unit below the ChatGPT answer, visually separated and labeled.**

It shows the advertiser name, favicon, title, description, image and landing-page link. Crucially, [it does not change the answer itself](https://promptwatch.com/glossary/chatgpt-ads).

Google advertisers hear _placement_ and think Display network or YouTube slot. Here, the important distinction is between the answer and the ad beneath it. I would start creative review there, not with a spreadsheet of headlines stripped of their surroundings.

If an assistant reports weak copy, I want to see the actual ad before approving another rewrite.

> **Ask your AI:** Show me the placement preview, its title and description, and the landing page it links to.

## Campaign

**A campaign is the top container that sets the objective, budget, dates and country.**

The hierarchy is [campaign → ad group → ad](https://help.openai.com/en/articles/20001209-launch-campaigns). Familiar enough. The dates deserve less casual treatment.

ChatGPT Ads paces a total budget over a default 60-day window when the end date is blank. An empty field is therefore not an absence of instructions. It gives the system a spending horizon you may not have intended.

For the Google version, I keep this [campaign structure playbook](https://groas.com/post/google-ads-best-practices-2026-campaign-structure-bidding-creative-measurement) separate. Similar containers do not guarantee identical rules.

> **Ask your AI:** List each campaign’s objective, budget, dates and country. Which campaigns have no end date?

## Ad group

**An ad group organizes one theme or intent inside a campaign and holds the context hints and bid for that theme.**

In the [campaign hierarchy](https://help.openai.com/en/articles/20001209-launch-campaigns), the campaign sets the money and market; the ad group describes the conversation you want to appear in.

The useful organizing idea is one buying moment, not one recycled keyword cluster. Mix three intents into a single group and the blended reporting becomes harder to act on. A cheap average does not tell me which intent deserves more money.

I would split distinct intents before trying to explain their combined performance.

> **Ask your AI:** List each ad group’s theme, hints and bid. Flag groups mixing different buying intents.

## Ad

**An ad is the title, copy, image and landing page the user can click.**

The [ad is the visible unit](https://help.openai.com/en/articles/20001209-launch-campaigns); campaigns and ad groups organize it.

Do not carry over the assumption that this works like a Google responsive ad mixing a collection of headlines. Review the ad as a unit, including its destination. A promising title followed by a page answering a different question is not a copy problem. It is a paid detour.

> **Ask your AI:** Show me each ad beside its landing page and the intent its ad group is designed to reach.

## Context hints

**Context hints are short, plain-language descriptions of the conversations where your product fits.**

They sit at ad-group level and guide matching, but [they are not exact-match keywords and do not guarantee delivery](https://www.chatadsai.com/guides/chatgpt-ads-glossary).

This is where old search habits become expensive. A keyword describes a query. A hint describes a conversational situation. HawkSEM illustrates the translation from [`best CRM for sales teams` to `marketing leaders comparing CRM for 50-person team`](https://hawksem.com/blog/chatgpt-ads-context-hints/).

Because the system matches conversational meaning, vague hints can reach broad discussions while very tight hints can restrict delivery. Narrower is not automatically more efficient if almost nothing runs.

I want to read the hints themselves, not an assistant’s reassuring summary of the targeting.

> **Ask your AI:** Read back every hint verbatim. Which are too broad, which may restrict delivery, and what reporting supports that assessment?

## Viewer eligibility

**Viewer eligibility determines which ChatGPT users can see ads.**

[Free and Go users are eligible](https://www.chatgptadlibrary.com/openai-ads), including logged-out traffic where allowed. Plus, Pro, Business, Enterprise, Edu and under-18 users are excluded.

Google advertisers may hear _eligibility_ and think policy approval. Here, it is an inventory question: the audience filter operates before an ad can compete.

That does not mean every enterprise buyer is unreachable. It means I would not mistake total ChatGPT usage for available advertising reach, or assume a strong hint can overcome a plan exclusion.

> **Ask your AI:** Which viewer restrictions affect this campaign, and what do we actually know about the resulting reach gap?

## Advertiser eligibility

**Advertiser eligibility determines whether your business is allowed to buy ads.**

It depends on the [billing country](https://www.chatgptadlibrary.com/openai-ads). Viewer eligibility filters the audience; advertiser eligibility checks the buyer.

This is the unromantic setup detail I would settle before discussing creative. What the deck calls _global rollout_, I call checking the billing address first. An eligible audience is not much use if the paying entity cannot complete setup.

> **Ask your AI:** Is our billing country eligible, and which business entity will appear on the invoice?

## Custom audience

**A custom audience is a matched list used to include people in your targeting.**

The cited inclusion floor is [25,000 matched users, with 100,000 recommended](https://www.chatadsai.com/guides/chatgpt-ads-glossary).

A Meta buyer may expect a small email list to start retargeting quickly. Here, list size can stop the exercise before creative matters. Matching reduces the uploaded file, and viewer eligibility further limits who can see ads.

I would check the matched count before writing a performance diagnosis. A list that cannot serve does not need better headlines.

> **Ask your AI:** How many users matched, are we above the inclusion floor, and what eligibility limitations remain?

## Daily budget

**Daily budget is a weekly average, not a hard daily spending cap.**

Spend can reach [2× the daily budget on one day and 7× over the Sunday-to-Saturday week](https://www.chatadsai.com/guides/chatgpt-ads-glossary).

That distinction matters more than the number in the budget field. The system shifts spending with conversation volume, then balances across the week. A $50 daily budget can therefore produce a $100 Tuesday without being off pace.

I would not call that overspending merely because Tuesday looks ugly. I would check the weekly allowance and actual daily spend together. Nor would I tell a client that $50 is the most the account can spend tomorrow.

> **Ask your AI:** Show actual spend by day this week, the weekly allowance, and our pace against it.

![Miniature tabletop model of a weekly ad budget as seven tiny day jars with coins spilling over](https://pub-87da24ecbbfc4c3bad6875f3aa013712.r2.dev/generated-images/07ecec51-e3c8-43b5-bd32-79a8b9a94fd9.png)

## Total budget and pacing

**Total budget is the amount allocated to the campaign’s flight; pacing determines how it is spread across time.**

Without an end date, the system uses a [default 60-day pacing window](https://www.chatadsai.com/guides/chatgpt-ads-glossary).

This is a different question from daily-budget averaging. One determines the overall spending horizon; the other explains variation within the week.

If the intention is a two-week test, I would set a two-week end date. Otherwise, a small test can trickle along for two months while everyone waits for enough activity to make a decision.

> **Ask your AI:** What total budget and dates are set, and what daily pace do they imply?

## Intelligent bidding (Maximize Results)

**Intelligent bidding is the default bidding mode that seeks the most results within your budget and settings.**

Previously called _Maximize Results_, it works with [Target CPA for conversions and Target CPC for clicks](https://admux.ai/chatgpt-ads/pricing).

Google experience helps with the vocabulary, but I would not treat this as imported tROAS logic. The system bids for conversation opportunities against the selected target. That makes the result you choose and the target you supply central to the decision.

For the Google mechanics, this [bidding strategies guide](https://groas.com/post/google-ads-bidding-strategies-2026-complete-guide-manual-cpc-smart-bidding-tcpa-troas) is the separate reference. Same industry, different instructions.

> **Ask your AI:** Which bidding mode is active, what result is it pursuing, and what target have we given it?

## Target CPA

**Target CPA is the average cost per conversion you ask Intelligent bidding to achieve.**

It is [an average target, not a ceiling](https://admux.ai/chatgpt-ads/pricing). A conversion costing more than the target does not, by itself, mean the system ignored it.

The conversion signal matters before the target does. Only [standard events such as `order_created` and `lead_created` support conversion optimization](https://www.chatadsai.com/guides/chatgpt-ads-glossary). A custom event will not become an optimization signal because the target looks sensible.

An aggressively low target can also restrict delivery. I would check signal quality and actual CPA before approving an efficiency target that mostly expresses optimism.

> **Ask your AI:** Which standard event are we optimizing for, how many fired in the last 14 days, and what CPA did we actually achieve?

## Target CPC

**Target CPC is the average click price you request when optimizing for clicks rather than conversions.**

It belongs to the [Intelligent bidding setup](https://admux.ai/chatgpt-ads/pricing), not the old mental model of a fixed manual bid.

The system can move around that average as it competes for conversations. I would therefore read CPC beside delivery and outcomes, not celebrate it alone. Cheap clicks are only useful if they reach the right buying moments.

> **Ask your AI:** What average CPC did each ad group pay, and what did those clicks produce?

## Enhancements

**Enhancements include opt-in text customization that adapts a headline to the conversation.**

[AI-suggested creative](https://promptwatch.com/glossary/chatgpt-ads) also drafts variants, but a suggested draft and a live contextual rewrite are different approval problems.

With text customization, the approved headline becomes a starting point rather than necessarily the final wording. That matters most when the copy contains pricing, product claims or compliance language.

My rule would be simple: approve the boundaries, not just the draft. If the wording must remain exact, do not treat rewriting as a harmless finishing touch.

> **Ask your AI:** Where is text customization enabled, and what can we verify about the wording that actually ran?

## Product-feed campaign

**A product-feed campaign builds ads from a catalog using a product set and an ad template.**

The format [pairs feed data with templates](https://promptwatch.com/glossary/chatgpt-ads). The template supplies details such as price, image and landing page while the conversation shapes the framing.

That puts catalog maintenance ahead of clever copy. A stale feed price can become a misleading live ad. No amount of headline testing repairs the wrong number at the source.

I would fix the catalog before asking an assistant for another creative variation.

> **Ask your AI:** Which product set and template supplied the ads, which products received clicks, and are their prices and destinations current?

## Visual ad

**A visual ad displays four to six images in the sponsored unit.**

The format is [in beta](https://promptwatch.com/glossary/chatgpt-ads). A Meta buyer may hear _carousel_ and start planning a swipe-by-swipe story. Here, the images appear together beneath an answer.

I would favor images that make sense at a glance rather than a sequence that needs patient navigation. Show the product doing the job; do not make the reader assemble the sales pitch.

> **Ask your AI:** Show the complete image set as it appears in the unit, and separate reported performance from assumptions about individual images.

## Sponsored Agent and Business Agent destination

**A Sponsored Agent lets users continue chatting inside ChatGPT; a Business Agent destination sends them to your agent.**

These [agent-based formats](https://promptwatch.com/glossary/chatgpt-ads) change what happens after the click. Instead of arriving at a conventional landing page, the user continues into a conversation.

That conversation has to do useful work. It can qualify interest and answer product questions, or it can ramble until the buyer leaves. A polished opening message is not the same as a working sales destination.

I would review the agent’s answers as seriously as landing-page copy.

> **Ask your AI:** What can we measure after an agent click, and where do users stop progressing toward a useful product answer?

## OpenAI Pixel and Conversions API

**The OpenAI Pixel and Conversions API connect browser and server events to activity after a ChatGPT ad click.**

The setup uses the [`oaiq` browser SDK and server events](https://www.chatadsai.com/guides/chatgpt-ads-glossary), with a shared Pixel ID, event name and Event ID for deduplication. The click identifier, `oppref`, lives in the first-party cookie `__oppref` for 30 days.

This is not a _drop one tag and forget it_ job. Browser reporting and conversion optimization are not interchangeable: the browser tag alone reports, while optimization requires the supported server-event setup and standard events such as `order_created` or `lead_created`.

Missing matching information can leave real outcomes disconnected from the advertising signal. Before changing bids in response to zero conversions, I would inspect the plumbing.

> **Ask your AI:** Show browser and server events by name for the last 14 days, including missing match keys and deduplication issues.

## Attribution window and reporting freshness

**The attribution window determines which later outcomes receive ad credit; reporting freshness determines when those outcomes appear.**

Reporting is aggregate-only and last-touch, with 7-, 14- or 30-day click windows and 0- or 1-day view windows. The columns also arrive at different speeds:

- **Impressions and clicks:** about 15 minutes.
- **Spend:** 7–8 hours.
- **Conversions:** 24–48 hours.

Those [reporting delays](https://topgrowthmarketing.com/how-to-measure-chatgpt-ads/) explain why live impressions can appear beside zero spend. User conversations are not shared, so I would not ask for a private chat transcript to explain an aggregate result.

The practical mistake is comparing fresh clicks with incomplete costs and conversions. Same-day CPA can look precise while its ingredients are still arriving.

> **Ask your AI:** State the attribution window and data cutoff. Recalculate performance using spend older than eight hours and conversions older than 48 hours.

## Keywords

**Keywords are not the targeting unit in ChatGPT Ads; context hints are.**

This is the term I would retire. Calling hints _keywords_ encourages the wrong work: pasting a match-type sheet into a system that reads conversational meaning, then reporting on coverage as though the old structure still applies.

Say you are spending $20k a month. Every hour your assistant spends defending keyword coverage is an hour it does not spend reviewing a hint, fixing a feed price or checking an event match through groas.

I want the assistant to use the right noun because the noun shapes the check. Ask it plainly: which hint brought delivery, what did it cost, and what did it return? Everything else in this glossary follows from that.

## Frequently Asked Questions

### Does a ChatGPT Ads daily budget cap what can be spent in a day?

No. Daily budget is a weekly average, and spend can reach 2× the daily budget on one day and 7× over the Sunday-to-Saturday week. A $50 daily budget can therefore produce a $100 Tuesday without being off pace, so check the weekly allowance and actual daily spend together rather than calling a single high day overspending.

### Are keywords the targeting unit in ChatGPT Ads?

No, keywords are not the targeting unit; context hints are. Context hints are short, plain-language descriptions of the conversations where your product fits, and they are not exact-match keywords and do not guarantee delivery. Pasting a keyword sheet into the system encourages reporting on coverage that no longer applies.

### Do ChatGPT Ads change the answer ChatGPT gives?

No. The placement is the sponsored unit below the ChatGPT answer, visually separated and labeled, showing the advertiser name, favicon, title, description, image and landing-page link. It does not change the answer itself, which is the important distinction between the answer and the ad beneath it.

### What happens if a ChatGPT Ads campaign has no end date?

ChatGPT Ads paces the total budget over a default 60-day window when the end date is blank, so an empty field still gives the system a spending horizon. If the intention is a two-week test, set a two-week end date; otherwise a small test can trickle along for two months.

### How large does a custom audience list need to be for ChatGPT Ads?

The cited inclusion floor is 25,000 matched users, with 100,000 recommended. Matching reduces the uploaded file, and viewer eligibility further limits who can see ads, so check the matched count before writing a performance diagnosis, because a list that cannot serve does not need better headlines.

### Why does a ChatGPT Ads campaign show impressions but zero spend or conversions?

Reporting columns arrive at different speeds: impressions and clicks in about 15 minutes, spend in 7–8 hours, and conversions in 24–48 hours. Reporting is aggregate-only and last-touch, so live impressions can appear beside zero spend. Avoid comparing fresh clicks with incomplete costs and conversions, since same-day CPA can look precise while its ingredients are still arriving.

### Can I optimize ChatGPT Ads bidding toward a custom conversion event?

No. Only standard events such as order_created and lead_created support conversion optimization, and a custom event will not become an optimization signal because the target looks sensible. Check the conversion signal before the target, and check signal quality and actual CPA before approving an aggressively low target that can restrict delivery.

### Will ChatGPT Ads rewrite my approved headline?

It can, if text customization is enabled. The approved headline becomes a starting point rather than necessarily the final wording, which matters most when the copy contains pricing, product claims or compliance language. If the wording must remain exact, do not treat rewriting as a harmless finishing touch; approve the boundaries, not just the draft.

## Run Your Marketing On groas

Plug ChatGPT, Claude or any AI agent into groas and let it do the work of a whole marketing team: $19 a month flat with every feature included, and autopilot whenever you’re ready.

[Start For $19](https://groas.com/pricing)

[Or have groas run it for you](https://groas.com/done-for-you)

## Structured data

```json
{"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://graph.groas.com/entity/groas","name":"groas","url":"https://groas.com/","description":"groas is the marketing infrastructure for your own AI: it connects your AI to all your marketing and gives it proprietary marketing data collected 24/7. Plug Claude, ChatGPT or any AI agent into groas and run all your marketing from it, for a tiny fraction of the usual cost, with the groas app there whenever you want to see more. It is the layer ChatGPT, Claude and any AI assistant run a business’s online marketing on. Through groas an assistant can do anything you could do natively in Google Ads, Bing Ads, Meta Ads and ChatGPT ads, on your website and in AI search, and, where the owner switches it on, can run connected campaigns by itself, so AI agents do the work a whole marketing team used to. groas also gives those agents large-scale proprietary marketing data it collects around the clock, on any schedule the owner chooses: every ad competitors run on Google and Meta, their landing pages and full funnels with the offers and prices in them, Reddit customer research, AI search answers from ChatGPT, Google AI Overviews, Gemini, Claude and Perplexity, and SEO data from Semrush and Ahrefs, backlinks included ($19 a month, flat, with everything included, for unlimited teammates on one shared workspace and unlimited ad accounts and websites, delivered as an MCP server). Separately, its done-for-you service has senior people and proprietary quantitative systems run paid ads, SEO and AI search for businesses, and white label for agencies, from $5,000 a month.","logo":{"@type":"ImageObject","url":"https://groas.com/icon-512.png","width":512,"height":512},"sameAs":["https://www.linkedin.com/company/groas/"]},{"@type":"WebSite","@id":"https://groas.com/#website","name":"groas","url":"https://groas.com/","inLanguage":"en-US","publisher":{"@id":"https://graph.groas.com/entity/groas"}},{"@type":"WebPage","@id":"https://groas.com/post/chatgpt-ads-glossary-18-terms-to-know-before-your-ai#webpage","url":"https://groas.com/post/chatgpt-ads-glossary-18-terms-to-know-before-your-ai","name":"ChatGPT Ads Glossary: 20 Terms to Check Before Your AI Spends","description":"ChatGPT Ads borrows familiar advertising terms but changes their meaning. Learn 20 terms and the questions to ask before approving campaign changes through groas.","inLanguage":"en-US","isPartOf":{"@id":"https://groas.com/#website"},"about":{"@id":"https://graph.groas.com/entity/groas"},"primaryImageOfPage":{"@type":"ImageObject","url":"https://pub-87da24ecbbfc4c3bad6875f3aa013712.r2.dev/generated-images/b2ae0138-abda-4900-9338-d0b5576c133d.png"},"breadcrumb":{"@id":"https://groas.com/post/chatgpt-ads-glossary-18-terms-to-know-before-your-ai#breadcrumb"},"dateModified":"2026-11-20T15:34:08.735Z","mainEntity":{"@id":"https://groas.com/post/chatgpt-ads-glossary-18-terms-to-know-before-your-ai#article"}},{"@type":"BreadcrumbList","@id":"https://groas.com/post/chatgpt-ads-glossary-18-terms-to-know-before-your-ai#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://groas.com/"},{"@type":"ListItem","position":2,"name":"Blog","item":"https://groas.com/blog"},{"@type":"ListItem","position":3,"name":"ChatGPT Ads Glossary: 20 Terms to Check Before Your AI Spends","item":"https://groas.com/post/chatgpt-ads-glossary-18-terms-to-know-before-your-ai"}]},{"@type":"BlogPosting","@id":"https://groas.com/post/chatgpt-ads-glossary-18-terms-to-know-before-your-ai#article","url":"https://groas.com/post/chatgpt-ads-glossary-18-terms-to-know-before-your-ai","mainEntityOfPage":{"@id":"https://groas.com/post/chatgpt-ads-glossary-18-terms-to-know-before-your-ai#webpage"},"isPartOf":{"@id":"https://groas.com/blog#blog"},"headline":"ChatGPT Ads Glossary: 20 Terms to Check Before Your AI Spends","description":"ChatGPT Ads borrows familiar advertising terms but changes their meaning. Learn 20 terms and the questions to ask before approving campaign changes through groas.","datePublished":"2026-11-20T15:34:08.663Z","dateModified":"2026-11-20T15:34:08.735Z","image":{"@type":"ImageObject","url":"https://pub-87da24ecbbfc4c3bad6875f3aa013712.r2.dev/generated-images/b2ae0138-abda-4900-9338-d0b5576c133d.png"},"author":{"@id":"https://groas.com/author/alexander-perelman#person"},"publisher":{"@id":"https://graph.groas.com/entity/groas"},"keywords":"before, chatgpt ads, ask, chatgpt, ads, terms, target, budget","wordCount":2422,"inLanguage":"en-US"},{"@type":"Person","@id":"https://groas.com/author/alexander-perelman#person","name":"Alexander Perelman","jobTitle":"Head Of Product @ groas","description":"My background includes Goldman Sachs and Stanford Computer Science.","image":"https://groas.com/media/blog/4f17cac3a81acc224cc1b2cfab8f94bbc295a3089d13e8421125771ef8d4064f.jpg","worksFor":{"@id":"https://graph.groas.com/entity/groas"},"sameAs":["https://www.linkedin.com/in/alexander-433793253/"]},{"@type":"FAQPage","@id":"https://groas.com/post/chatgpt-ads-glossary-18-terms-to-know-before-your-ai#faq","mainEntity":[{"@type":"Question","name":"Does a ChatGPT Ads daily budget cap what can be spent in a day?","acceptedAnswer":{"@type":"Answer","text":"No. Daily budget is a weekly average, and spend can reach 2× the daily budget on one day and 7× over the Sunday-to-Saturday week. A $50 daily budget can therefore produce a $100 Tuesday without being off pace, so check the weekly allowance and actual daily spend together rather than calling a single high day overspending."}},{"@type":"Question","name":"Are keywords the targeting unit in ChatGPT Ads?","acceptedAnswer":{"@type":"Answer","text":"No, keywords are not the targeting unit; context hints are. Context hints are short, plain-language descriptions of the conversations where your product fits, and they are not exact-match keywords and do not guarantee delivery. Pasting a keyword sheet into the system encourages reporting on coverage that no longer applies."}},{"@type":"Question","name":"Do ChatGPT Ads change the answer ChatGPT gives?","acceptedAnswer":{"@type":"Answer","text":"No. The placement is the sponsored unit below the ChatGPT answer, visually separated and labeled, showing the advertiser name, favicon, title, description, image and landing-page link. It does not change the answer itself, which is the important distinction between the answer and the ad beneath it."}},{"@type":"Question","name":"What happens if a ChatGPT Ads campaign has no end date?","acceptedAnswer":{"@type":"Answer","text":"ChatGPT Ads paces the total budget over a default 60-day window when the end date is blank, so an empty field still gives the system a spending horizon. If the intention is a two-week test, set a two-week end date; otherwise a small test can trickle along for two months."}},{"@type":"Question","name":"How large does a custom audience list need to be for ChatGPT Ads?","acceptedAnswer":{"@type":"Answer","text":"The cited inclusion floor is 25,000 matched users, with 100,000 recommended. Matching reduces the uploaded file, and viewer eligibility further limits who can see ads, so check the matched count before writing a performance diagnosis, because a list that cannot serve does not need better headlines."}},{"@type":"Question","name":"Why does a ChatGPT Ads campaign show impressions but zero spend or conversions?","acceptedAnswer":{"@type":"Answer","text":"Reporting columns arrive at different speeds: impressions and clicks in about 15 minutes, spend in 7–8 hours, and conversions in 24–48 hours. Reporting is aggregate-only and last-touch, so live impressions can appear beside zero spend. Avoid comparing fresh clicks with incomplete costs and conversions, since same-day CPA can look precise while its ingredients are still arriving."}},{"@type":"Question","name":"Can I optimize ChatGPT Ads bidding toward a custom conversion event?","acceptedAnswer":{"@type":"Answer","text":"No. Only standard events such as order_created and lead_created support conversion optimization, and a custom event will not become an optimization signal because the target looks sensible. Check the conversion signal before the target, and check signal quality and actual CPA before approving an aggressively low target that can restrict delivery."}},{"@type":"Question","name":"Will ChatGPT Ads rewrite my approved headline?","acceptedAnswer":{"@type":"Answer","text":"It can, if text customization is enabled. The approved headline becomes a starting point rather than necessarily the final wording, which matters most when the copy contains pricing, product claims or compliance language. If the wording must remain exact, do not treat rewriting as a harmless finishing touch; approve the boundaries, not just the draft."}}]}]}
```
