Skip to content

Channel decisions

ChatGPT Ads Results: Early Numbers

ChatGPT ads results so far: sourced early returns and one-account benchmarks, each with source and date. Not averages.

A results clipboard, magnifying glass, and evidence folder.

Early public numbers for ChatGPT ads are sparse and uneven. OpenAI’s March 2026 pilot update emphasized trust and relevance signals rather than advertiser ROAS. Trade press has described underwhelming early CTRs for some brands, while one agency published a 15-day single-account benchmark. None of these is a universal average.

The table below labels every figure. Where OpenAI does not publish an advertiser CTR or ROAS benchmark, this page says so.

Results table (source + date on every row)

Read each row as that source’s claim for its window — not a forecast for your account.

ClaimSource typeSource + date
OpenAI: no impact on consumer trust metrics; low ad dismissal; improving relevance (pilot signals)Platform (consumer metrics — not advertiser ROAS)OpenAI Testing ads in ChatGPT update — 26 Mar 2026
Some brands reported weak early CTRs and opaque measurement in first weeksTrade press qualitative (no single CTR published in the article fetch)Performance Marketing World — article on early returns
~$60K spend, ~$89K revenue, 1.49x blended ROAS, ~$1.72 CPC, 2.35% CVR over 15 days on scaled campaignsThird-party single-account benchmark — not an OpenAI averageOpascope — updated 15 Jun 2026
Day-to-day ROAS roughly 0.2x–2.9x in that same Opascope windowThird-party single-account volatilityOpascope — updated 15 Jun 2026
OpenAI published average advertiser CTR / CPC / ROASNot published as a single public figure on the checked OpenAI pagesChecked 23 October 2026 — do not invent

How to use these numbers

Use OpenAI’s eligibility and billing docs for setup. Use first-party revenue and your own conversion definitions to judge a test. Treat agency benchmarks as hypotheses. For channel choice, compare business outcomes — see vs Google Ads.

What remains unpublished

As of 23 October 2026, the OpenAI help and ads.openai.com pages cited in this cluster do not publish a single advertiser-average CTR, CPC, or ROAS for ChatGPT ads. If a vendor page shows a number without a dated method, treat it as marketing copy until you can reproduce it in your account.

Common questions

What results are ChatGPT ads seeing?

Public data is limited. OpenAI has shared positive consumer-trust pilot signals. Trade press has described weak early returns for some brands. One agency published a 15-day single-account ROAS/CPC snapshot. OpenAI has not published a single advertiser CTR or ROAS average on the pages we checked.

Can I use Opascope’s 1.49x ROAS as my forecast?

No. It is labelled as one account’s scaled campaigns over 15 days. Use it only as a comparison point while you run your own measured test.

Where should I track my results?

Use Ads Manager plus your own analytics and sales records. Reconcile definitions before comparing channels.

Sources and verification

Every numeric row cites a dated third-party or OpenAI consumer-pilot source. No invented platform-wide CTR/ROAS average. Your account may expose a different set of features. Examples are illustrative unless a dataset is explicitly supplied.

Take the next step

Compare with Google Ads ↗