ContentsStart0%
What do these cases say about ChatGPT Ads cost?
Nine public ChatGPT Ads reports show widely different costs and outcomes. Their markets, currencies, campaign periods and attribution methods differ. Each number needs its own context before it can inform a budget.
For cost comparisons, start with the same advertiser, currency and period. The five-market breakdown below keeps those fields visible. These historical observations are not a current rate card.
The original Traditional Chinese analysis was published on 18 October 2026.
This article compiles advertiser disclosures and interviews published by named sources. LS-Labs — AI Search and Brand Visibility Research did not run these campaigns. The nine rows identify public reporters or publishers; some underlying accounts are unnamed. They cannot establish a platform average or a reliable expected return.
Nine public ChatGPT Ads case reports
CPC is cost per click; CPM is cost per thousand impressions; CTR is clicks divided by impressions. A source’s bare $ symbol is retained when it does not explicitly identify the currency. Such amounts are not converted or pooled with USD figures. “Not disclosed” is not zero.
Each case lists its source, period, currency, spend, click and impression metrics, and outcome.
| Reporter / source | Period and market | Currency | Spend | CTR and sample | CPC | CPM | Outcome and attribution |
|---|---|---|---|---|---|---|---|
| Opascope | Unnamed account; 15 days; dates and market not disclosed. Page updated 15 June 2026. | $; currency not specified | About $60,000 | CTR and click count not disclosed; CVR 2.35% | About $1.72 | Not disclosed | About $89,000 revenue; ROAS 1.49× using its own tracking. Covers campaigns selected for scaling, not all tests. Industry not disclosed. |
| Common Thread Collective | High-value ecommerce; post says 34 days and lists 5/22 and 6/22; exact calendar scope and market unclear. | $; currency not specified | $9,620 | CTR 0.94%; 136,000 weekly impressions, not a full-period total | $4.41 | Not disclosed | $19,000–38,000 attributed revenue; estimated ROAS 3.3–6.8× using Triple Whale. Published figures do not reconcile to a common spend denominator. |
| Hostinger | Hosting service; several weeks; exact dates and market not disclosed. | $; currency not specified | Nearly $70,000 | CTR and sample counts not disclosed | Not disclosed | Above $65 | Reports initial and later purchases. Counts, revenue and full attribution details not disclosed. |
| SE Ranking | About two weeks; US, Canada, Australia and New Zealand. Page dated 10 August 2026. | Not disclosed | Not disclosed | CTR 1.30%; 97,000+ impressions; 1,263 clicks; 3 campaigns, 8 ad groups, 48 ads | Not disclosed | Not disclosed | Very few sign-ups; stopped the test. “Very few” does not mean zero. The article’s external CPC quote is a separate advertiser’s evidence. |
| Floyd Blaikie’s team | B2B integrations service; several weeks. Identified-organization data covers the past week; market not disclosed. | Spend: CAD. CPC/CPM: $ without a separate currency label | About CAD 7,000 | CTR 0.7%; weekly subset: 336 paid clicks, 146 identified organizations | $9.29; currency unspecified | $64.34; currency unspecified | 5 of 146 identified organizations fit the ideal customer profile; no conversions reported at that point. Weekly subset and total spend cover different periods. |
| Synter | 2–10 June 2026; US, UK, Canada, Australia and New Zealand; API extraction 11 June. | USD, explicitly stated | US$4,428.84 | Overall CTR 0.35%; 129,663 impressions, 448 clicks | US$9.89 overall | US$34.16 overall | Conversion volume too small for the published comparison; no usable revenue or ROAS disclosed. Market detail below. |
| Choice OMG | Two weeks in June 2026; Ontario and Alberta, Canada; 3 campaigns. | $; currency not specified | $415 | CTR about 0.6%; nearly 9,000 impressions; 60 clicks | About $7 | Not disclosed | MediaPost interview: no interested buyers or conversions. Small test; full attribution settings not disclosed. |
| Cleverly | B2B lead generation; campaign dates and market not disclosed. Report published 28 August 2026. | Not disclosed | Not disclosed | CTR not quantified; platform: 57 clicks; Google Analytics: fewer than 20 visits | Not disclosed | Not disclosed | Reported zero conversions and paused. Reason for the click/visit discrepancy remains unresolved. |
| Case relayed by Camino5 | Second-hand account of an unnamed campaign; dates and market not disclosed. Report published 28 August 2026. | $; currency not specified | About $1,000 | Clicks reported; count not disclosed | Not disclosed | Not disclosed | Heard about clicks with no corresponding Google Analytics data. Not a verified Camino5 campaign or evidence of zero conversions. |
Case evidence rechecked on 25 October 2026. Campaign periods, sample periods and publication dates are shown separately where available. Camino5’s second-hand account has weaker evidential support than a first-hand disclosure.
Synter: five markets, one currency
The Synter account report explicitly states USD. The US campaign launched first; the other four markets were added on 7 June. The reporting window covers 2–10 June 2026, so the markets had different delivery histories.
| Market | Impressions | Clicks | CTR | CPC | CPM | Spend |
|---|---|---|---|---|---|---|
| US | 103,133 | 337 | 0.33% | 10.62 | 34.69 | 3,577.61 |
| UK | 8,989 | 65 | 0.72% | 5.10 | 36.90 | 331.73 |
| Canada | 9,181 | 33 | 0.36% | 8.33 | 29.95 | 274.96 |
| Australia | 5,714 | 10 | 0.17% | 17.59 | 30.78 | 175.88 |
| New Zealand | 2,646 | 3 | 0.11% | 22.89 | 25.95 | 68.66 |
| Total | 129,663 | 448 | 0.35% | 9.89 | 34.16 | 4,428.84 |
New Zealand’s three clicks are a small sample. No revenue is disclosed for a profitability comparison.
How to read the performance claims
Revenue needs a matching attribution denominator
Opascope reports revenue from campaigns it chose to scale. That selection limits its value as a forecast for a new account. Its reported ROAS also excludes the question of profit after product, delivery and other costs.
Common Thread Collective reports $9,620 spend, $19,000–38,000 attributed revenue and an estimated ROAS range of 3.3–6.8×. Dividing that revenue range by the published total spend does not reproduce the reported ROAS. The public account lacks the aligned periods or denominators needed to reconcile it. Preserve the claim as attributed reporting; avoid choosing the highest number as a budget assumption.
Clicks, suitable customers and conversions are different outcomes
Floyd Blaikie’s weekly subset identifies 146 organizations, of which five fit the target customer profile. Those five are not five sales. SE Ranking reports very few sign-ups, which must remain distinct from a claim of zero.
A reporting gap needs investigation. A campaign that cannot be found in analytics does not supply a measurable conversion rate. Platform clicks and site visits may also use different counting rules; a mismatch alone does not establish the cause.
What to measure in a test
The useful next step is a measurement plan tied to the business outcome you need. The following is an editorial checklist drawn from the limits of these reports, not a tested budget recommendation.
- Choose the outcome and loss limit. Define a purchase, qualified lead or activated account, then set a tolerable spend ceiling before delivery begins.
- Keep scope alongside each metric. Record market, currency, dates, campaign identifiers and attribution window. Keep spend and attributed outcomes aligned.
- Retain both platform and site records. Label paid campaigns consistently and investigate discrepancies. The GA4 source guide explains the separate task of observing AI referral traffic; paid campaigns still need their own labels.
- Review customer quality and economics. Inspect qualified leads, completed purchases and margin. CPC and CTR alone cannot answer whether continued spend is worthwhile.
Market eligibility, available controls and pricing can change. This evidence review does not verify present access to the ad platform or provide an account setup tutorial.
Sources and method
The table links to six first-hand disclosures or publisher accounts: Opascope, Common Thread Collective, Hostinger, SE Ranking, Floyd Blaikie and Synter. Three further rows come from MediaPost’s 28 August 2026 interviews and reporting. All source pages were reopened for this adaptation; missing fields remain unknown.
The selection follows the nine cases in the Traditional Chinese source article. It is a curated desk review with publication and selection bias. Campaign totals are not combined. Modeled industry conversion estimates, organic AI referral studies and anonymous platform success stories are outside this table, so their figures cannot be mistaken for additional observed campaigns.
Evidence checked: . Publication dates are not substitutes for campaign dates. The review makes no claim that the disclosed results have been independently audited against the advertisers’ private accounts.
