ChatGPT ads performance: do they work, what they cost and how they compare with Google Ads

ChatGPT ads performance is uneven: a few brands report cheaper acquisition, while independent tests report few or no conversions. Entry costs have fallen to small daily budgets, but Google Ads converted better in both same-advertiser tests cited here.

The AI Ads launch post shows how AirOps tracks ChatGPT ads, the paid side of AI search, beside organic answers. Ads Library collects observed ads. Ad Load % is the share of tracked ChatGPT answers carrying at least one ad.

This guide weighs published results and ChatGPT ads pricing against Google Ads benchmarks. Every figure carries its source, date, and denominator, because the numbers rarely measure the same thing. It ends with a cost per conversion formula and a scale, hold, or stop scorecard for your own test.

Are ChatGPT ads worth it?

ChatGPT ads are worth a measured test when your buyers use ChatGPT Free or Go. They have not yet proven themselves as a channel for scaled spend. Treat early budget as research money, with a clear stop rule set before launch. The sections below show why that caution holds.

OpenAI says as much in its own documentation. ChatGPT Ads has no cross-advertiser performance benchmarks yet, according to OpenAI's ads help center FAQ. The same FAQ calls the platform a beta and warns that delivery and spend fluctuate. Any ChatGPT ads review you read rests on one advertiser's test or a panel estimate.

Practitioner sentiment reflects that split. Matt Hammel, Co-founder & COO at AirOps, says the market is "very divided on how useful it is today." He describes agencies with mature Google and Meta programs holding back, while other marketers find early wins in specific verticals. So ChatGPT ads performance depends on your category and your buyers.

A fair ChatGPT ads review of your own funnel starts with two checks. Your buyers need to use the free tiers, and your conversion tracking needs to work. You also need room for a test that produces zero qualified leads.

How are ChatGPT ads performing?

ChatGPT ads results split by who reports them. The strongest wins come from partner results that OpenAI published, and the weakest come from independent agency tests. Neither group gives you a market average.

OpenAI's October 5 measurement post lists early partner findings that OpenAI chose to publish. WeightWatchers saw an attributed cost per acquisition (CPA) 15.3% below its blended paid-search benchmark, per DV Rockerbox. Dose drew 67% of incremental purchases from net-new customers, per WorkMagic. These are vendor- and partner-reported figures with no spend, sample size, or dates attached. Treat them as the best case so far.

Workshop Digital's B2B test, published August 26, 2026, gives the clearest independent numbers. The agency spent $2,319 of its own budget across three campaigns, at a $3.14 CPC. The post reports a 1.1% CTR. The 1.06% used here is recalculated from its 739 clicks and 69,630 impressions.

The campaigns produced zero MQLs, while the agency's Google Search campaigns produced two. The authors call it an imperfect comparison, since search and chat capture different intent. Still, it reports every input, which few tests do.

Read every row in that table as one advertiser's experience. A partner dashboard shows association between ads and sales, which does not prove the ads caused them. Your ChatGPT ads results will depend on your offer and your tracking.

ChatGPT ad CTR and why the benchmarks disagree

Reported ChatGPT ad CTR runs far below Google's search median, and each source divides clicks by a different base. That difference decides which comparisons hold up. Check the denominator before you compare any two CTR figures.

Search Engine Journal's Graphite report, published September 29, 2026, puts the US rate at 0.73%. Graphite built it from a Similarweb panel, with data running through September 16. The denominator is clicks per ad-bearing exchange, meaning one message plus its reply, pooled across advertisers.

The same report places English-language markets between 0.6% and 0.75%. Similarweb sells the ad-tracking tool and Graphite is a growth agency, so both have a stake.

Workshop Digital's test, covered above, measures CTR per impression from an ad server. Its 1.06% on ChatGPT against 5.2% on Google Search is a same-advertiser comparison. That makes it the fairest CTR comparison available, even though it covers one account.

Google's broad baseline comes from the WordStream and LocaliQ 2026 benchmarks, which put median search CTR at 6.64%. The sample covers 13,474 US search campaigns from April 2025 to March 2026. It combines Google Ads and Microsoft Ads, and the publisher sells ad-management software.

Graphite's per-reply rate and WordStream's per-impression median do not belong in the same ratio. Compare CTRs only within one advertiser and one denominator. Even then, CTR tells you about creative pull, not customer cost.

Alex Hallenbeck, Lead Product Manager, Enterprise at AirOps, puts it plainly: "Cheap clicks, don't mean cheap customers."

He adds that tighter targeting toward a qualified audience often pushes CPC up sharply. Conversion rate tends to rise with it, which improves return on ad spend (ROAS). So read ChatGPT ad CTR as a creative signal, and leave cost decisions to cost per conversion.

How much does it cost to advertise on ChatGPT?

Advertising on ChatGPT now starts with a small daily budget on CPC bidding. Your actual ChatGPT ads cost comes from an auction and has moved almost every month since launch.

The early drop was steep. Digiday reported in April 2026 that cost per thousand impressions (CPM) fell from $60 at launch to as low as $25. The same article says minimum commitments fell from $250,000 to $50,000. Those CPM figures came from agency anecdotes rather than a rate card. The timeline below lays out each change.

ChatGPT ads CPC vs CPM, and which to bid

Bid on CPC unless reach is your goal, because clicks tie spend to visits. OpenAI's ads basics guide recommends a starting maximum bid of $3 to $5 per click. OpenAI frames that as a starting bid rather than a price.

The same guide describes a relevance-weighted, second-price auction. Your bid caps what you pay, and competing ads plus relevance set the final ChatGPT ads CPC. OpenAI publishes no formula for the relevance weighting, so you learn your real price only by running.

The guide also covers oCPC. It bills per valid click and optimizes delivery toward conversions. That option suits you once your conversion tracking fires reliably.

Pick ChatGPT ads CPM bidding only when you want reach in answers and accept paying for views. Pick CPC or oCPC when you pay for traffic or conversions. Revisit that choice after your first test shows which goal matters.

ChatGPT ads minimum spend and budget pacing

ChatGPT ads minimum spend fell from a $250,000 pilot commitment to $50,000 by April, per Digiday. As of August 2026, self-serve campaigns start at $25 a day, per a Search Engine Land guide. StubGroup's CEO wrote it, and his agency runs ChatGPT campaigns for clients.

The floor matters less to your ChatGPT ads cost than how fast the platform spends. OpenAI's FAQ, cited above, says daily spend can reach twice your daily budget. Weekly spend can reach seven times it. Set each daily cap at an amount you can afford doubled.

Are ChatGPT ads better than Google Ads?

ChatGPT ads are not better than Google Ads for direct response on today's evidence. Google converted better in both same-advertiser tests cited here. ChatGPT reaches buyers at a different moment, mid-conversation with a recommendation already on screen.

Later ChatGPT ads CPM figures sit far from April's lows. Search Engine Journal's six-month roundup reports Hostinger CPMs above $65 and a Blaikie CPM of CAD 64.34. The same roundup lists Synter CPCs of $10.62 in the US and $5.10 in the UK. These are self-published tests from a handful of advertisers, so treat them as data points.

Cost comparisons swing on setup. Alex Hallenbeck, Lead Product Manager, Enterprise at AirOps, says "some people have said it's cheaper," while others report higher costs. He ties the variance to targeting. No published test isolates that effect yet, so treat it as his read of customer accounts until someone measures it.

Workshop Digital paid 54% less per click on ChatGPT than on Google Search. Synter's US CPC of $10.62 sat well above WordStream's $5.42 median. So ChatGPT ads vs Google Ads has no single cost answer, only an answer for your own account and market.

Run your own split before you trust anyone else's figure. Start small, and keep both channels' setups identical. AirOps growth solutions tie those paid metrics to the organic visibility data you already track.

The cost per conversion math that settles the comparison

Divide CPC by conversion rate, and compare that number across channels. Cost per conversion is the one ChatGPT ads performance number that travels cleanly between platforms.

WordStream's medians give you a Google baseline: $5.42 divided by 8.18% is about $66 per conversion. That lines up with the same report's $66.69 median cost per lead. The illustration below applies the same math to a cheap ChatGPT click at StubGroup's 0.21% conversion rate.

A $3 click looks cheap until it converts at 0.21%. Run the numbers for your own accounts in four steps.

1. Pull CPC and conversions for the same date range on both channels.
2. Count the same conversion event on both, such as a demo request or purchase.
3. Divide each channel's CPC by its conversion rate.
4. Compare each result with the margin you earn per customer.

Where ChatGPT ads land relative to organic answers

AirOps' fintech ads study found ads rose from near zero in April to about 53% of answers in the study's final week. When those answers named brands, the ad usually belonged to a competitor. That pattern changes how you read ChatGPT ads performance next to organic results.

The study covered 2,826,899 ChatGPT answers about 131 fintech brands over sixteen weeks in 2026, ending with the week of July 13. Of those, 207,785 carried at least one observed ad, a 7.35% average across the window.

The study's page puts it directly: when the organic answer contains brand context, the ad "overwhelmingly intercepts a competitor." Track paid and organic visibility separately, and remember that ad load says nothing about CTR.

That intercept pattern has a mirror image on the buying side. Alex Hallenbeck, Lead Product Manager, Enterprise at AirOps, says brands' ad prompts are "not areas, where they're being organically mentioned." He describes that as the pattern in the majority of cases. Paid placement, read that way, extends reach into answers your organic work has not won.

The separation runs in both directions. OpenAI's January 2026 advertising principles state that ads do not influence the answers ChatGPT gives. Buying an ad buys the slot beside the answer, and organic work still shapes the answer itself.

ChatGPT ads for business and who should test them

Test ChatGPT ads for business when your buyers use ChatGPT Free or Go. OpenAI's principles, cited above, exclude ads from Plus, Pro, Business, and Enterprise subscriptions. A buyer on a paid work plan never sees your ad. Check your analytics for ChatGPT referral traffic before you commit budget. Start with that check before you plan creative.

That limit shapes B2B fit. Product-led growth (PLG) companies fit best, because their buyers often start on free tools. Matt Hammel, Co-founder & COO at AirOps, sees B2B traction "particularly for PLG-style products, where there's, like, a free plan." Those buyers live on the free tiers where ads run.

The flip side matters most for enterprise sellers with sales-led deals. AirOps covers it in an earlier webinar recap: judge ads on impact, not impressions, "especially as premium users may never see them." For those buyers, organic influence is the route into the answer.

So ChatGPT ads for business suit consumer brands and PLG software first. Sales-led enterprise software belongs in the hold column until eligibility changes.

How to judge ChatGPT ads performance with a scale, hold, or stop scorecard

Decide on cost per conversion against your Google baseline, after you reconcile clicks and read incrementality. Score each signal below at the end of your test.

Kevin Indig, author of Growth Memo, makes the case in his recent webinar recap. He argues that measurement must move past last-click attribution toward incrementality and self-reported attribution. You can watch his full session on AI search winners and losers below.

OpenAI agrees the work is unfinished. Its October 5 post, cited earlier, calls incrementality measurement early and plans geo-experiments with Haus, Measured, and WorkMagic. Until independent lift studies exist, run your own holdout before ChatGPT ads move from test to line item.

AirOps for ChatGPT ads performance tracking

The cost and CTR figures in this guide leave out where your ads and competitors' ads run beside organic answers. AirOps AI Ads, live for Enterprise customers with expert support, tracks that placement. Ad Load %, Paid Share of Voice, and Ads Library show where your ads and competitors' ads appear across tracked prompts. Observed ads also cover Google AI Mode answers.

You can connect spend, impressions, clicks, CTR, CPC, and CPM to the same market data as your organic work. That pairing puts your spend, CTR, and CPC next to the share of answers you hold. Playbooks and Brand Kit help you create and update ads. AirOps ad experts can help you choose first opportunities, set up the channel, and support measurement.

Work with us