ChatGPT Ads CTR: What Early Numbers Mean for Self-Employed

The click-through rate for ChatGPT Ads more than doubled between March and September 2026—from low starting values to an average of 0.68 percent. That sounds like progress, but it says nothing about qualified customers or sustainable costs. For self-employed people in the DACH region, what matters isn't the click, but the qualified inquiry.

Here's the thing: A higher CTR can mean more people are curious. But it can just as easily mean more people are clicking who ultimately aren't a fit. Offline, you're doing fine. Online, you don't want to collect clicks—you want inquiries that lead to contracts. For that, you need a test that separates interest from real demand.

What the First ChatGPT Ads Data Shows

OpenAI launched the public advertising test on February 9, 2026 in the US. Ads initially appeared for logged-in adult users on the free Free and Go plans. Source

The ads are separate from ChatGPT responses. They're not supposed to influence which answer the model generates. Additionally, no conversation content is shared with advertisers. Source

On May 5, 2026, OpenAI began rolling out a beta self-serve ads manager. Companies could book and manage campaigns directly. At the same time, OpenAI introduced CPC bidding. This allowed advertisers to optimize not just for impressions (CPM), but also for click goals. Source

The previous entry barrier of $50,000 for campaign tests was removed. This made the platform accessible to smaller advertisers in principle. Source

In August 2026, OpenAI announced the expansion of ChatGPT Ads to Europe. Along with CPM and CPC, conversion optimization and additional measurement capabilities were introduced. Source

On September 23, 2026, OpenAI announced an expansion to Southeast Asia and Taiwan. There too, ads would only appear on the Free and Go plans. Source

An analysis of Similarweb data, reported by The Growth Syndicate, cites an average CTR of 0.68 percent for ChatGPT Ads. The upper quartile of advertisers achieved about 1 percent CTR. Particularly strong brands reached approximately 1.57 percent. The observed peak value is listed as 5.4 percent. Source

These values are not official OpenAI benchmarks. They shouldn't be used as expected values for an individual DACH test. For comparison: For search ads on Google and Microsoft, the same analysis cites an average CTR value of 6.66 percent, based on a LocaliQ benchmark for more than 16,000 campaigns. Source

The comparison is only valid to a limited extent. Search ads often respond to explicit purchase or solution intent. ChatGPT Ads appear in a conversational research context. A lower CTR may therefore be related to different user intent and says little about lead quality on its own.

What a Realistic Test Range Is for Small DACH Providers

• below 0.5 percent CTR: reason to review context, offer, and ad copy
• about 0.5 to 1 percent: plausible early value, but no statement about profitability yet
• above 1 percent: interesting signal, provided it generates qualified conversations or inquiries
• well above 1.5 percent: examine more closely whether the result is due to a very narrow audience, strong brand offer, or small data set

These ranges are working assumptions, not official industry standards. The published sources document the aggregated value of 0.68 percent and the mentioned quartile and peak values, but not a universal CTR for German self-employed people.

Why a Higher Click Rate Isn't Enough

The central metric for small providers shouldn't be CTR, but cost per qualified inquiry.

A simple measurement chain is: impressions → clicks → sessions or chat starts → contacts → qualified inquiries → initial conversations → proposals → won customers → revenue or contribution margin.

CTR only measures: clicks divided by impressions times 100. More important for business value is: advertising costs divided by qualified inquiries.

Example as a model calculation, not as a platform benchmark:

• 10,000 impressions
• 0.68 percent CTR
• about 68 clicks
• 10 percent of clicks initiate contact
• 30 percent of contacts meet qualification criteria
• Result: about 2 qualified inquiries

With a budget of €300, in this example the cost would be approximately €150 per qualified inquiry. Whether that's sustainable depends on order value, close rate, and contribution margin.

An offer with 0.5 percent CTR and high inquiry quality can be more economical than an offer with 1.5 percent CTR that attracts many curious but unsuitable visitors.

OpenAI has therefore extended the platform beyond click measurement to include pixel, Conversions API, and conversion optimization. Source

Which Offers Are Suitable for Testing

Suitable offers have a clearly recognizable problem, specific target audience, brief explanation, clear next action, sufficiently high customer value, and measurable inquiry or booking.

Good initial test candidates for self-employed people can be:

• Initial consultation for specialized advisory
• Local service with high margin
• B2B audit or analysis
• Workshop with clear target audience
• Paid diagnosis
• Inquiry for a specific project
• Appointment for a qualified preliminary discussion

Less suitable are general image campaigns, very inexpensive products with low order value, offers without clear benefit, services with long and difficult-to-measure decision phases, highly explanatory services without prepared conversation entry, and products that can't be meaningfully continued in a chat context.

For small budgets, the test should ideally include one offer, one target audience, and one central action. Multiple services in parallel make evaluation uncertain with small data volumes.

How to Define a Qualified Lead

A qualified lead should be defined before starting, for example:

• appropriate geographic area
• realistic need
• minimum budget
• desired timeframe
• appropriate company size
• specific decision-making role

Prepare Landing Page and Conversation Entry

The landing page should answer the same question that led to the ad. A general homepage text often creates additional friction.

A suitable structure for a small test:

• Headline with problem and target audience
• One specific outcome
• Brief explanation of the process
• Evidence through case study, reference, or qualification
• Price range or minimum scope, if appropriate
• One primary CTA
• Brief qualification questions

The conversation entry shouldn't just say "Learn more." More precise examples:

• "Check if your project qualifies for funding."
• "Book a 20-minute initial review for your business."
• "Get an assessment of whether an audit makes sense in your case."

What's Different About Click-to-Chat Formats

With click-to-chat formats, the next step can be a conversation with an AI agent instead of a website visit. Digiday reported on corresponding tests with selected customers in September 2026. Source

A chat entry should ask at most a few business-relevant questions:

• What's this about?
• Where is the company located?
• How large is the project?
• When is support needed?
• What would be a realistic budget range?

AI pre-qualification doesn't automatically replace human review. For evaluation, you must define when a chat counts as qualified.

Measure Clicks, Conversations, and Qualified Inquiries

At minimum, the following events should be tracked: impression, click, landing page visit, form start, form completion, chat start, qualified chat, booked appointment, actually held conversation, proposal, close.

OpenAI cites pixel-based measurement and the Conversions API as central platform extensions. Source

For small providers, a lean setup is initially sufficient:

• Own UTM parameters
• Separate landing page or clear campaign path
• CRM source "ChatGPT Ads"
• "Qualified" event
• Monthly attribution from inquiry to revenue

A Possible UTM Structure

utm_source=chatgpt
utm_medium=paid
utm_campaign=[offer_audience]
utm_content=[creative_a]

The Most Important Formulas

• Click-through rate: clicks ÷ impressions
• Cost per click: costs ÷ clicks
• Lead rate: leads ÷ clicks
• Qualification rate: qualified inquiries ÷ leads
• Cost per qualified inquiry: costs ÷ qualified inquiries
• Appointment rate: held conversations ÷ qualified inquiries
• Close rate: customers ÷ qualified inquiries
• Cost per customer: costs ÷ won customers

For European campaigns, it's particularly important which data is actually available. An August 2026 report describes aggregated data such as impressions and clicks for advertisers, with no access to individual conversation content. Source

Data protection and consent must be aligned with the tracking deployed, target region, and respective legal basis. The sources document available measurement products but don't replace legal review for Germany, Austria, or Switzerland.

Costs and Budgets for Small Businesses in the DACH Region

An official, uniform DACH CPC or minimum price is not published in available sources. Therefore, no blanket "average ChatGPT Ads price" should be claimed.

Historically, there was a minimum requirement of $50,000 for early tests. This was removed with the self-serve rollout in May 2026. Source

For smaller providers, budget shouldn't be derived from a supposed platform minimum size, but from the maximum sustainable price per qualified inquiry.

Calculation: maximum price per qualified inquiry = contribution margin per new customer × expected close rate.

Example as model calculation:

• Contribution margin per customer: €2,000
• Close rate of qualified inquiries: 20%
• Theoretical maximum acquisition value: €400 per qualified inquiry
• Sensible test threshold: below that, to account for sales, dropouts, and uncertainty

A Conservative Budget Range for a Small DACH Test

• €300 to €500 media budget: initial signal test for offer and message
• €750 to €1,500: more sensible if multiple conversion stages are to be measured
• €2,000 or more: only if customer value is sufficiently high and follow-up capacity exists

These budget levels are planning examples, not published ChatGPT Ads prices. Actual CPC, CPM, or available scope may vary by country, audience, format, auction, and platform access.

In addition to media budget, small businesses should plan for costs for landing page, tracking, CRM, appointment qualification, creative variants, and inquiry follow-up.

A Conservative Test Plan for Small Budgets

Here's how it works:

Week 1: Foundations

• Select one offer
• Define target customer and exclusion criteria
• Create landing page or chat entry
• Define conversion events
• Set up CRM source

Week 2: Launch

• Test only one core message
• Prepare two to three creative variants
• Distribute budget evenly
• No daily changes without sufficient data basis

Week 3: Quality Check

• Review search or conversation context
• Document irrelevant inquiries
• Analyze landing page dropoffs
• Refine qualification questions
• Look at cost per qualified inquiry, not just CTR

Week 4: Decision

• Determine winner by qualified inquiries and cost per inquiry
• Include conversion rate and close quality
• Only increase budget if operational processing works
• With poor quality, first check offer and conversation entry, not just ad copy automatically

A campaign shouldn't be judged as success or failure after a few individual clicks. With small budgets, results are statistically unstable. Close rates in particular often require more time than click data.

A sensible reason to stop would not be "CTR below industry value," but for example: no qualified inquiries despite sufficient clicks, cost per qualified inquiry above economic threshold, high number of unsuitable contacts, or lack of capacity for timely processing.

What Results You Evaluate After Four Weeks

After four weeks, you check:

• Offer fit: Do users understand the offer and recognize a specific reason to make contact?
• CTR: Is the click-through rate significantly below or above the available reference value of 0.68 percent? Source
• Click quality: Do users arrive on the page with appropriate expectations or are they looking for something else?
• Landing page conversion: How many clicks lead to form start, chat, or appointment?
• Qualification rate: How many contacts meet the predefined criteria?
• Cost per qualified inquiry: Is the value compatible with contribution margin and close rate?
• Sales follow-up: How quickly are inquiries answered and how many conversations actually take place?
• Close quality: Do proposals and orders result, or just information requests?
• Scalability: Does lead quality remain stable at higher budget?
• Data quality: Are tracking, UTM parameters, and CRM attribution complete enough to make a decision?

What I'd Do Differently: Common Mistakes

I know this: You see a number that looks good and think it's working. But a high CTR alone doesn't help if the inquiries don't fit. Here are the most common mistakes I see in small tests:

• Treating CTR as the main goal and not measuring qualified inquiries
• Directly transferring a Google Ads benchmark to ChatGPT Ads, even though search intent and conversation context differ
• Working with a general homepage instead of a problem-focused landing page
• Testing an inexpensive or very general offer, even though inquiry processing costs are high
• Having no definition of what "qualified" means
• Being unable to attribute chat starts, forms, and phone contacts to the same campaign
• Only measuring the first click, even though OpenAI now supports pixel, Conversions API, and conversion optimization Source
• Changing too many variants simultaneously: offer, audience, CTA, and landing page
• Making a final decision after individual clicks or few leads
• Increasing budget before manually checking inquiry quality

Best Practices: How to Test Properly

The trick is: You don't need many campaigns. You need one clear campaign with clean measurement. This works:

• Test a clear offer instead of a general brand message
• Adapt the CTA to the desired action: initial review, appointment, inquiry, or chat
• Create a written definition for qualified inquiries before starting
• Test two to three message variants, but initially keep offer and audience constant
• Set up conversion tracking early; not after the first test month
• View ChatGPT Ads as a conversation and intent channel, not just banner advertising
• Use the same language and problem definition after the click as in the ad
• Make price range, minimum project size, or exclusion criteria transparent if they reduce unsuitable leads
• Classify leads by quality: A (immediately suitable), B (potentially suitable), C (unsuitable or pure information request)
• Base decisions on cost per qualified inquiry, appointment rate, and close value
• When scaling, first ensure processing capacity

More on ChatGPT Ads and advertising in AI assistants in the topic guide.

Frequently Asked Questions

What is a good CTR for ChatGPT Ads?

An average CTR according to available data is 0.68 percent. The upper quartile reaches about 1 percent, strong brands approximately 1.57 percent. For a small DACH test, 0.5 to 1 percent is a plausible starting value. However, what matters isn't the CTR, but the quality of inquiries.

How much budget do I need for a ChatGPT Ads test?

An initial signal test is possible from €300 to €500 media budget. For meaningful measurement across multiple conversion stages, €750 to €1,500 makes more sense. Higher budgets are only worthwhile if customer value is sufficiently high and you can process inquiries promptly.

How long should a ChatGPT Ads test run?

Plan for at least four weeks. In the first two weeks, you launch the campaign and collect initial data. In week three, you check inquiry quality. In week four, you evaluate cost per qualified inquiry, appointment rate, and close quality. With small budgets, results are statistically unstable; close rates often require more time than click data.

Which offers are suitable for ChatGPT Ads?

Suitable offers have a clearly recognizable problem, specific target audience, brief explanation, clear next action, and sufficiently high customer value. Examples: initial consultation for specialized advisory, local service with high margin, B2B audit, workshop with clear target audience, paid diagnosis, or inquiry for a specific project.

How do I measure qualified inquiries with ChatGPT Ads?

Track at least these events: impression, click, landing page visit, form start, form completion, chat start, qualified chat, booked appointment, actually held conversation, proposal, close. Use UTM parameters, a separate landing page, and a dedicated CRM source for ChatGPT Ads. Define before starting when a lead counts as qualified.

Are ChatGPT Ads better than Google Ads?

That can't be said universally. Google search ads often respond to explicit purchase or solution intent and achieve an average 6.66 percent CTR. ChatGPT Ads appear in a conversational research context and achieve an average 0.68 percent CTR. A lower CTR doesn't automatically mean worse quality. What matters is which platform brings you qualified inquiries at sustainable costs.

The Step You've Been Postponing for Years

You know you have to start with advertising at some point. You know referrals alone aren't enough. You know other providers are growing online while you watch. But you don't know where to start. You don't want to burn money again. You don't have time to learn every platform.

I get it. But look: ChatGPT Ads aren't more complicated than Meta Ads or Google Ads. The platform is new, the principles are the same. A clear offer, a suitable audience, clean measurement. That's enough to start.

You don't have to know everything. You just have to begin. Choose an offer. Define who fits. Create a landing page that answers the same question as the ad. Set up tracking. Start with €300 to €500. After four weeks, check if inquiries fit and costs are sustainable. If yes, you scale. If no, you know what needs fixing.

The first test isn't a masterpiece. It's a signal. It shows you whether the offer works, whether the message lands, whether the platform fits you. And it shows you what to change next. That's more than you have today.

If you don't want to do it alone or want to be faster: Nelija builds the campaign for you, sets up tracking, and reviews inquiries with you. No contract, no hidden costs. Try Nelija for free.