ChatGPT Ads Are Underdelivering. Here's What to Do.

Should SMBs be buying ChatGPT Ads right now?

Not yet. The early numbers from agencies testing ChatGPT Ads show a 0.6% CTR against a typical 2% benchmark, at around $7 per click, with clicks reported by the platform not matching what shows up in Google Analytics. That is not a channel you scale. It is a channel you monitor while others pay for the learning curve.

What is actually happening with ChatGPT Ads?

OpenAI launched its advertising product quietly, and a handful of agencies moved fast to test it. The results, reported by Stacked Marketer, tell a consistent story: impressions without intent.

Users inside ChatGPT are in a fundamentally different cognitive mode than users on Google Search. They are generating, exploring, and thinking out loud. They are not in a buying frame. When an ad interrupts that flow, clicks happen but purchase intent does not follow. A 0.6% CTR is the symptom. The cause is a mismatch between the surface and the intent signal advertisers depend on.

The tracking gap makes this worse. When platform-reported click counts do not reconcile with Google Analytics, you cannot trust the data you would use to optimize. That is not a rounding error. It is a structural problem that makes spend decisions unreliable.

How does this compare to what OpenAI is actually earning?

Here is the irony: OpenAI's ad revenue is reportedly doing fine. The business model is working for OpenAI. The ROI for advertisers is a different question entirely.

This pattern is not new. Early Facebook Ads, early YouTube pre-roll, and early programmatic display all went through a phase where platform revenue grew while advertiser results lagged. The platform figures out monetization first. Advertiser ROI catches up later, or it does not, and the format dies. ChatGPT Ads are somewhere in that early, uncertain window.

The platform making money and advertisers making money are two different things. Right now only one of those is confirmed.

What does the Claude news mean for SMBs?

Separately, Anthropic has been building out Claude's capabilities in ways that matter for business operators. The "twin" reference in the Stacked Marketer headline points to Claude's expanding model lineup, with distinct versions optimized for different tasks. For SMBs already using Claude in workflows, that is worth tracking because the right model for a writing task is not the right model for a data analysis task, and pricing differs across tiers.

For operators not yet using Claude, the short version is this: the competitive pressure between OpenAI and Anthropic is producing faster capability improvements than anyone predicted 18 months ago. That is good for users. It also means the tool you evaluated six months ago may not reflect what is available today.

What should SMBs actually do with AI advertising channels?

There is a framework worth using here. Advertising channels fall into three categories for a small or mid-size business:

• Core channels: Proven ROI, defensible attribution, scale with confidence.
• Test channels: Emerging, some signal, small budget experiments to build internal knowledge.
• Watch channels: Too early to spend, but worth monitoring quarterly.

Right now, ChatGPT Ads belong in the watch column for most SMBs. The exception would be a business with a large experimental budget and a dedicated paid media team that can tolerate ambiguous attribution and build institutional knowledge before competitors do. Most SMBs do not fit that description.

What to watch for before moving ChatGPT Ads to "test" status

| Signal | What it means | |---|---| | CTR climbs above 1.5% in independent reports | Intent alignment is improving | | Platform and GA4 click data reconcile within 10% | Attribution is trustworthy enough to optimize | | Case studies from businesses in your category with verified ROAS | Proof it works for your type of offer | | OpenAI releases conversion tracking documentation | They are taking advertiser ROI seriously |

None of those boxes are checked today. Watch for them.

What about AI-assisted advertising on existing platforms?

This is where SMBs should actually be spending attention. Google's Performance Max, Meta's Advantage+ campaigns, and LinkedIn's AI-driven audience tools are mature products with real attribution infrastructure and years of optimization data behind them. The AI layer on these platforms is improving faster than most operators realize.

Using AI tools to write ad copy, generate creative variants, and analyze campaign performance on those established channels is where the ROI is right now. Not on a new ad surface where the tracking does not work.

If you are running paid media, the highest-leverage move in 2025 is probably getting better at prompting AI to produce and test ad creative faster, not chasing a new inventory source with broken attribution.

What we'd actually do

• Put ChatGPT Ads on a 90-day review cadence. Set a calendar reminder, read the independent agency reports as they come out, and revisit the decision with fresh data. Do not spend yet.
• Audit your attribution setup on existing channels first. If your Google Analytics and platform data do not reconcile on Meta or Google campaigns, fix that before you add another surface to the confusion. Clean data on proven channels beats early access to broken data anywhere.
• Test AI for creative production, not AI ad inventory. Use Claude, ChatGPT, or your preferred tool to generate 10 ad copy variants in the time it used to take to write two. Run those variants on channels where you trust the attribution. That is the AI advertising play that is actually working right now.

If you want to work through your paid media AI strategy with people who are running this for real clients, that conversation is happening at skool.com/aiforbusiness.