ChatGPT Ads Show for a Quarter of Commercial Search Queries

ChatGPT now shows ads in 25–30% of commercial search queries — roughly a quarter — putting it on par with Google AI Mode, according to an SE Ranking study of 50,000 prompts published in August 2026. But 14% of those ads don’t match the conversation that triggered them, and early traffic skews low-quality.

The short version

ChatGPT serves sponsored results in 25–30% of commercial search queries, a rate that now matches Google AI Mode, according to an SE Ranking analysis of 50,000 prompts across 20 categories. The same study found that 14% of served ads are a poor content match for the conversation that triggered them — one in seven. Advertisers are getting clicks, but the audience quality is weak, and media buyers argue the format is only ready for mid-funnel campaigns, not upper-funnel reach or bottom-funnel conversion.

Key facts

• ChatGPT ads appear in 25–30% of commercial queries — on par with Google AI Mode (SE Ranking, August 2026)
• 14% of served ads don’t match the conversation topic, roughly one in seven
• ChatGPT ads use contextual hints rather than traditional keyword targeting
• Early traffic exists but skews low-quality; media buyers recommend mid-funnel use only

What happened

On August 12, 2026, SEO platform SE Ranking published an analysis of 50,000 ChatGPT prompts across 20 commercial categories, finding that OpenAI’s chatbot now shows paid ads for roughly a quarter of commercial search queries — between 25% and 30%. TechCentral.ie summarized the findings on August 14, reporting the rate is about as often as Google AI Mode does TechCentral.ie.

The study is the first broad third-party quantification of how aggressively ChatGPT is monetizing commercial queries since OpenAI began letting advertisers buy placement beneath chat responses in early 2026. The ad format is still young — OpenAI runs it through a ChatGPT Ads Manager beta — but a 25–30% appearance rate means ads have moved from a limited pilot to a mainstream presence in commercial conversations.

How does ChatGPT actually decide which queries get ads?

ChatGPT ads do not use traditional keyword targeting. Advertisers hand OpenAI contextual hints: natural-language descriptions of the conversations where they want their ads to appear, plus a few keyword-like phrases. SE Ranking stated it directly: “ChatGPT ads do not use traditional keyword targeting. Instead, advertisers provide contextual hints: natural-language descriptions of the conversations in which they want to appear, along with a few keyword-like phrases.”

That contextual model is the likely reason 14% of ads are irrelevant. In one out of seven cases, the ad served does not match the subject of the chat — a mismatch rate SE Ranking called “a lot,” and a poor basis for rolling out the advertising proposition broadly at this point. MediaPost covered the same finding under the headline “ChatGPT Serves Ads In Results On Irrelevant Topics, Data Shows” MediaPost.

A separate Adthena analysis of 29 million queries surfaced a related pattern: 83% of ChatGPT ad triggers do not exist in traditional search, meaning advertisers cannot simply port their Google keyword lists into ChatGPT PYMNTS. WIRED’s independent test of 500 ChatGPT questions catalogued which advertisers appeared most often and reinforced the relevance concern WIRED.

Is ChatGPT ad traffic any good?

The study’s most consequential finding for performance marketers is that ChatGPT ads do generate clicks — but not necessarily the right ones. SE Ranking concluded the traffic quality is weak, describing the system as an unreliable supplier of potential customers and noting that lower pricing would not change that.

That distinction matters because a click from a curious browser is not the same as a click from a buyer. If ChatGPT ads surface alongside broad, exploratory questions rather than high-intent purchase queries, the audience skews toward researchers rather than purchasers. It is the same quality problem that has defined zero-click and AI-search traffic more broadly, and it echoes the mixed early CTR results advertisers reported through spring and summer 2026.

Who is affected, and when?

The 25–30% ad rate affects three groups right now, in live August 2026 behavior rather than a roadmap:

• Advertisers testing ChatGPT Ads must accept that today’s inventory is mid-funnel. Omnicom Media’s Sander Kwint told TechCentral the format is above all suited to mid-funnel campaigns, and that upper-funnel reach and brand building are not yet appropriate uses.
• Google and its advertisers face a structural shift: ChatGPT now monetizes commercial queries at a rate matching Google’s own AI Mode, adding a second large AI surface competing for the same demand.
• Searchers are seeing sponsored results in roughly a quarter of their commercial conversations — including logged-out users, per AdExchanger AdExchanger. OpenAI has also broadened eligible categories, with Search Engine Journal reporting the company now allows some health and finance ads.

The practical upshot is that campaigns must be structured around conversations and intent rather than keywords or audiences, with heavy emphasis on relevant content and landing pages. This is the same playbook that applies to the newer Business Agent conversation campaign type, which targets agentic commerce scenarios.

What this means (our take)

The “quarter of commercial queries” headline is eye-catching, but the relevance and quality findings are the real story for operators. A 14% irrelevance rate would be unacceptable in Google Ads, where advertisers bid on explicit keywords and expect near-exact intent matching. ChatGPT’s contextual model trades that precision for conversational reach — and the early data shows the trade is not fully working yet.

For now, treat ChatGPT Ads the way you would a new display or mid-funnel channel, not a replacement for paid search. Structure campaigns around conversation topics and intent rather than keyword lists, and measure against assisted-conversion and mid-funnel metrics instead of last-click ROAS. Keep Google Ads as your bottom-funnel conversion engine and let ChatGPT experiments target the consideration stage. OpenAI is actively tuning relevance through ad dismissals and relevance improvements, so the mismatch problem may narrow over the next few quarters — but it is a live risk today.

What to do now

1. Audit relevance before scaling. If you are in the ChatGPT Ads beta, review which conversation topics triggered your ads each week and flag mismatches immediately.
2. Write contextual descriptions, not keyword lists. OpenAI’s targeting wants natural-language descriptions of the conversations where your product belongs, plus a few keyword-like phrases.
3. Measure mid-funnel metrics. Track engagement, assisted conversions, and branded-search lift rather than last-click ROAS, which will understate a mid-funnel channel.
4. Pair with bottom-funnel paid search. Keep Google Ads for high-intent capture and use ChatGPT for consideration-stage touchpoints.

FAQ

Sources

• TechCentral.ie — ChatGPT shows adverts for a quarter of commercial search queries (August 14, 2026)
• SE Ranking — ChatGPT Ads research, 50,000 prompts across 20 categories (August 12, 2026)
• MediaPost — ChatGPT Serves Ads In Results On Irrelevant Topics, Data Shows (August 12, 2026)
• WIRED — I Asked ChatGPT 500 Questions. Here Are the Ads I Saw Most Often
• PYMNTS — 83% of ChatGPT Ad Triggers Don’t Exist in Traditional Search
• AdExchanger — ChatGPT Ads Have Begun Showing Up For Logged-Out Users