Advertiser data via Similarweb; ad-program facts from OpenAI.
OpenAI’s pivot to advertising — $100M annualized in weeks, B2B SaaS flooding in, and the highest-intent inventory on the internet now for sale.
What Happened
On February 9, 2026, OpenAI quietly switched on advertising inside ChatGPT for Free and Go tier users in the United States. Within weeks, the program had reportedly reached a $100 million annualized revenue run-rate — a number OpenAI has not contradicted and that TechCrunch and Axios both corroborated at launch. Ads now appear in roughly 49% of US replies and about 26.5% of replies globally, according to Similarweb data.
New Similarweb advertiser data reveals who is buying: the top spenders are overwhelmingly B2B SaaS — Monday.com, JotForm, Shopify, Cursor, Resume.io, HubSpot, Aikido Security, Canva, and Shutterstock lead the rankings. The US accounts for approximately 70.9% of all ChatGPT ad impressions. A note on the data: Similarweb’s raw impression counts appear identical across its country and advertiser breakout tables, which is almost certainly a display artifact. The numbers to trust are the rankings, the share percentages, and OpenAI’s own confirmed figures — not the absolute impression counts.
The speed of the platform’s structural evolution is arguably more significant than the revenue figure itself. OpenAI launched with a $60 flat CPM and a $200,000 minimum buy — a premium, direct-sold model designed for large brands. Within roughly ten weeks, it flipped to a self-serve, no-minimum CPC auction any small business can enter with a credit card. That is the AdWords democratization playbook, compressed into a single quarter.
The key insight: OpenAI did not gradually test its way into advertising. It launched at premium CPMs, validated demand in weeks, then immediately democratized access — the precise sequence Google ran between 1999 and 2002, compressed into a single quarter. The $100M run-rate is not the story. The structural shift from model lab to consumer-media company is.
The Structural Read
There are three interlocking dynamics here, and each one matters independently.
1. The subscription ceiling forces ads. ChatGPT has hundreds of millions of free users. You cannot convert them all to $20/month subscriptions — the math breaks before you reach meaningful penetration. Advertising is the only mechanism that monetizes attention at scale without a paywall, which is exactly the logic that built Google and Meta. OpenAI arriving at this conclusion is not a failure of the subscription model; it is the natural second engine of any consumer platform that reaches mass scale. The ~$100M annualized run-rate within weeks tells you the advertiser demand was already there, waiting. This development is central to OpenAI’s IPO story — the company is reframing itself as a deployment company, not just a model lab, and advertising revenue is proof of that consumer-scale thesis.
2. The advertiser mix is the moat. The reason Monday.com, HubSpot, Shopify, Cursor, and Canva are first in line is not coincidence. ChatGPT captures the highest-intent inventory in existence: people literally asking “what tool should I use for X.” A user typing “what’s the best project management software for a 10-person team” into ChatGPT is further down the purchase funnel than almost any Google search query. That is arguably the most valuable ad surface ever created, and B2B SaaS — with its high LTV, measurable CAC, and tool-evaluation purchase journeys — is structurally the perfect fit. The US concentration at 70.9% of impressions reflects both where ChatGPT’s highest-intent commercial users are and where the self-serve auction launched first.
3. The trust tightrope is real. Ads threaten the single property that made ChatGPT the default for millions of users: the perception that answers are neutral. OpenAI maintains that its responses remain independent and conversations private — a claim that is structurally identical to the one Google has made for 20 years, with mixed results. The moment users begin suspecting that a sponsored result shaped an answer, the utility premium evaporates. This is not hypothetical; it is the central tension of attention monetization, and it now sits inside the most-used AI product on earth.
The Google-ification Thesis
Attention monetization has come to AI — and it changes everything downstream
Once a platform monetizes attention through auctions, its incentives shift. Product decisions, ranking logic, and interface design all begin to accommodate the revenue layer — sometimes imperceptibly, sometimes not. Google’s 25-year arc from “don’t be evil” to ad-shaped search results is the reference case. OpenAI is now on the same path, with the same structural forces, and no reason to expect a different outcome. The question is not whether this happens but how fast.
Map of AI — The Attention Layer
“Whoever controls the attention layer controls the distribution tax on every application above it. ChatGPT’s ad launch is not a monetization event — it is a claim on the attention layer itself.”
Top ChatGPT Advertisers by Impression Share (Similarweb Rankings)
Rankings and relative share — not raw counts, which appear artifactual in the source data.
Three Implications
GOOGLE AND META FEEL THIS IMMEDIATELY
Every B2B SaaS dollar going into ChatGPT auctions is a dollar that was previously flowing to Google Search and LinkedIn. The high-intent query — “best CRM for a startup
91,000+ executives read Business Engineer for the AI strategy frameworks cited by ChatGPT, Claude, and Perplexity.
Sources: openai.com · techcrunch.com · axios.com · mlq.ai · cloro.dev









