OpenAI Moves ChatGPT Ads to CPC Amid Revenue Growth

OpenAI has fundamentally changed how it sells advertising within ChatGPT, abandoning its initial impressions-based model for a cost-per-click pricing structure. This strategic pivot, occurring just ten weeks after ads were first introduced, places the company in direct competition with giants like Google and Meta for performance-driven marketing budgets. The shift was prompted by a rapid decline in the value of ad impressions, forcing a move to a model that ties revenue directly to user engagement. This change underscores the intense financial pressures on OpenAI as it seeks to monetize its flagship product at scale.

Initially launching with a cost-per-thousand impressions model in February, OpenAI charged advertisers a premium $60 CPM. That rate quickly eroded, falling to as low as $25, making the volume-dependent approach unsustainable. The new system allows advertisers to set bids between $3 and $5 per click, while the minimum spend requirement has been significantly reduced from $250,000 to $50,000. This adjustment aims to attract a broader range of brands by offering a more familiar and accountable pricing metric.

These sponsored placements appear at the bottom of ChatGPT responses for users on the free and $8-per-month Go tiers, clearly labeled and visually separated from the AI’s answer. For product-related queries, the ads can resemble the shopping cards seen on Google. OpenAI emphasizes user privacy, stating that advertisers cannot access personal conversations, chat history, or identifiable user data. They receive only aggregated performance metrics. Targeting is purely contextual, based on the topic of the ongoing chat rather than demographic profiling.

The company’s move into advertising has been remarkably swift. After hiring a longtime Google ads executive in mid-2024, OpenAI launched its initial test with major brands like Target and Adobe. The pilot quickly scaled, generating an annualized run rate exceeding $100 million within two months. Internal revenue projections are ambitious, targeting $2.5 billion from ads in 2026, scaling to $100 billion by 2030. To support this growth, OpenAI has rapidly built out its advertising infrastructure, partnering with ad-tech firms for programmatic buying, developing conversion tracking tools, and launching a global self-serve platform for advertisers.

This aggressive expansion marks a stark reversal for CEO Sam Altman, who had previously been publicly critical of advertising models, calling them a “last resort.” He now argues that ad-supported access is essential to democratizing AI for billions of users who cannot afford subscriptions. This philosophy directly contrasts with rivals like Anthropic and Perplexity, which have positioned their Claude and Perplexity AI assistants as explicitly ad-free, premium alternatives. Anthropic even ran Super Bowl commercials criticizing the very concept of ads in AI, a campaign that correlated with a notable jump in its user base.

The industry is now split on a core question: do users perceive AI chatbots more like utilitarian search engines, where ads are an accepted nuisance, or as private conversational partners, where commercial intrusions break trust? Early surveys suggest significant user skepticism, with a majority reporting that ads in AI search reduce their trust in the results. A consumer boycott movement has also gained traction, highlighting the potential reputational risk for platforms that monetize through advertising.

Underlying this strategic shift is acute financial pressure. Despite generating massive revenue growth, OpenAI is projected to lose approximately $14 billion this year due to enormous computing and research costs. The company does not anticipate profitability until 2030. Advertising revenue represents the fastest available path to bridge this gap without further subscription hikes or massive funding rounds. The overall market for AI search advertising is forecast to explode from $1 billion to nearly $26 billion in just a few years, and OpenAI aims to capture a major share from the outset.

The critical test will be whether the move to cost-per-click can attract advertisers who found the initial CPM model both expensive and difficult to measure. Early reports indicated that the advertising test was struggling to meet expectations due to a lack of mature measurement tools. The new CPC model provides a simple, understood metric: the click. However, the long-term success of OpenAI’s advertising platform will depend on whether those clicks convert into real value for businesses, and whether ChatGPT becomes an indispensable channel for marketers navigating the new landscape of AI-driven search.