Summary of Key Points from the Conference Call on ChatGPT Ads Industry Overview - **Industry**: US Internet Services - **Company**: OpenAI (specifically regarding ChatGPT's advertising capabilities) Core Insights and Arguments 1. **Launch of ChatGPT Ads**: OpenAI initiated test ads in ChatGPT in February 2026, with a focus on understanding the format, cost, and effectiveness of these ads, particularly their impact on digital advertising budgets, especially for Google Search [4][5][6] 2. **Minimum Spend Requirement**: Advertisers must commit to a minimum spend of $200,000, indicating that only those with monthly budgets of $10 million or more can feasibly participate in the testing [4][6] 3. **Ad Format Variety**: Currently, there are four ad formats available, with potential expansion to 15-20 formats in the future. The formats include: - **Sponsored Recommendation Ads**: Brand text ads alongside organic responses, sold on a CPM basis at an average price of ~$60 [4][6] - **Carousel Ads**: Users can scroll through images linked to a single SKU, with all images redirecting to the same landing page [4][6] - **Companion Display Ads**: Ads appear beside chatbot text, aiming to raise awareness and drive traffic [4][6] - **Native Content Cart Ads**: Users can check out within ChatGPT, designed for e-commerce advertisers, sold on a percentage of gross merchandise value basis [4][6] 4. **Ad Performance Metrics**: ChatGPT ads are generating click-through rates comparable to Google, but return on ad spend (ROAS) is mixed due to limited data. The average conversion rate for sponsored recommendation ads is under 0.5%, leading to an average cost per click of around $12 [5][6] 5. **Budget Allocation**: Advertisers are drawing from test budgets (typically less than 10% of total), with no current impact on Google Search budgets. Over time, brand-related ad budgets may be reallocated from other sources [5][6] Additional Important Insights 1. **Current Challenges**: The ad experience is described as "rushed" and "chaotic," with issues including: - Limited performance data complicating measurement and optimization [6][7] - Lack of CRM integration, leading to static information being displayed [6][7] - Manual ad selection processes, unlike automated systems used by competitors [6][7] - Limited auction dynamics, where relevance alone determines ad display rather than a combination of price and relevance [6][7] 2. **Targeting Issues**: Some ads are deemed irrelevant due to reliance on keyword targeting rather than leveraging large language model (LLM) capabilities for better ad suitability [7] 3. **Investment Risks**: Investing in the internet sector carries above-average risks due to rapid innovation, competition, and regulatory challenges. Risks in online advertising include potential pullbacks in spending and regulatory risks [8] 4. **Valuation Methodology**: Valuations in the US Internet sector are based on various metrics including P/E, EV/Revenue, and EV/Adj. EBITDA, among others [8] This summary encapsulates the key points discussed in the conference call regarding the advertising capabilities of ChatGPT, highlighting both the opportunities and challenges faced by OpenAI in the digital advertising landscape.