OpenAI initiated a test for ChatGPT advertisements in February, but this highly anticipated move into commercialization is experiencing a rushed and disorganized start. According to a UBS report dated February 23rd, titled "ChatGPT Advertising Expert Call Summary," an expert managing approximately $800 million in digital advertising budgets stated directly that the current ChatGPT ad experience is "worse than the early days of Google, Meta, or even Snapchat." Although the market hopes that ChatGPT ads can challenge the dominance of Google Search, the current ad budgets are solely derived from experimental spending by advertisers (representing less than 10% of total budgets), having no impact on Google so far. A high minimum commitment of $200,000, a chaotic user experience, and unsatisfactory return on investment indicate that the path to commercialization for ChatGPT ads remains long. High Barrier to Entry: Open Only to Major Players OpenAI has set a minimum spending commitment of $200,000 for ChatGPT ads, effectively limiting participation to large advertisers with monthly budgets exceeding $10 million. The testing phase began in February, with its duration still undetermined. Despite the high barrier, market participation interest is significant. The expert observed adoption across a wide range of verticals—including furniture retailers, automotive, subscription-based consumer goods, luxury brands, B2B, and the food industry. Advertisers can independently choose which ad formats they wish to test. Four Ad Formats: From Brand Exposure to Closed-Loop Transactions ChatGPT currently offers four ad formats, though the expert anticipates this could expand to 15-20 in the future: Sponsored suggestions serve as the primary brand advertisement, inserting branded text into the natural language responses to user queries. Priced on a CPM basis, the current average cost is approximately $60. Carousel ads allow users to scroll through multiple images, each clickable to a specific product. However, the functionality is still immature, with each image linking to only one SKU and all images redirecting to the same landing page on the merchant's website. It lags behind offerings from Google and Meta. Companion display ads appear alongside the chat text and are a display version of sponsored suggestions. The expert noted that advertisers can directly repurpose existing creative assets, lowering the adoption barrier. This format is positioned for brand advertising, with success measured by a combination of increased awareness and traffic generation. Native shopping cart ads may be the closest form to an "AI agent advertisement." Users can view product details and complete checkout within the ChatGPT interface without redirecting to the seller's website. This format charges a percentage of the Gross Merchandise Volume. However, this closed-loop transaction model is also contentious. The expert indicated it suits commoditized products, but premium brands may hesitate as they lose opportunities for upselling, cannot reinforce their brand image through their official website, and are unable to collect user data for remarketing. Budget and Returns: The Gap Between Expectation and Reality Current ChatGPT ad budgets are entirely sourced from advertisers' experimental funds, having no impact on Google Search budgets yet. The expert pointed out that brand advertising budgets typically have a ceiling, and future spending might be reallocated from other channels. Performance-based budgets, however, could represent incremental spending if ChatGPT can consistently demonstrate improved Return on Ad Spend. Regarding ad performance, data is mixed. ChatGPT ads are generating "good click-through rates," but the Return on Ad Spend is inconsistent. With limited data volume and small sample sizes, drawing definitive conclusions is difficult. On pricing, the $60 CPM is significantly higher than social media and streaming video ads. However, the expert believes the comparison should be made against Google Search ads. Calculating based on a $2 cost-per-click for Google Search and an assumption of 200 conversions per thousand clicks, the cost per conversion would be $40. By this standard, ChatGPT's $60 CPM is not considered expensive. The core issue lies with the conversion rate. The average conversion rate for sponsored suggestions is below 0.5%, implying a cost per click of around $12. Based on this, the expert believes the Return on Ad Spend must improve for the platform to remain commercially viable. Current Chaos: Worse Than Early Google and Meta The expert bluntly described the current ChatGPT ad experience as "rushed" and "chaotic," even "worse than the early experiences of Google, Meta, or Snapchat." Key issues include: A lack of performance data makes it difficult for advertisers to measure and optimize their ad units. The expert expects this to be a priority fix for OpenAI. The absence of CRM integration means ads display static information—they continue to show even if a specific product is out of stock or unavailable. Fully manual operation is the biggest pain point. While Google and Meta offer automated tools for creative development, selection, and placement, ChatGPT advertisers currently must manually select all content. The expert believes OpenAI's generative AI and agent capabilities should easily address this. Limited bidding mechanisms are another major flaw. Currently, if two advertisers bid on the same ad slot, OpenAI only shows the ad deemed more relevant. In contrast, Google uses a combination of bid price and relevance to determine ad placement. Ad delivery and targeting issues are also prominent—in some instances, ChatGPT displays irrelevant ads to users. The expert speculates this might be because ChatGPT currently uses keyword targeting rather than its LLM capabilities to judge ad-to-content suitability, a problem that should be a "quick fix." Outlook: Potential Improvement in 6-12 Months Despite the current array of problems, the expert still anticipates that within the next 6-12 months, ChatGPT will evolve into a more scalable advertising platform. The minimum spending threshold is also expected to decrease, allowing more advertisers to participate in testing. For investors, however, the core question remains: When will ChatGPT advertising truly begin to erode Google Search's market share? The chaotic start and underwhelming returns suggest that day may arrive later than the market currently expects.