As OpenAI aggressively scales its new advertising business, early enterprise partners are voicing a consistent frustration: while the conversational intent is high, the measurement tools are stuck in the “spreadsheet era.”
Despite paying a premium $60 CPM (comparable to NFL Sunday Night Football rates) and committing to a $200,000 minimum spend, early advertisers report that OpenAI currently provides almost no data to prove a direct Return on Investment (ROI).
1. The Measurement “Black Box”
Performance marketers, accustomed to the granular attribution of Google and Meta, are finding ChatGPT’s current reporting to be severely limited.
- Basic Metrics Only: Early testers receive only weekly CSV reports containing two data points: total impressions and total clicks.
- The Attribution Gap: There is currently no native way to track conversions, demographic insights, or purchase attribution. Advertisers cannot see which specific prompts triggered their ads or the context of the conversations they appeared in.
- Low Click-Through Rates (CTR): While Google Search averages a 29% click-through rate for high-intent queries, ChatGPT’s ads are seeing a measly 1–3% CTR. This is because ChatGPT’s primary goal is to resolve a query within the interface, often removing the need for a user to click away to a brand’s website.
2. Technical and Operational Hurdles
Beyond the data gap, the “Ads Manager” is still in its infancy, leading to significant logistical friction for early adopters.
- Manual Execution: There is no fully automated self-serve platform yet. Large brands are reportedly exchanging spreadsheets and emails with OpenAI staff to launch and adjust campaigns.
- Budget “Under-Spend”: One agency executive revealed that some clients have only spent 15–20% of their committed budgets because the ad-matching logic is so restrictive that impressions aren’t being triggered frequently enough.
- Lack of Programmatic Access: OpenAI has no current plans to introduce real-time bidding or programmatic auctions, keeping the ecosystem closed and curated for now.
3. The “Conversation Insights” Roadmap
To address these complaints, OpenAI has teased a more robust “Conversation Insights” dashboard, expected to roll out later in 2026.
| Feature | Current Status (March 2026) | Planned for Late 2026 |
| Reporting Format | Manual CSV / Weekly | Real-time Dashboard (Ads Manager) |
| Primary Metric | Impressions / Clicks | Intent Signals & Sentiment Score |
| Attribution | None (Post-click only) | In-Chat Purchase Integration |
| Targeting | Basic Contextual | Agentic Upselling & Semantic Mapping |
4. Why Brands Are Still Staying
Despite the “blind” spending, the $200,000 entry bar hasn’t deterred major players in travel, finance, and e-commerce.
- First-Mover Advantage: Brands want to “train” the AI to recognize their products as the authoritative answer before the platform becomes crowded.
- Brand Safety: OpenAI has strictly blocked ads from appearing alongside sensitive topics like politics, health, and mental health, making it one of the cleanest environments for high-end brands.
- High-Value Audience: Ads are specifically targeted at the Free and “Go” ($8/month) tiers, capturing a massive demographic that is actively looking for answers and recommendations.
“OpenAI is building some of the most sophisticated AI on earth, but their ad reporting is stuck in 2005,” noted one marketing lead. “We’re flying blind, but we’re flying in a very high-value direction.”
