This GPT has literally saved me ad accounts.

Key Concepts

• Custom GPT: A tailored version of the GPT model, configured with specific instructions and knowledge.
• Ad Policy (Facebook/TikTok/Google): The rules and guidelines governing acceptable advertising content on these platforms.
• Disapproval Rate: The estimated probability an ad will be rejected by the platform’s review process.
• Policy Violation: Content that breaks the platform’s advertising rules.
• Ad Copy: The text used in an advertisement.

---

Mr. Page Social Ads Policy GPT: A Detailed Overview

The video details the creation and testing of a custom GPT – “Mr. Page Social Ads Policy GPT” – designed to assist users in navigating the complexities of Facebook and TikTok (with planned expansion to Google) advertising policies. The GPT is built with the core instruction to function as an expert in these ad policies, specifically analyzing ad copy and images to predict potential disapproval and suggest revisions.

Functionality and Testing Methodology

The creator tested the GPT using an intentionally problematic ad. The initial ad copy stated, “This meta ads course will make you millions of dollars overnight,” accompanied by an image of cash. The GPT immediately flagged this ad, assigning it a 95% chance of disapproval. The reasoning provided centered on the use of “unrealistic expectations” – a direct violation of Facebook’s advertising guidelines. The GPT then offered a revised ad copy: “Unlock the potential meta ads with our comprehensive course.”

Iterative Refinement and Results

Despite the revised copy, the GPT still predicted a 70% chance of disapproval, this time due to the continued use of the cash image. The GPT identified the image as “insinuating unrealistic expectations in earning potential.” It recommended replacing the image with visuals depicting work-related activities – someone using a computer, attending a webinar, or studying.

Following this suggestion, the ad copy remained as “Unlock the potential meta ads with our comprehensive course,” but the image was changed to align with the GPT’s recommendation. The GPT then predicted the revised ad should have “no issues running and getting approved” by both Facebook and TikTok ad policies. This demonstrates the GPT’s ability to iteratively refine ad content based on policy feedback.

Practical Application and Limitations

The creator emphasizes that the GPT isn’t intended for every ad created, but rather as a “first line of defense” for ads facing specific policy issues. It’s presented as a brainstorming tool to help circumvent policy violations. The video highlights the importance of understanding why an ad is flagged, not just blindly following suggestions.

Connection to Larger Course

The development of this GPT is linked to a larger, in-development Facebook ads course. The creator mentions a dedicated module within the course specifically addressing ad policy. Access to this course is available via a pre-save link in the creator’s profile.

Notable Quote

“If you’re having issues with a specific policy, this can be a great first line of defense and help you brainstorm some creative ways to get around policy violations.” – The creator, describing the GPT’s intended use case.

Technical Terms Explained

• GPT (Generative Pre-trained Transformer): A type of large language model capable of generating human-quality text. Custom GPTs are tailored versions of these models.
• Meta Ads: Advertising on platforms owned by Meta (formerly Facebook), including Facebook and Instagram.

Synthesis

The Mr. Page Social Ads Policy GPT represents a practical application of AI to address a common pain point for digital marketers: navigating complex and often ambiguous ad policies. The video demonstrates the GPT’s ability to identify potential violations, provide actionable feedback, and iteratively refine ad content to increase the likelihood of approval. While not a replacement for thorough policy understanding, it offers a valuable tool for brainstorming and troubleshooting ad compliance issues.