Descriptions:
The Ben AI channel walks through three production-ready AI systems for Meta advertising management, built with n8n and Airtable. Each system has been sold to marketing agencies, e-commerce brands, and SaaS companies, and the video provides enough architectural detail to understand how to replicate them.
The first is an automatic product ad generator. A user uploads a product image and notes to an Airtable interface, which triggers an n8n workflow that uses Perplexity to research the product, generates five ad prompt variations via an LLM, and produces actual ad images through NanoBanana Pro hosted on Replicate — all for just cents per generation. An optional reference competitor ad can guide the visual style. Human approval is built in before images are generated, and a final workflow produces matching ad copy.
The second system is a Meta ad intelligence tool that automatically scrapes all of a competitor brand’s active ads from the public Facebook Ad Library on a weekly schedule, analyzes image, video, and text creatives, and generates a strategy report summarizing competitor positioning. The video explains how to find the correct Facebook Ad Library URL for any competitor, using Gymshark as a live example. Both systems are template-based: cloning requires importing n8n JSON, duplicating the Airtable base, wiring webhooks, and connecting API keys for Perplexity, the chosen LLM provider, and Replicate.
📺 Source: Ben AI · Published January 13, 2026
🏷️ Format: Workflow Case Study
