Automating Social Media Content with LLMs and n8n

Most small businesses know they should post on social media regularly, but the reality of creating quality content daily across LinkedIn, Twitter, Instagram, and Facebook is overwhelming. This n8n workflow solves that problem by generating a week of content in minutes, complete with platform-specific formatting, hashtags, and scheduling.

The Content Generation Pipeline

The workflow starts with a weekly trigger every Monday morning. An LLM generates 5-7 content ideas based on your niche, recent industry news, and top-performing past posts. Each idea is then expanded into platform-specific posts — short and punchy for Twitter, professional for LinkedIn, visual-first for Instagram. The workflow outputs all content into a Google Sheet for review.

Key Workflow Components

• Schedule Trigger — fires every Monday at 8 AM to generate the weekly content batch
• RSS Feed Node — pulls recent articles from industry blogs for content inspiration
• LLM Node — generates content ideas and writes platform-specific posts
• Google Sheets Node — stores all generated content for human review and approval
• Buffer/Hootsuite API — schedules approved posts for publishing throughout the week

Adding Human-in-the-Loop Review

While AI-generated content is impressive, a human review step ensures brand consistency and catches occasional hallucinations. The workflow sends a Slack notification with the content batch for review. A team member approves, edits, or rejects each post directly in the Google Sheet. A second n8n workflow monitors the sheet and automatically schedules approved posts.

This pipeline reduces social media content creation from 5-10 hours per week to under 30 minutes of review time. Scale it further by adding image generation via DALL-E or Stable Diffusion, A/B testing different post styles, and training the LLM on your brand voice with few-shot examples.