n8n Workflow: Generate LinkedIn Posts with Gemini + Sheets

Generate 3 high-performing LinkedIn post options with Google Gemini + log results to Google Sheets

This n8n workflow takes your bullet points via a webhook, uses Google Gemini to generate three LinkedIn post variations, scores them for performance, saves everything in Google Sheets, and returns an HTML page showing the recommended post plus all alternatives.

What this workflow does

• Receives input: A POST request hits the webhook with bullet_points, tone, post_type, and an optional author_name.
• Normalizes and configures: The workflow normalizes fields and sets the target Google Sheets document ID and worksheet name (including an expected worksheet named “Posts”).
• Generates exactly three variations: Using Google Gemini, it creates three distinct LinkedIn post drafts from your bullet points and options.
• Scores each variation: Gemini is then used again to score each draft on hook strength, readability, and engagement potential.
• Selects a recommendation: It totals the three scores per variation, selects the highest-scoring version as the recommended post, and prepares a breakdown for output.
• Logs to Google Sheets: Appends the generated variations, totals, input metadata, and the recommended index into a worksheet.
• Returns HTML output: Responds with a formatted HTML page showing the recommended post and all three variations with their scores.

Use cases

• SaaS operators turning product updates (bullet points) into LinkedIn posts in a consistent tone.
• Automation engineers building a repeatable “content draft → evaluation → storage” pipeline using n8n.
• Teams testing different post types and choosing the best-performing draft automatically.

Technical details

• Nodes / components: set, code, webhook, sticky note, Google Sheets, respond to webhook.
• Integrations: Google Gemini (attached to both generation and scoring steps) and Google Sheets OAuth2.
• Google Sheets setup: Create a worksheet named “Posts” (or update the workflow) with headers including generated_at, bullet_points, tone, post_type, variation_1, score_1, variation_2, score_2, variation_3.