Building a Social Media Content Engine with Claude, Notion, and n8n
Social media consistency is one of those things that everyone knows matters and almost no one actually maintains. The problem isn't lack of content. Most organizations have plenty to say. The problem is the friction between having something to say and actually saying it.
I just built something that removes most of that friction, and I did it entirely through conversation with Claude using a set of connected tools. The result is a system where I can batch-create weeks or months of social posts in a single focused session, schedule them in Notion, and let automation handle the rest.
The Problem with Social Media Workflows
The typical social media workflow looks something like this: you have an idea, you open Twitter or LinkedIn, you write something, you post it. Simple enough for one post. But consistency requires doing this repeatedly, which means repeatedly context-switching into "social media mode." That mental overhead adds up, and eventually the posting stops.
The alternative is scheduling tools like Buffer or Hootsuite. These work, but they introduce another interface to manage, another login to remember, another place where content lives separate from everything else.
What I wanted was different. I wanted social media scheduling to live where I already work. I wanted to create posts in bulk when I'm in the right headspace, then forget about them until they go live.
The Solution: Notion as a Content Calendar
The answer turned out to be surprisingly straightforward. I created a database in Notion called Social Posts. Each entry has the content I want to post, the platform (Twitter, LinkedIn, or both), a publish date, and a status field.
The status field is the key. When I'm drafting posts, they stay in "Draft" status. When they're ready to go live, I flip them to "Ready." An automation checks hourly for posts that are marked Ready with a publish date in the past, sends them to the appropriate platform, and updates the status to "Posted."
That's it. I now have a single place where I can see everything that's scheduled, everything that's been posted, and everything I'm still working on.
Where the Content Comes From
Here's where it gets interesting. Gadoci Consulting has a substantial library of articles, frameworks, and educational content. All of that material contains ideas worth sharing on social media, but extracting those ideas has always been manual work.
Now I can have Claude read through existing articles and generate social posts based on key concepts, interesting quotes, or actionable insights. In one session, I can create dozens of posts that are all grounded in content that already exists. The posts aren't generic filler. They're drawn from real intellectual property.
This transforms social media from a constant creative burden into a batch processing task. Spend a few hours creating a month's worth of content, schedule it all, and move on to other work.
The Technical Stack
The automation runs on n8n, an open-source workflow automation tool. Every hour, a scheduled workflow queries the Notion database for posts that are ready to publish. It filters them by platform and sends each post to the appropriate social network through webhook-triggered workflows I'd already built. After posting, it updates the Notion record to reflect the new status.
What's notable is how I built this. The entire system was designed and implemented through conversation with Claude, using MCP (Model Context Protocol) connections to Notion and n8n. I described what I wanted, Claude helped architect the solution, and we built it step by step without me leaving the conversation interface.
This is what L2 automation looks like in practice. Not a massive enterprise deployment, but a practical workflow that solves a real problem with tools that connect to each other.
Why This Matters Beyond Social Media
The specific use case here is social media scheduling, but the pattern is what matters. The pattern is: take a task that requires repeated manual effort, find where the friction lives, and build a system that removes that friction while keeping you in control of the content.
Social posts still require human judgment about what's worth saying. The automation doesn't generate content on its own. It just handles the scheduling and posting so I can focus on the creative work in concentrated batches rather than scattered fragments.
This is the difference between using AI tools and operationalizing AI. Using Claude to write a single social post is helpful. Building a system where Claude helps create content that flows into a scheduling database that triggers automated posting is a workflow. Workflows compound. One-off uses don't.
Getting Started
If you want to build something similar, the components are:
A Notion database with fields for content, platform, publish date, and status. This becomes your content calendar and the single source of truth for what's scheduled.
An n8n workflow (or equivalent automation tool) that queries Notion on a schedule, filters for posts ready to publish, and triggers the actual posting through platform APIs or webhooks.
A connection to Claude that lets you generate content at scale. This could be through the API, through Claude's interface with MCP connections, or through whatever tooling fits your workflow.
The technical implementation took a few hours. The value is ongoing. Every week, I can batch-create content without thinking about the posting mechanics. The system handles that part.
The Bigger Picture
We talk a lot about AI transformation as if it requires massive initiatives and enterprise-wide change management. Sometimes it does. But sometimes transformation looks like this: one person, one afternoon, building a system that saves hours every week and actually ships.
The tools exist. Claude can connect to your databases and automation platforms. The question is whether you'll use that capability to solve real workflow problems or just ask it questions one at a time.
I know which approach compounds.