What this workflow actually automates
This isn't a "click one button, get a finished article" workflow — it's a pipeline that removes the repetitive parts (research gathering, first-draft writing, formatting, publishing) so a human editor spends their time on the part that matters: judgment and quality control.
The workflow, step by step
1. Trigger: A schedule node (say, every Monday at 9am) or a webhook from a content calendar tool kicks off the run.
2. Topic input: A Google Sheets or Airtable node reads your next queued topic and target keyword.
3. Research step: An HTTP node calls a search/research API (or a Perplexity-style tool) to pull current facts and top-ranking competitor angles on the topic.
4. Draft generation: An AI node (OpenAI, Claude or your model of choice) drafts the article using the research as grounding context, following a prompt template that encodes your brand voice and structure rules.
5. SEO pass: A second AI node — or a call to a tool like NeuronWriter's API — checks the draft against target keywords and suggests edits.
6. Human review queue: Instead of auto-publishing, the workflow posts the draft to a Slack channel or a "Review" status in your CMS. This step is non-negotiable — publishing AI drafts without a human check is how sites end up with factual errors and thin content penalties.
7. Publish: Once approved, an HTTP node pushes the final post to your CMS (WordPress, Webflow, etc.) via its API.
Why the human-review step matters more than the automation
The real risk with AI auto-publishing isn't that it's slow — it's that it's fast enough to publish errors at scale before anyone notices. Keep a human in the loop for anything that touches facts, pricing, or claims about your product.
Getting started
n8n's free self-hosted tier is enough to build this entire workflow at zero software cost — you'll only pay for the AI API calls themselves, which for a weekly blog post workflow is a small, predictable cost.