Creating consistent, high-quality blog content is one of the most powerful ways to grow your brand and build SEO authority. But if you're also producing YouTube videos or podcasts, you know how difficult it can be to repurpose that content efficiently.

Instead of manually converting video transcripts into articles, I built an automation pipeline that does it for me. With a simple setup using n8n, Google Drive, ChatGPT, and WordPress, I now generate full blog drafts from transcripts with almost no manual input.

Here's exactly how it works—and how you can build your own version of it.

Why Automate Blog Creation?

If you're already creating video content, you have a library of material waiting to be repurposed. But transcribing and formatting manually is inefficient and repetitive. Automating this process offers several benefits:

  • Save hours of content repurposing time
  • Maintain a consistent blogging schedule
  • Eliminate human error during formatting
  • Scale your content operations with minimal resources

This workflow takes a transcript and turns it into a formatted, SEO-friendly blog post—all posted directly into WordPress as a draft, ready for review.

Tools Used in the Workflow

To make this system work, I used the following:

  • n8n: A no-code/low-code workflow automation platform.
  • Google Drive: For storing and triggering new transcripts.
  • OpenAI ChatGPT (via GPT Assistant): For writing the actual blog content.
  • WordPress: The publishing platform where posts are created as drafts.

Step-by-Step Breakdown of the Workflow

Step 1: Monitoring Google Drive for New Transcripts

The automation starts with n8n continuously monitoring a specific Google Drive folder. This is where my video editor drops newly created transcript files after editing a video.

  • Trigger type: File Created
  • Interval: Every 1 minute
  • Folder monitored: "Video Transcripts"

As soon as a new file appears, the workflow begins.

Step 2: Downloading and Extracting the Transcript

Using the “Download File” node in n8n, the system grabs the uploaded transcript file using its ID. Then, with the “Extract from File” node, it pulls the plain text from the document—ready for processing.

This step requires no special formatting. As long as the transcript is readable, the system handles it.

Step 3: Generating the Blog Post with ChatGPT

Next, the extracted transcript is sent to a pre-trained OpenAI GPT Assistant. The assistant uses a structured prompt to turn the transcript into a formatted blog post.

Prompt Highlights:

  • Minimum 750 words
  • Includes title, overview, main content, and conclusion
  • Business casual tone
  • Output in WordPress-ready HTML markdown
  • Slug included for SEO purposes
  • JSON format: title, content, slug

The assistant is fine-tuned with 13 previous blog examples stored in vector memory, helping it maintain consistent structure and tone.

Step 4: Parsing the JSON Output

The response from ChatGPT is in JSON format, but n8n requires each field to be separate. So, I created a quick function to split the JSON into its components:

  • Blog Title
  • Blog Content (formatted in HTML)
  • URL Slug

This ensures that everything is cleanly mapped for WordPress publishing.

Step 5: Pushing the Draft to WordPress

Once the blog content is generated and formatted, it's automatically pushed into WordPress as a draft post. For this, n8n uses the WordPress node with custom credentials.

Important Setup Notes:

  • You must use your default WordPress login path (typically wp-admin). If it's customized, the API may fail.
  • The “password” required is not your actual login password. Instead, go to your WordPress profile settings and generate a new Application Password.
  • Once credentials are in place, map the title, content, slug, and author fields.

After configuration, n8n sends the post to WordPress, where it appears instantly in the Drafts section—ready for human review.

Testing the Workflow

To test the system, I uploaded a sample transcript file to the connected Google Drive folder. After about one minute (due to the node polling interval and GPT response time), the draft appeared in WordPress.

It worked exactly as intended. The title, content, and formatting looked good, though I still recommend a final human pass before publishing for grammar or tone adjustments.

Key Takeaways and Tips

  • Start with a clean, consistent transcript. The better your transcript quality, the better your AI-generated blog will be.
  • Train your GPT assistant well. Using a few examples of past blog posts improves output dramatically.
  • Keep the workflow modular. Each step (fetching, generating, formatting, posting) can be tested individually.
  • Use drafts, not auto-publish. This gives your team a chance to review and refine the content before going live.

Frequently Asked Questions

(1) How does n8n detect new transcripts in Google Drive?

n8n uses a trigger node configured to watch a specific Google Drive folder. It checks the folder every minute for any new file uploads. When a new transcript file is detected, it automatically initiates the blog creation workflow.

(2) Do I need coding knowledge to set this up?

Not necessarily. n8n is a low-code automation tool, and the process largely involves configuring nodes using its visual interface. While some light scripting can help fine-tune specific parts (like parsing JSON), the entire workflow can be created without writing full programs.

(3) Can this workflow work with other file types or just plain text?

Yes, this workflow can support different file formats like .txt, .docx, or even PDFs if configured properly. The key is using the appropriate node in n8n to extract text from the file. However, clean, plain-text transcripts usually yield the best results when generating blog content.

(4) Is the content generated by ChatGPT ready to publish?

The content is generally 80 to 90 percent ready. The GPT assistant does a solid job with structure and tone, especially when trained with sample blogs. That said, it's best to have a human editor review and make final tweaks to ensure clarity, accuracy, and brand consistency.

(5) What if my WordPress site uses a custom login URL?

If your WordPress admin URL has been changed from the default wp-admin, the API connection might fail. You must use the standard wp-admin path when setting up the credentials in n8n to successfully connect and push content.

(6) Is it possible to extend this workflow further?

Absolutely. You can add extra steps to the workflow, such as automatically generating featured images, scheduling posts, assigning categories and tags, or even triggering email notifications when a new draft is created. The flexibility of n8n and the OpenAI API makes it easy to scale and customize the process based on your publishing needs.

Final Thoughts

This workflow has saved me hours of manual work each week. Whether your team is small or you're running content solo, automating blog creation with n8n and ChatGPT can help you stay consistent and scale up without burning out.

If you work with video content regularly and want to keep your blog active without writing from scratch, this system could be a game-changer.

Want to skip the setup? I've shared the full workflow on our blog—download it, plug in your credentials, and you're ready to go.

Let me know in the comments if you'd like a follow-up on adding images, SEO meta fields, or auto-publishing. There's a lot more we can automate from here.