Nov 24, 2024
Step by step
The process begins with selecting a niche and keywords. The speaker uses an AI niche hunter tool (link provided in the video description) to identify sub-niches and related keywords within a chosen niche. For example, choosing "Smart Home Devices" might lead to sub-niches like "Smart Home Security," from which you'd select a cluster of 20 keywords.
Next, the automation is built in Make.com. This involves connecting various modules in a specific sequence:
Webhook Trigger: A custom URL is created; clicking this URL initiates the entire automation process. A closing window module is also added for user experience.
Airtable Data Source: The automation pulls data from Airtable, specifically the "Record ID" of a particular article (each article has a unique record). This record contains the target keyword.
Claude AI (Outline Generation): Claude AI receives the keyword as a prompt and generates a detailed SEO-optimized content outline in JSON format. This step uses a custom AI prompt (available via a link in the video description). The speaker reinforces the keyword for Claude by including it in a second message. A JSON creator module then formats the outline into a structured JSON data.
Claude AI (Article Writing): Claude AI uses the JSON outline to generate the main body of the article in Markdown format. The instruction here is to use the H2 headings from the outline to create a conversational main body, excluding the title, introduction, and conclusion (those will be added later).
OpenAI (Image Generation): OpenAI is used in two steps: First, to generate an image prompt based on the keyword. Second, to actually create the image using the generated prompt. This ensures a more relevant image than simply using the keyword alone.
Markdown to HTML Conversion: The Markdown article is converted into HTML format for compatibility with Google Docs.
Google Docs Integration: A Google Doc is created with a title derived from the keyword. The generated HTML, along with the image URL from OpenAI, is added to the Google Doc.
Airtable Update: The automation updates the corresponding Airtable record with information like the completed Google Doc URL, the image URL and the generated article text.
WordPress Publishing: Finally, the content is published as a draft in WordPress, allowing for manual review and final adjustments before publishing live.
The entire process is visual in Make.com. You can view each module's inputs and outputs, allowing for adjustments as needed. Importantly, the speaker emphasizes the need for human input at both the keyword selection phase (to focus on profitable topics) and at the end (to review and improve the AI-generated content). Tools like Ahrefs can help evaluate keyword competition, and the speaker strongly advises humanizing the content after generation, which would involve making it more engaging, adding unique perspectives and adding more images.
Keyword Research
Automation with Make.com
Content Creation and Formatting
Publishing
This detailed lecture walks through the process of setting up an AI-based blogging system, emphasizing the balance between automation and strategic human involvement for optimizing content creation and monetization.