Testing the article generation process

Step-by-Step Process

Step 1: Primary Keyword Input

1. Keyword Upload
   Keywords will be added by the admin via bulk upload or manually, forming the list for the Cat-Site.
2. Keyword Retrieval
   The system retrieves keywords from the Cat-Site-specific keyword library.
3. Single Keyword Processing
   • The list is randomized.
   • One keyword is picked (usually from the top) and marked as used.
   • The picked keyword is removed from the pool until 50% of keywords have been used.
   • After 50% usage, the system randomizes the list again and makes all keywords available for use.
4. Usage Table
   • The Usage Table will display the order in which keywords are used.
   • Admin can edit, reorder, add, or remove keywords at any time.
   • This table is created after the list is randomised and updated every time the 50% usage mark is reached.

Step 2: Keyword Cluster Generation

Goal:
Expand the primary keyword with a cluster of 10–15 semantically related keywords to enhance the SEO and content relevance.

• From the primary keyword, generate 9 additional related keywords using AI. The number of keywords is dynamic, with the default set to 5 but can go up to 15.
• These keywords should be SEO-optimized, relevant, and aligned with the primary keyword.

AI Prompt Example:
“You are a top-tier SEO strategist. Given the primary keyword ‘federal retirement’, generate a cluster of 15 semantically related keywords that are highly relevant for the financial services industry, commonly associated with search engine results, and ideal for internal linking and building authority.”

Notes:

• The generated keywords will be stored in a JSON array.
• Future content can randomize keyword clusters for internal linking or content expansion.

Step 3: Title Generation Using AI Assistant

Goal:
Create 5 SEO-optimized titles using the primary keyword and keyword cluster.

• Titles should be click-worthy, emotional, informative, and relevant to users aged 45–65.
• Each title must:
  • Be under 70 characters.
  • Contain the primary keyword.
  • Be non-salesy, educational, and resonate with the target audience.

AI Prompt Example:
“Create 5 compelling, SEO-friendly titles using the primary keyword ‘KEYWORD’ and relevant correlative keywords. Ensure each title is under 70 characters and resonates emotionally with readers aged 45–65.”

System Notes:

• The top-ranked title will be marked as Default Title.
• Admin can edit, delete, or replace any title if needed.
• The system logs title performance for future testing and optimization.

Step 4: Article Structure Creation

Goal:
Generate an optimized article structure based on the primary keyword and its cluster.

How It Works:

• The system creates a clear article structure with:
  • Introduction hook.
  • 4-7 H2 sections based on common search intents.
  • Suggested H3s for subtopics where necessary.
  • A Key Takeaways section at the end.

AI Prompt Example:
“Using the keyword ‘KEYWORD’ and related terms, create an SEO-optimized article structure with an introduction, 4–7 H2s, supporting H3s, and a key takeaways section.”

System Notes:

• The structure is saved as a content schema or article template.
• H2s are used as anchor points for internal linking.

Step 5: Loop & Continuation

Goal:
Repeat steps 2-4 until the required number of Default Titles is generated.

• Article count should match the Cat-Site schedule + 20% buffer.
• Titles are created immediately after the CRON job is run.
• Admin can edit or replace titles as necessary.
• The system will generate articles in the order of the oldest title first.
• Pending Article Title List: A dynamic dropdown showing all available Cat-Sites and titles for selection.

Step 6: Article Creation

Goal:
Generate a 1,000-word educational article based on the selected title and structure.

AI Prompt Example:
“Write a 1,000-word educational article titled: ‘[SELECTED TITLE]’ using the structure: [OUTLINE]. The article should be professional, educational, and designed for federal employees aged 45–65. Include a call-to-action and a Key Takeaways section.”

System Notes:

• Articles will be saved with keyword, title, and structure metadata.
• The system will generate slug, canonical tag, and meta title/description for SEO.

Step 7: Metadata & Schema Generation

Goal:
Generate structured metadata for Google and AI crawlers.

AI Prompt Example:
“Using the following article, generate: SEO meta title (under 70 characters), meta description (under 155 characters), and JSON-LD schema markup for a BlogPosting.”

System Notes:

• This metadata will be automatically inserted into the <head> tags or CMS fields.

Step 8: Proofreader Review

Goal:
The proofreader reviews and approves the article before publishing.

What Proofreaders Do:

• Edit or regenerate the article if needed.
• Add or change the image.
• Approve or reject the article.

System Notes:

• If an article is not manually reviewed, it is automatically approved after a set time.
• Proofreaders also verify grammar, readability, and factual accuracy.
• The system tracks edits and approvals.

Step 9: Fail-Safe & Auto-Publishing

Goal:
Ensure content is published even if proofreading tasks are delayed.

• 2 days before publishing: PR is notified.
• 1 day before publishing: Admin is notified.
• If no PR review is done, the system auto-selects pre-generated articles for publishing.
• Articles not reviewed are flagged with a ‘PR’ tag.

Notification Example:
“For <CatSite1>, the Proofreader failed to review these articles: Title1, Title2, Title3. These articles are now in the ‘Available for Sponsorship Library.'”