Most AI writers fail miserably when you try to force them into strict SEO guidelines while maintaining a high word count. We have all seen tools generate generic fluff that might hit a length target but fails every single readability check that matters for ranking.
I recently stumbled upon a goldmine of a post by an innovator on Reddit who didn’t just write a prompt; the author engineered a complete system for bulk-producing long-form content that adheres to strict SEO rules. This expert designed a workflow that combines deep content generation with the technical rigidity required by plugins like Yoast.
The “Yoast-First” Architecture
The brilliance of this approach lies in how the creator reversed the typical writing process. Usually, you generate text and then fix the SEO. The original poster built a prompt that acts as a compliance officer before writing a single word of the body content. It forces the AI to structure the article based on SEO logic: slugs, meta descriptions, and keyword density, rather than just narrative flow.
This isn’t just about dumping keywords into a paragraph. The author of the prompt specifically instructed the AI to handle complex formatting tasks that usually require manual editing, such as creating markdown tables, avoiding heading level mentions, and integrating LSI (Latent Semantic Indexing) keywords naturally. It solves the biggest bottleneck in AI blogging: the time spent cleaning up the output.
📌 Why This Prompt Architecture Wins
Strict Readability Constraints
One of the hardest things to get an LLM to do is write simply. Left to its own devices, AI loves complex sentence structures and passive voice. The creator of this workflow explicitly added constraints to keep passive voice under 10% and transition word usage above 30%. These aren’t random numbers; they are the specific metrics Yoast uses to judge readability. By baking these instructions into the system prompt, the expert ensures the output sounds punchy and human rather than academic and robotic.
Forced Structural Depth
The prompt doesn’t just ask for a long article; it demands a comprehensive outline with 25+ headings and subheadings before writing begins. This is a crucial “chain of thought” technique. By forcing the model to plan the skeleton of a 3,000-word piece first, the author prevents the AI from losing the plot halfway through. The requirement to include tables, lists, and FAQs ensures the content is visually broken up, which increases time-on-page metrics for human readers.
The Bulk Execution Strategy
Writing one optimized article is great, but the real power here is the bulk workflow the LinkedIn user shared. By setting up variables like `[PROMPT]` in the master script, you can feed a CSV file containing hundreds of keywords, target audiences, and tones into the system. This transforms a writing task into a database management task. You aren’t copy-pasting prompts; you are running a script that iterates through your keyword strategy automatically.
💡 The Master Prompt
Here is the refined prompt that the original poster shared. You can copy this directly into your preferred LLM.
System Instructions:
“Using markdown formatting, act as an Expert Article Writer and write a fully detailed, long-form, 100% original article of 3000+ words using headings and sub-headings without mentioning heading levels. The article must be written in simple English, with a formal, informative, optimistic tone.
Output this at the start (before the article):
Focus Keywords: SEO-friendly focus keyword phrase within 6 words
Slug: SEO-friendly slug using the exact [PROMPT]
Meta Description: within 150 characters, must contain exact [PROMPT]
Alt text image: must contain exact [PROMPT], describes the image clearlyOutline requirements:
Before writing, create a comprehensive Outline for [PROMPT] with 25+ headings/subheadings. Put the outline in a table, include natural LSI keywords, and ensure complete topic coverage.Article requirements:
Include a click-worthy title containing a number, power word, and sentiment word.
Write the Meta Description immediately after the title.
Ensure [PROMPT] appears in the first paragraph and as the first H2.
Write 600–700 words under each main heading.
Add at least 1 table (comparison, checklist, steps, etc.) and 6 FAQs.
End with a clear Conclusion.On-page / Yoast-style checks:
Keep passive voice ≤ 10%.
Use transition words often (aim 30%+).
Include [PROMPT] in at least one subheading and naturally 2–3 times across the text (density ~1.3%).Link suggestions:
Add Inbound link suggestions (3–6 internal pages) and Outbound link suggestions (2–4 credible sources) after the conclusion.Now generate the article for: [PROMPT]“
✅ How to Run the Bulk Workflow
The creator also outlined a method to scale this using a simple CSV file. This is perfect if you are managing a content calendar for multiple clients or sites.
The Setup:
Create a CSV file with columns for: `keyword`, `country`, `audience`, and optional fields like `tone` or `internal_links`.
Populate the rows with your SEO research (20–200 keywords).
The Execution: Use a tool that allows variable replacement (like ChatGPT Team/Enterprise or an API wrapper script). For every row in your CSV, the script replaces the `[PROMPT]` placeholder in the master prompt with your specific keyword.
This method ensures that every single article follows the exact same high-quality structure: title logic, meta data, and formatting, while being unique to the specific keyword topic.
This is a massive efficiency boost for anyone in content marketing. You can find the original discussion and more details on the specific tools the author uses in the full post below.
💡 FAQ & Troubleshooting
How do I use this prompt for bulk article generation?
To generate articles in batches, create a CSV file containing columns for your keyword, country, audience, tone, internal links, and external sources. For every row in your CSV, replace the [PROMPT] placeholder with your specific keyword to generate the articles sequentially using the defined rules.
What specific SEO metrics does the prompt enforce?
The prompt is designed to pass specific Yoast-style checks. It instructs the AI to keep passive voice usage below 10%, ensure transition words are used in at least 30% of sentences, and maintain a natural keyword density of approximately 1.3% without keyword stuffing.
What is the expected length and structure of the output?
The prompt requests a long-form article exceeding 3000 words. It requires an initial table outline with 25+ headings/subheadings, followed by content sections of 600–700 words each. The structure must also include a click-worthy title, meta description, at least one data table, and a dedicated FAQ section.
Made a bulk version of my Yoast article prompt (includes the full prompt + workflow)
byu/LongjumpingBar in