From Keyword to Tech Article Outline: This AI Prompt Kills Writer's Block
Dario Radečić
Staring at a blank document when you have a great article idea is the worst part of technical writing.
You know what you want to write about, but turning that single topic into a structured outline feels overwhelming. Most writers waste hours trying to figure out what sections to include, what order makes sense, and whether they're missing important points. The result? You either end up with a messy outline that doesn't flow, or you procrastinate and ditch the idea entirely.
AI can solve this problem in under a minute. Instead of struggling with structure, you give AI a topic and it analyzes top-performing articles on Google to see what works.
The result is a 90% ready outline in Markdown format. You just add your personal touch and start writing.
This article shows you the exact prompt that transforms any topic into a complete article outline.
Let's dive in.
Why Manual Outline Creation Wastes Your Time
Hours.
That's how long most writers spend figuring out article structure before they write a single word. You have a topic in your head, but the moment you open a blank document, everything falls apart.
Your brain jumps between different angles. Should you start with the basics or dive straight into advanced concepts? What examples should you include? Before you know it, you've spent an hour thinking about structure without writing a single word.
You don't need to figure this out from scratch. Thousands of articles already exist on your topic. The best ones follow proven patterns, and an LLM with internet access can analyze them for you.
This prompt fixes that. It helps you:
Skip the guesswork by analyzing what actually works in your niche
Get a research-backed structure in under a minute
Focus your energy on writing instead of planning
Build on patterns from top-ranking articles in your field
Let me show you the prompt next.
The Article Outline Researcher and Writer Prompt
Here's the prompt you can use with ChatGPT, Claude, or any other LLM that has a web search capability:
You are a professional tech writer. Your task is to analyze a topic submitted by the user and create an article outline that includes:
- A working article title
- A full markdown outline with:
- H2 sections
- H3 subsections under each H2 (if needed)
- Bullet points for each subsection explaining what will be covered
Each bullet point must describe a concept, insight, technique, or example that should be addressed. The outline must be complete enough to guide a writer through drafting the article without missing key points.
### Target audience
Tech professionals, software developers, data scientists, and other experienced IT practitioners.
### Research instructions
- Use the top 5 Google search results for the topic (or closely related terms)
- Only include blog posts and articles from reputable publications
- Exclude Reddit, Stack Overflow, forums, and Q&A sites
For each selected article:
- Extract the title, H2/H3 headings
- Analyze the content under each heading
- Derive relevant bullet points from the content (not just the heading text)
- Look for tone, structure, patterns, and scope
### Writing style
- Write in American English
- Use conversational and simple to understand language - most readers aren't native English speakers
- Use title casing for H2 headings (This is the Title Casing)
- Use sentence case for H3 headings (This is the sentence casing)
- Always use contractions (e.g., “don’t” instead of “do not”) to keep the tone natural.
- Avoid fluff words and filler content - the article is already dense and packed with technical jargon
### Output format (Markdown)
Return the article outline in this exact structure:
~
# [Working Article Title]
- [Intro: what the article will cover]
- [Intro: target audience context]
- [Intro: article value proposition]
## [H2 Section Title]
[One sentence describing this section’s purpose]
### [H3 Subsection Title]
- [Topic or concept to explain]
- [Key insight or point to emphasize]
- [Example or case study to reference]
### [H3 Subsection Title]
- ...
## [Another H2 Section Title]
[One sentence describing this section’s purpose]
### [H3 Subsection Title]
- ...
[Continue pattern for all sections...]
## Conclusion
- [Key takeaway to reinforce]
- [Next steps or action items]
~
The article should aim to be comprehensive and self-contained. Don’t rush or skip important topics. Include as many H2/H3 sections as needed to make it a complete resource, but avoid unnecessary sections—be economical and purposeful in your structure. Keep in mind that the final article should have 2000-5000 words, so keep the outline at a level that can contain that much content - don't exceed it.
### Final instruction
Now, ask the user to submit a topic (and any optional context or requirements). Do not ask them for anything else.
Testing the Prompt
Let's see how it works in practice.
Say you want to write an article comparing alternatives to Python's pandas library. Copy the prompt above and paste it into your LLM:
Once it asks for your input, explain your topic and add any optional context:
The AI takes about 30 seconds to perform web search and research. This isn't deep research - I tried that approach initially, but it kept ignoring my instructions and writing research documents instead of outlines. Useful, but not what we need.
The AI will present you with a Markdown article outline:
Copy this outline into a new Markdown document for refinement. AI delivers a 90% ready structure, but the final 10% is what makes it yours. Add your personal insights, adjust sections based on your experience, and include examples that show your expertise.
Talk soon,
Subscribe to my newsletter
Read articles from Dario Radečić directly inside your inbox. Subscribe to the newsletter, and don't miss out.
Written by