# LinkedIn Post Creation Rules for Marcos Heidemann

## 🚫 **NEVER DO:**

### 1. Role-Based Opening Lines
- ❌ "As a Principal [anything]..."
- ❌ "As someone working in [industry]..."
- ❌ "In my role as [title]..."
- ❌ "Being a [profession]..."

**Why:** These openings are cliché and don't reflect authentic personal voice.

### 2. Statistics Without Sources
- ❌ Any statistic, percentage, or data point without a valid URL source
- ❌ Vague attributions like "recent study shows" or "experts say"
- ❌ Unverifiable claims presented as facts

**Why:** Credibility is paramount for technical professionals.

## ✅ **MANDATORY REQUIREMENTS:**

### 1. Source Attribution for ALL Statistics
- **Format:** Every statistic MUST include a valid, clickable URL
- **Placement:** Include sources either inline or at the end of the post
- **Verification:** Sources must be verifiable and current
- **Example:** 
  ```
  42% of executives say GenAI is "tearing their company apart" 
  (Source: https://example.com/actual-study-link)
  ```

### 2. Authentic Opening Styles
Instead of role-based openings, use:
- **Direct insight:** "Here's what caught my attention about..."
- **Observation:** "Something interesting happened in the AI space..."
- **Question-driven:** "Why are we seeing this pattern in..."
- **News reaction:** "The latest developments in [topic] reveal..."
- **Personal take:** "I've been thinking about..."

## 📝 **PREFERRED POST STRUCTURE:**

### 1. Hook (First Line)
- Start with insight, observation, or intriguing statement
- Avoid generic professional introductions
- Make it specific and engaging

### 2. Context/Development
- Present the main topic or news
- Include relevant technical details
- Provide necessary background

### 3. Analysis/Perspective
- Share genuine insights or implications
- Connect to broader trends or challenges
- Avoid obvious statements

### 4. Engagement Question
- Ask specific, thought-provoking questions
- Target peer professionals (ML engineers, data scientists)
- Encourage technical discussion

### 5. Sources (MANDATORY)
- List all sources with valid URLs
- Use format: "Sources:" followed by numbered list
- Ensure links are accessible and current

## 🎯 **TONE AND STYLE GUIDELINES:**

### Voice Characteristics:
- **Technical but accessible** - Use proper terminology without being overly academic
- **Curious and analytical** - Show genuine interest in understanding implications
- **Balanced perspective** - Acknowledge both opportunities and challenges
- **Conversational but professional** - Avoid corporate speak

### Content Preferences:
- **Focus on implications over announcements** - What does this mean rather than what happened
- **Technical depth** - Include specific metrics, architectures, or methodologies when relevant
- **Real-world connection** - Bridge research/announcements to practical applications
- **Critical thinking** - Don't just celebrate developments, analyze them

## 📊 **SOURCE REQUIREMENTS:**

### Acceptable Sources:
- ✅ Peer-reviewed research papers with DOI links
- ✅ Official company announcements with direct URLs
- ✅ Reputable tech publications (with specific article links)
- ✅ Government reports with direct PDF/webpage links
- ✅ Industry surveys with methodology documentation

### Unacceptable Sources:
- ❌ "Recent studies" without links
- ❌ Social media posts as primary sources
- ❌ Paywalled content without accessible alternatives
- ❌ Dead links or temporary URLs
- ❌ Secondary reporting without primary source verification

## 🔧 **HASHTAG STRATEGY:**

### Core Tags (Always Include):
- `#MachineLearning`
- `#ArtificialIntelligence` 
- `#MLOps`

### Contextual Tags (Choose 2-3 based on content):
- `#TechLeadership` (for strategy/management topics)
- `#GenAI` (for generative AI content)
- `#DataScience` (for data-focused posts)
- `#Innovation` (for breakthrough/research topics)
- `#TechTrends` (for industry analysis)

### Total Hashtag Limit: 5-7 maximum

## 📋 **POST LENGTH GUIDELINES:**

### Optimal Structure:
- **Hook:** 1 line
- **Context:** 2-3 sentences
- **Analysis:** 2-4 bullet points or short paragraphs
- **Engagement question:** 1-2 sentences
- **Sources:** As needed
- **Hashtags:** 5-7 tags

### Character Count Target:
- **Minimum:** 300 characters (for algorithm visibility)
- **Optimal:** 600-1000 characters (best engagement)
- **Maximum:** 1300 characters (before "see more" truncation)

## 🎭 **ENGAGEMENT OPTIMIZATION:**

### Question Types That Work:
- **Technical implementation:** "How are you handling [specific challenge]?"
- **Experience sharing:** "What's been your experience with [technology/approach]?"
- **Prediction/opinion:** "Where do you see [trend] heading?"
- **Problem-solving:** "What solutions have you found for [specific issue]?"

### Avoid Generic Questions:
- ❌ "What do you think?"
- ❌ "Agree or disagree?"
- ❌ "Thoughts?"

## 🔄 **QUALITY CONTROL CHECKLIST:**

Before posting, verify:
- [ ] No role-based opening ("As a...")
- [ ] All statistics have valid source URLs
- [ ] Sources are accessible and current
- [ ] Authentic voice and genuine insights
- [ ] Specific engagement question for ML/DS professionals
- [ ] 5-7 relevant hashtags
- [ ] 600-1000 character target met
- [ ] Technical accuracy verified
- [ ] No corporate speak or generic language

## 📚 **EXAMPLES OF GOOD OPENINGS:**

Instead of "As a Principal ML Engineer, I've been following AI developments..."

Use:
- "The latest AI benchmarks reveal something unexpected..."
- "Here's why the $2B funding for Thinking Machines caught my attention..."
- "Something's not adding up in the enterprise AI adoption numbers..."
- "The gap between research and production is widening, and here's why..."
- "While everyone's celebrating Claude 4's benchmarks, we're missing the bigger picture..."

---

**Last Updated:** July 28, 2025
**Review Schedule:** Monthly updates based on feedback and performance analysis