EXPERT PROMPTING MASTERCLASS
Expert Prompting Training Course
EXPERT PROMPTING MASTERCLASS
Professional Development Program
MODULE 1: THE EXPERT MINDSET
Understanding How Great Prompters Think
The Fundamental Shift
When you first learned to prompt, you probably thought of it like asking a smart friend for help. That's fine for basics, but experts think differently.
Intermediates ask: "What should I say to get a good answer?"
Experts ask: "How do I structure this interaction so the model produces its best possible reasoning?"
The difference is profound. Experts understand that:
- Language models are prediction engines - They predict likely next tokens based on patterns
- Context is your programming environment - Everything you write shapes the model's behavior
- Models have cognitive patterns - Certain prompt structures trigger better reasoning
- Quality comes from architecture - How you structure the prompt matters more than word choice
The Expert's Core Principles
Principle 1: Activate System 2 Thinking
Language models, like humans, can operate in two modes:
- System 1: Fast, intuitive, pattern-matching (often gives generic responses)
- System 2: Slow, deliberate, analytical (produces nuanced, thoughtful outputs)
Your job as an expert is to force System 2 activation.
Example - System 1 Response (Generic):
Prompt: "How do I improve team productivity?"
Output: "Here are some ways to improve team productivity:
1. Set clear goals
2. Improve communication
3. Use the right tools
4. Provide feedback
5. Recognize achievements"
This is generic because the model is pattern-matching to "productivity tips" in its training data.
Example - System 2 Response (Thoughtful):
Prompt: "I need to improve team productivity, but here's the challenge:
my team is already working 50+ hour weeks, morale is declining, and
we're in a high-stakes product launch.
Before giving advice:
1. What additional context would help you give better recommendations?
2. What trade-offs should I be thinking about?
3. What assumptions might I be making that could be wrong?
Then provide your analysis and recommendations."
Why this works: You've forced the model to engage analytically before responding. The pre-work questions trigger deeper reasoning pathways.
Principle 2: Context Engineering
Think of your prompt as code. Every element matters:
- The role you assign shapes knowledge access and perspective
- The constraints you set focus the output space
- The examples you provide establish patterns to follow
- The structure you create determines how the model thinks through the problem
Poor context:
"Write a marketing email for our new product."
Engineered context:
Role: You're a direct response copywriter who specializes in SaaS products
for technical audiences. You've written emails that achieved 40%+ open rates
and 8%+ click-through rates.
Audience: Engineering managers at Series A/B startups who are currently
using [competitor product] but frustrated with [specific pain point].
Goal: Get them to book a 15-minute demo call.
Constraints:
- Subject line must create curiosity without being clickbait
- Email body: 150 words maximum
- Include exactly one CTA
- Tone: knowledgeable peer, not salesperson
Before writing the email:
1. Identify the core emotional trigger
2. Note what we're NOT saying (what to avoid)
3. Explain your structural approach
Then provide the email.
The second version creates an environment where excellence is more likely.
Principle 3: Intelligent Iteration
Experts never accept first outputs. They iterate systematically:
The Expert Iteration Loop:
- Get baseline → Identify gaps → Add specific constraints → Re-generate
- Get variation → Compare approaches → Synthesize best elements → Refine
- Stress test → Find weaknesses → Strengthen → Polish
Example in action - Iteration 1 (Baseline):
"Explain blockchain to a non-technical audience."
[Get response, evaluate]
Iteration 2 - Add constraints:
"That's too abstract. Rewrite it:
- Use only one analogy (choose the best one)
- Include a concrete example of a real problem it solves
- Address the #1 objection people have
- Keep it under 200 words"
[Get response, evaluate]
Iteration 3 - Refine:
"Better. Now make it more engaging:
- Start with a surprising fact or question
- Remove any jargon that remains
- End with a clear 'so what' moment"
[Get response, evaluate]
Iteration 4 - Polish:
"Final pass: tighten the language. Every sentence should be essential."
[Get final response]
This takes 5-10 minutes but produces dramatically better results than spending 30 minutes trying to write the "perfect" first prompt.
Practice Exercises
✍️ PRACTICE EXERCISE 1.1: Activate System 2
Your turn. Take this generic prompt and transform it using System 2 activation:
Generic prompt:
"Give me ideas for reducing customer churn."
Your enhanced prompt should:
- Provide specific context about the business
- Include pre-work questions that force analysis
- Request structured thinking
- Set clear output requirements
Spend 5 minutes writing your enhanced version. Then test it in your preferred AI tool.
✍️ PRACTICE EXERCISE 1.2: Context Engineering
Scenario: You need to write a performance review for a team member who's technically excellent but struggles with communication.
Bad prompt:
"Write a performance review for John."
Your task: Engineer a complete context that includes:
- Your role and relationship
- Specific details about John's work
- The purpose and audience of the review
- Tone and format requirements
- Any constraints or sensitivities
Write your engineered prompt and test it.
The Expert Self-Assessment
Before moving to the next module, rate yourself honestly:
I understand that:
- ☐ Prompts structure thinking, not just request information
- ☐ Generic inputs produce generic outputs
- ☐ Iteration is where quality comes from
- ☐ Context engineering is a skill I can develop
- ☐ Different prompts trigger different reasoning modes
If you checked all boxes, you're ready for Module 2.
MODULE 2: ADVANCED PROMPTING FRAMEWORKS
The Techniques That Produce Excellence
Now that you understand the expert mindset, let's build your technical toolkit. These are battle-tested frameworks that work across all AI models.
Framework 1: Layered Prompting
The single most powerful technique experts use is building prompts in conceptual layers rather than dumping everything at once.
The Five Layers
- Layer 1: Role & Expertise - Who is the AI in this interaction? What knowledge should it access?
- Layer 2: Context & Constraints - What's the situation? What are the boundaries?
- Layer 3: Task Structure - How should the AI approach this? What's the thinking process?
- Layer 4: Output Specifications - What format? What length? What elements must be included?
- Layer 5: Quality Controls - How will we validate? What checks should happen?
Layered Prompting in Action
Scenario: You need a content strategy for a product launch.
❌ Intermediate Approach:
"Create a content strategy for launching our new project management
tool targeting remote teams."
✅ Expert Approach:
LAYER 1 - ROLE & EXPERTISE:
You are a content strategist who has launched 20+ B2B SaaS products.
Your specialty is creating content engines that generate qualified leads
with minimal paid spend. You think in terms of customer journey stages
and content formats that actually convert.
LAYER 2 - CONTEXT & CONSTRAINTS:
Product: Project management tool with unique async collaboration features
Target: Remote-first teams, 10-50 people, currently using Asana or Trello
Timeline: 8 weeks pre-launch, ongoing post-launch
Budget: $15K (must include content creation and promotion)
Team: 1 content writer, 1 designer, founder available for thought leadership
LAYER 3 - TASK STRUCTURE:
Approach this in phases:
Phase 1 - Strategic Foundation:
- Define our 3 core content pillars based on product differentiation
- Map content types to each funnel stage (awareness → consideration → decision)
- Identify our unique POV that will stand out
Phase 2 - Content Planning:
- Create 8-week pre-launch calendar
- Specify content format, channel, and goal for each piece
- Note dependencies and production timelines
Phase 3 - Distribution Strategy:
- Organic channels (which platforms, why, frequency)
- Paid promotion (where to allocate budget for maximum impact)
- Partnership/collaboration opportunities
LAYER 4 - OUTPUT SPECIFICATIONS:
Deliver as:
1. Strategic narrative (2-3 paragraphs explaining the approach)
2. Content calendar (table format: Week | Content Piece | Format |
Channel | Goal | Owner)
3. Budget allocation (breakdown of $15K)
4. Success metrics (3-5 KPIs we should track)
LAYER 5 - QUALITY CONTROLS:
After completing the strategy:
- Identify the 3 riskiest assumptions you've made
- Note what additional information would strengthen this plan
- Provide 2-3 alternative approaches we should consider
Let's begin with Phase 1.
Why this is powerful:
- The AI knows exactly what expertise to draw from
- Clear constraints prevent generic advice
- Phased approach enables complex reasoning
- Specific output format ensures usability
- Quality controls catch gaps and assumptions
When to Use Layered Prompting
✅ Use it for:
- Complex business problems
- Creative projects requiring strategy
- Technical documentation
- Any task where quality matters more than speed
❌ Don't use it for:
- Simple factual queries
- Quick edits or formatting
- Straightforward tasks with obvious approaches
Framework 2: Constitutional Prompting
Give the AI a "constitution"—a set of principles that govern how it should think and respond throughout your interaction.
The Power of Constitutions
A constitution creates consistency across multiple exchanges and embeds quality standards directly into the model's behavior.
Example Constitution:
OPERATING PRINCIPLES FOR THIS CONVERSATION:
1. Specificity over Generalization
- Provide concrete examples, not abstract concepts
- Use numbers, names, and specific scenarios
- Replace "often" with "in X% of cases" when possible
2. Reasoning Transparency
- Show your thinking process, not just conclusions
- Explain why you chose approach A over approach B
- Note when you're uncertain
3. Productive Disagreement
- Present counter-arguments to your own recommendations
- Identify when conventional wisdom might be wrong
- Challenge my assumptions if you spot flaws
4. Practical Orientation
- Every recommendation must be actionable
- Include what it costs (time, money, complexity)
- Flag what could go wrong
5. No Platitudes
- Ban: "think outside the box," "synergy," "leverage," "circle back"
- If it sounds like it came from a corporate memo, rewrite it
- Be direct and human
ACKNOWLEDGE THESE PRINCIPLES, THEN: Help me decide whether to
build or buy a customer support ticketing system.
The AI will follow these principles throughout the conversation, producing much higher quality responses with less hand-holding.
Constitutional Prompting Templates
For Analysis Work:
ANALYTICAL PRINCIPLES:
1. Data before opinions - cite sources and numbers
2. Acknowledge uncertainty - if confidence <80%, say so
3. Multiple perspectives - always include alternate interpretations
4. Falsifiability - state what evidence would prove you wrong
Now analyze: [your task]
For Creative Work:
CREATIVE PRINCIPLES:
1. Original over familiar - avoid the first idea that comes to mind
2. Specific over general - use concrete details and sensory language
3. Surprising over expected - look for the unusual angle
4. Purposeful over decorative - every element should serve the goal
Now create: [your task]
For Technical Work:
TECHNICAL PRINCIPLES:
1. Correctness over cleverness - working code beats elegant code
2. Explain trade-offs - note what you're optimizing for and against
3. Consider maintenance - flag what will be hard to change later
4. Security-conscious - point out potential vulnerabilities
Now build: [your task]
Practice Exercise
✍️ PRACTICE EXERCISE 2.1: Build Your Constitution
Scenario: You're working with an AI to develop your business strategy for the next year.
Your task: Write a 4-5 principle constitution that will ensure the AI gives you the kind of strategic thinking you need (not generic business advice).
Consider:
- What bad habits do generic business recommendations have?
- What qualities do you value in strategic advice?
- What should the AI prioritize or avoid?
Write your constitution, then test it with a real strategic question.
Framework 3: Chain-of-Thought Scaffolding
Don't just ask for reasoning—provide the exact scaffolding the model should use.
The Scaffolding Principle
When you provide a thinking structure, the model produces dramatically more thorough and logical outputs.
Without Scaffolding:
"Should we expand our product to serve enterprise customers? Explain your reasoning."
Result: You'll get some pros and cons, but the analysis will be shallow.
With Scaffolding:
Should we expand our product to serve enterprise customers?
Use this analytical framework:
STEP 1 - OPPORTUNITY ASSESSMENT:
- Market size and growth rate for enterprise segment
- Current competitors and their positioning
- Our unique advantages in this segment
STEP 2 - CAPABILITY ANALYSIS:
- What product capabilities do we have?
- What gaps exist for enterprise needs?
- Estimated development cost and time for each gap
STEP 3 - GO-TO-MARKET CHALLENGE:
- What does enterprise sales require (team, process, timeline)?
- Expected CAC and sales cycle length
- First-year revenue realistic forecast
STEP 4 - RISK ASSESSMENT:
- What could go wrong?
- What could distract from our core business?
- What's the opportunity cost?
STEP 5 - DECISION FRAMEWORK:
- Under what conditions is this a YES?
- Under what conditions is this a NO?
- What information would change your recommendation?
STEP 6 - RECOMMENDATION:
Based on the above analysis, provide a clear recommendation with:
- Your confidence level (1-10)
- The 3 most important factors in your decision
- Next steps if we proceed
- Alternative approaches we should consider
Why this works: Each step builds on the previous one, forcing comprehensive analysis rather than surface-level thinking.
Advanced Scaffolding Patterns
The Comparison Scaffold:
Compare [Option A] vs [Option B] using this structure:
FOR EACH OPTION:
1. Core strengths (top 3)
2. Critical weaknesses (top 3)
3. Best-case scenario
4. Worst-case scenario
5. Hidden costs or complications
THEN COMPARE:
6. Which is better for [specific criterion]?
7. Which has lower risk?
8. Which has higher upside?
9. What would make you choose one over the other?
10. Is there a hybrid approach that takes the best of both?
The Problem-Solving Scaffold:
Problem: [describe problem]
STEP 1 - PROBLEM DISSECTION:
Break this into sub-problems. What are the 3-4 core issues?
STEP 2 - ROOT CAUSE ANALYSIS:
For each sub-problem, what's causing it? Go at least 2 levels deep.
STEP 3 - SOLUTION GENERATION:
For each root cause, generate 2-3 potential solutions.
Rate each: Impact (1-10), Difficulty (1-10), Time to implement.
STEP 4 - DEPENDENCIES & SEQUENCING:
What needs to happen first? What depends on what?
STEP 5 - RECOMMENDATION:
Given the above, what's the optimal sequence of actions?
What resources are needed? What could block success?
The Research Scaffold:
Research topic: [topic]
PHASE 1 - LANDSCAPE MAPPING:
- What are the 5-7 key subtopics I need to understand?
- Who are the 3-5 leading experts or authorities?
- What are the major schools of thought or competing frameworks?
PHASE 2 - EVIDENCE GATHERING:
- What does the data show? (cite specific studies/sources)
- Where is there consensus?
- Where is there active debate or uncertainty?
PHASE 3 - CRITICAL EVALUATION:
- What's the quality of available evidence?
- What biases might exist in the research?
- What questions remain unanswered?
PHASE 4 - SYNTHESIS:
- What can we conclude with high confidence?
- What requires more investigation?
- What are the practical implications?
✍️ PRACTICE EXERCISE 2.2: Create Your Scaffold
Choose one of these scenarios:
- A) You need to decide whether to hire a senior developer or two junior developers for your team.
- B) You're evaluating whether to rebrand your company.
- C) You're deciding which of three marketing channels to focus on.
Your task: Create a complete analytical scaffold (6-8 steps) that would force thorough, logical analysis of your chosen scenario.
Then test it with an AI and see the difference in output quality.
Framework 4: Few-Shot Mastery
The most underrated expert technique: showing the AI exactly what good looks like through examples.
The Power of Examples
Few-shot learning means providing 2-4 examples of exactly what you want before asking the AI to generate something new.
Why it's powerful:
- Examples are more precise than descriptions
- The AI can pattern-match to your specific quality bar
- You control style, tone, format, and structure
- Works across all models
Few-Shot Structure
I need [type of output]. Here are examples of exactly what I want:
EXAMPLE 1:
[paste example]
EXAMPLE 2:
[paste example]
EXAMPLE 3:
[paste example]
KEY PATTERNS TO NOTICE:
- [element 1 that's important]
- [element 2 that's important]
- [element 3 that's important]
Now create one for: [your new case]
Few-Shot Examples in Practice
Scenario: Product Descriptions
I need a product description for our new wireless earbuds.
Here are 3 examples of our brand voice:
EXAMPLE 1 (Phone Case):
"Your phone survives your life. Military-grade protection meets
slim design—because bulk is not a security strategy. Drops, dings,
and the occasional coffee catastrophe? Handled. $24."
EXAMPLE 2 (Laptop Stand):
"Your neck deserves better than hunching. Aluminum stand that puts
your screen at eye level and your posture in the 21st century.
Folds flat, works anywhere, fixes your workspace. $49."
EXAMPLE 3 (USB-C Cable):
"The cable that stays plugged in. Reinforced connector survives
12,000+ bends (we tested). Charges fast, transfers faster, looks
better than it needs to. 6ft of frustration-free tech. $15."
PATTERNS TO MATCH:
- Opens with customer pain point or desire
- Technical benefits in plain language
- One personality line that's memorable
- Crisp, confident tone
- Ends with price
Now write one for: Wireless earbuds with active noise cancellation,
24-hour battery, $129.
The AI will now match your exact style and structure.
Scenario: Email Responses
I need to respond to customer complaints. Here's how we do it:
EXAMPLE 1:
"Hi Marcus—you're absolutely right, that's not the experience we
want you to have. I've refunded your account and added 3 months free.
Your patience while we sorted this out means a lot. —Jenna"
EXAMPLE 2:
"Sarah, I completely understand the frustration. We shipped a
replacement today (tracking: [link]) and you can keep or donate
the damaged one. If this one isn't perfect, I'll handle it personally.
—Jenna"
EXAMPLE 3:
"Tom—that's on us. We've updated your subscription and you won't be
charged for the months you couldn't access the app. I've also flagged
this with our engineering team so it doesn't happen to anyone else.
Thanks for giving us a chance to make it right. —Jenna"
STYLE NOTES:
- Use customer's name
- Acknowledge they're right to be upset
- State the solution immediately
- Personal sign-off
- Under 50 words
- Warm but professional
Customer complaint: "I was charged twice this month and can't get
through to support. This is ridiculous."
Write the response:
When Few-Shot Works Best
✅ Perfect for:
- Brand voice consistency
- Specific formatting requirements
- Matching a particular style or tone
- Creating variations of a proven format
- Training the AI on your company's approach
❌ Less useful for:
- Completely novel creative work
- When you don't have good examples
- Simple, straightforward tasks
✍️ PRACTICE EXERCISE 2.3: Few-Shot Training
Your task: Find 3 examples of something you need to create regularly (emails, social posts, reports, code snippets, etc.).
Create a few-shot prompt that includes:
- Your 3 examples
- Explicit pattern notes
- A new case to generate
Test it and compare the output quality to what you'd get without examples.
Framework 5: Constraint-Based Creativity
Counter-intuitive truth: More constraints = better creativity.
The Constraint Paradox
When you give an AI total freedom, it produces generic outputs. When you add smart constraints, it produces original work.
Why constraints work:
- They eliminate the obvious/generic options
- They force the model to search deeper in possibility space
- They create novel combinations
- They give you more control over the output
Types of Powerful Constraints
1. Elimination Constraints
Remove the obvious approaches to force originality.
Bad:
"Write about leadership"
Good:
"Write about leadership. Constraints:
- Cannot use the words: vision, empower, inspire, team, goal
- Cannot mention any CEOs or business leaders
- Must draw all examples from sports, arts, or parenting
- Must include one counter-intuitive claim"
2. Format Constraints
Force a specific structure that creates clarity.
"Explain quantum computing. Format:
- Exactly 4 paragraphs
- Paragraph 1: One-sentence definition
- Paragraph 2: Why it matters (one concrete example)
- Paragraph 3: Main challenge (why it's hard)
- Paragraph 4: Where we are today (one number/milestone)
- Total: under 150 words"
3. Perspective Constraints
Force an unusual angle.
"Explain the importance of user research. But:
- Write from the perspective of a user who's tired of being researched
- Make me sympathize with both sides
- End with something both researcher and user would agree on"
4. Combination Constraints
Force unexpected combinations.
"Create a marketing campaign for our accounting software. Requirements:
- Use the structure of a movie trailer
- Include exactly one moment of humor
- Reference a historical event
- End with a question, not a call-to-action"
The Constraint Formula
[Task] + [What to avoid] + [Format requirement] + [Unusual element]
= Original output
Example:
Write a job posting for a senior engineer.
Constraints:
- Avoid: "rockstar," "ninja," "passionate," "fast-paced"
- Format: 4 short paragraphs max, no bullet points
- Include: One surprising perk that isn't typical
- Include: One honest challenge of the role
- Tone: How you'd describe it to a friend, not how HR would write it
✍️ PRACTICE EXERCISE 2.4: Constrain for Quality
Take one of these generic tasks and add 4-5 constraints that will force better output:
- A) "Write a blog post about time management"
- B) "Create a social media post about our new product"
- C) "Explain blockchain technology"
Test your constrained version against the generic version. Notice the difference.
🎯 MODULE 2 CHECKPOINT
You've now learned five advanced frameworks:
- Layered Prompting - Build prompts in conceptual layers
- Constitutional Prompting - Set operating principles
- Chain-of-Thought Scaffolding - Provide thinking structures
- Few-Shot Mastery - Show exactly what you want
- Constraint-Based Creativity - Use limits to force originality
Before moving on, complete this integration exercise:
Complex Challenge: You need to create a competitive analysis comparing your product to three competitors.
Your task: Design a prompt that combines at least 3 of the frameworks above. Then execute it and evaluate the quality of output.
This is where the magic happens—when you start combining frameworks.
MODULE 3: MODEL-SPECIFIC MASTERY
Getting the Best from Each AI Tool
Not all AI models are created equal. Each has distinct strengths, weaknesses, and quirks. Experts know how to optimize for each one.
ChatGPT (GPT-4) Mastery
Model Characteristics
- Strengths: Creative tasks, conversational flow, broad knowledge, coding, accessibility
- Weaknesses: Can be verbose, sometimes overconfident, may require more constraint for precision
- Best for: Brainstorming, content creation, explanations, code generation, conversational interactions
GPT-4 Optimization Techniques
Technique 1: Custom Instructions as Environment Variables
ChatGPT Plus allows custom instructions. Experts use these like programming environment variables.
Example Custom Instructions:
WHAT WOULD YOU LIKE CHATGPT TO KNOW ABOUT YOU:
- I'm a B2B SaaS founder, technical background, 8 years experience
- My company: project management tool, 50 customers, $30K MRR
- I value: directness, speed, practical over theoretical
- I know: basic marketing, growth tactics, product development
- I need help with: strategy, copywriting, prioritization
HOW WOULD YOU LIKE CHATGPT TO RESPOND:
- Match my tone: if I'm casual, be casual; if formal, be formal
- Show reasoning in tags before complex answers
- When giving advice, always include: what could go wrong
- For code, always explain trade-offs between approaches
- Challenge my assumptions if you spot logical gaps
- Keep responses under 300 words unless I ask for comprehensive analysis
- Never use: "delve," "leverage," "synergy," "circle back"
This creates consistency across all your conversations.
Technique 2: Verbosity Control Through Language
GPT-4 can be wordy. Control this through your language:
For concise output:
"Explain X. Be ruthlessly concise. Every sentence must be essential."
"Give me the 3-bullet version."
"Pretend I'm in an elevator and you have 30 seconds."
For comprehensive output:
"Give me a deep dive on X. Be thorough."
"Walk me through this step-by-step, assuming I know nothing."
"I want the full picture, not a summary."
Technique 3: Multi-Turn Sculpting
Instead of trying to get perfect output in one prompt, use multiple rapid turns to sculpt exactly what you want:
Turn 1: "Draft a cold email for enterprise sales."
[Review output]
Turn 2: "Too salesy. Rewrite as a peer reaching out with something
genuinely valuable. One specific insight they'd care about."
[Review output]
Turn 3: "Better. Now cut it by 40% without losing impact."
[Review output]
Turn 4: "Perfect length. Replace the opening line with something that
shows I've done my homework about their company."
[Review output]
Total time: 5 minutes. Result: Excellent email.
Technique 4: Temperature Control Through Instructions
You can't directly control temperature, but you can simulate it:
For creative/exploratory work (high temperature simulation):
"Give me 10 ideas for [X]. Prioritize unusual over obvious.
I want at least 3 that make me think 'that's weird but interesting.'"
For precise/factual work (low temperature simulation):
"Explain [X]. Stick to facts. If you're uncertain about anything,
explicitly flag it as uncertain."
Technique 5: Code Optimization
For coding tasks, GPT-4 needs specific direction:
Basic request:
"Write a Python function to validate email addresses."
Expert request:
"Write a Python function to validate email addresses.
Requirements:
- Use regex for validation
- Handle edge cases: plus addressing, subdomains, international domains
- Return tuple: (is_valid, error_message)
- Include docstring with examples
- Type hints required
Then:
- Explain why you chose this regex pattern
- Note what valid emails this might reject
- Show 3 test cases (2 valid, 1 invalid)"
✍️ PRACTICE EXERCISE 3.1: GPT-4 Sculpting
Task: You need GPT-4 to write a product announcement.
Step 1: Write a basic prompt and get output
Step 2: Use 3-4 sculpting turns to improve it
Step 3: Document what each turn accomplished
Notice how much faster this is than trying to write the "perfect" first prompt.
Claude (Sonnet 4.5) Mastery
Model Characteristics
- Strengths: Long-context reasoning, nuanced analysis, following complex instructions, structured thinking, precision, coding
- Weaknesses: Can be formal, sometimes cautious
- Best for: Analysis, research synthesis, complex reasoning, detailed documentation, working with large documents
Claude Optimization Techniques
Technique 1: Extended Context Exploitation
Claude excels with large amounts of context. Feed it everything relevant:
I'm uploading 4 documents:
1. Our Q1-Q3 board presentations (63 pages)
2. Customer interview notes (24 interviews)
3. Product roadmap (15 pages)
4. Competitive analysis (8 companies)
Task: Identify gaps between what customers are asking for and
what's on our roadmap. Then recommend 3 features we should prioritize
in Q4.
Requirements:
- Reference specific customer quotes and board slides
- Note any contradictions between documents
- For each recommendation: customer evidence, competitive context,
estimated effort
- Flag assumptions you're making due to missing information
Claude will synthesize across all documents effectively.
Technique 2: Artifact-Driven Workflows
Claude can create "artifacts" (structured documents it can then edit). Use this for iterative work:
Create a Product Requirements Document for [feature] as an artifact.
Include:
- Problem statement
- User stories (3-5)
- Technical requirements
- Success metrics
- Open questions
Then, act as 3 different stakeholders and critique it:
1. Engineering Lead: Technical feasibility concerns
2. Customer Success: User adoption risks
3. Product Marketing: Positioning challenges
After each critique, update the PRD to address valid concerns.
This creates a living document that improves through iteration.
Technique 3: Meta-Cognitive Prompting
Claude responds exceptionally well to prompts about its own thinking:
Analyze [business problem].
Before providing your analysis:
1. What framework or mental model is most useful here?
Why did you choose it over alternatives?
2. What key information is missing that would significantly
change your analysis?
3. What's the strongest counter-argument to the conclusion
you're about to reach?
4. What assumptions are you making? Rate your confidence
in each (1-10).
Now provide your analysis.
This produces remarkably thoughtful outputs.
Technique 4: Structured Thinking Frameworks
Claude excels when given explicit analytical structures:
Evaluate whether we should expand into the European market.
Use this framework:
SECTION 1 - OPPORTUNITY QUANTIFICATION:
Market size, growth rate, competitive landscape
[Claude analyzes]
SECTION 2 - CAPABILITY ASSESSMENT:
What do we have? What do we need? What are the gaps?
[Claude analyzes]
SECTION 3 - RISK MAPPING:
Regulatory, operational, financial, competitive risks
Rate each: Likelihood (1-10), Impact (1-10)
[Claude analyzes]
SECTION 4 - SCENARIO MODELING:
Best case, base case, worst case (with specific assumptions)
[Claude analyzes]
SECTION 5 - DECISION FRAMEWORK:
Go/No-Go criteria based on the above
[Claude provides recommendation]
Technique 5: Long-Form Content Development
For reports, documentation, or comprehensive analysis:
I need a comprehensive guide on [topic] for [audience].
Structure:
- Executive summary (key takeaways)
- Main sections (5-7, you determine the best breakdown)
- Each section: explanation, examples, common mistakes, best practices
- Conclusion with action items
Approach:
1. First, outline the structure and get my approval
2. Then write section by section
3. After each section, I'll review and you'll refine
4. We'll iterate until we have the complete guide
Let's start with the proposed outline.
✍️ PRACTICE EXERCISE 3.2: Claude Analysis
Task: Give Claude a complex business scenario with multiple uploaded documents (or simulate by providing detailed context).
Use meta-cognitive prompting to force deeper analysis.
Compare the output quality to what you'd get from a simpler prompt.
Perplexity Mastery
Model Characteristics
- Strengths: Research, finding current information, source synthesis, citations
- Weaknesses: Shorter responses, less creative, more constrained
- Best for: Research, fact-finding, comparing sources, staying current
Perplexity Optimization Techniques
Technique 1: Research Query Structuring
Perplexity works best with well-structured research queries:
Weak:
"What's happening with AI regulation?"
Strong:
"Compare AI regulation approaches across US, EU, and China (2024-2025):
Focus on:
- Major legislation passed or proposed
- Definition of 'high-risk AI systems' in each jurisdiction
- Enforcement mechanisms and penalties
- Industry response and compliance challenges
Prioritize primary sources: government documents, official statements,
policy papers over news aggregation.
Provide specific examples of companies affected by each regulatory approach."
Technique 2: Source Quality Control
Tell Perplexity what types of sources you want:
Research [topic].
Source requirements:
- Prioritize: Academic papers, government reports, company filings,
technical documentation
- Avoid: News aggregators, opinion pieces, marketing content
- Date range: Last 6 months
- Include: Contradicting viewpoints if they exist
For each major claim, cite the specific source and publication date.
Technique 3: Iterative Research Refinement
Start broad, then narrow based on results:
Turn 1: "Overview of quantum computing commercial applications 2024-2025"
[Review results]
Turn 2: "Focus on the top 3 nearest-term viable applications.
Which companies are leading in each?"
[Review results]
Turn 3: "For quantum optimization (the #1 application), what are
the specific technical blockers and estimated timelines to commercial
viability? Include expert predictions."
[Review results]
Turn 4: "Find case studies or pilot programs currently testing
quantum optimization. What results have they published?"
Technique 4: Comparative Research
Use Perplexity for side-by-side comparisons:
"Compare [Technology A] vs [Technology B] for [use case]:
Create a comparison across:
1. Technical maturity (with evidence/milestones)
2. Current adoption rate (specific numbers)
3. Cost considerations (range of pricing)
4. Key limitations (technical, not marketing speak)
5. Expert predictions for 2025-2026
For each point, cite specific sources and note if experts disagree."
Technique 5: Deep Dive Research Mode
For comprehensive research projects:
I need a comprehensive research report on [topic].
Phase 1: Landscape
- Identify the 5-7 key subtopics I need to understand
- Who are the 3-5 leading experts/authorities?
- What are the major frameworks or schools of thought?
Phase 2: Evidence
- For each subtopic, what does recent research show? (cite studies)
- Where is there consensus vs. active debate?
- What are the quality/limitations of available evidence?
Phase 3: Current State
- What's happening now? (last 3-6 months)
- What are the emerging trends?
- What predictions are experts making?
Phase 4: Synthesis
- What can we conclude with high confidence?
- What requires more investigation?
- What are practical implications?
Let's start with Phase 1.
✍️ PRACTICE EXERCISE 3.3: Research Challenge
Choose a topic you need to research for work or a project.
Task: Use Perplexity with:
- A well-structured initial query
- Source quality requirements
- At least 2 rounds of iterative refinement
Document the insights you gained vs. what a simple Google search would give you.
🎯 MODULE 3 CHECKPOINT
You now understand how to optimize for:
- ChatGPT (GPT-4) - Creative, conversational, needs sculpting
- Claude - Analytical, handles complexity, loves structure
- Perplexity - Research-focused, needs specific queries
- Image & code generation tools
Integration Exercise:
Scenario: You're launching a new product and need:
- Market research (which tool?)
- Positioning strategy (which tool?)
- Marketing copy (which tool?)
- Visual assets (which tool?)
Your task: Map each task to the optimal tool and write the expert-level prompt for each one.
MODULE 4: TROUBLESHOOTING LIKE AN EXPERT
Diagnosing and Fixing Common Problems
Even expert prompts sometimes produce suboptimal outputs. The difference is that experts know how to diagnose and fix problems quickly.
The Expert Troubleshooting Framework
- IDENTIFY THE GAP - What specifically is wrong with the output?
- DIAGNOSE THE ROOT CAUSE - Why did the model produce this?
- APPLY THE FIX - What prompt modification addresses the root cause?
- VALIDATE - Did it work? If not, iterate.
Let's apply this to common problems.
Problem 1: Generic / Bland Output
The Symptom
The AI gives you advice that sounds like it came from a corporate handbook or generic blog post.
Example:
Prompt: "How can I improve team productivity?"
Output: "To improve team productivity:
1. Set clear goals and expectations
2. Foster open communication
3. Provide the right tools and resources
4. Recognize and reward good work
5. Encourage work-life balance"
This is useless—it's what everyone already knows.
The Diagnosis
Root cause: The model is pattern-matching to generic training data. It's giving you the "average" response to this type of question.
The Fix
Add specificity and constraints:
I need to improve team productivity, but here's the specific situation:
- Team: 7 engineers, 2 designers
- Current state: Shipping features, but missing deadlines by 30%
- Constraints: Already working 50+ hours/week, morale is decent
- Tools: Using Jira, Slack, GitHub
- Recent changes: Switched to 2-week sprints 3 months ago
Constraints on your advice:
- No generic productivity tips
- Focus on root cause, not symptoms
- Consider that time/effort is maxed out
- Suggest what to STOP doing, not just what to add
- Include how to validate if changes are working
What's actually going on, and what should I do about it?
The Result
Now you get specific analysis of potential bottlenecks (unclear requirements? too much context-switching? wrong sprint length?) and actionable, non-obvious advice.
Problem 2: Factual Errors / Hallucinations
The Symptom
The AI states something confidently that's incorrect or makes up details.
The Diagnosis
Root cause: Language models predict plausible-sounding text, not truth. They can't inherently distinguish fact from fiction.
The Fix
Strategy 1: Verification Loops
Research [topic].
After your response:
1. List every factual claim you made
2. Rate your confidence in each (1-10)
3. For claims with confidence <8, mark as [VERIFY]
4. Suggest how I could independently verify [VERIFY] claims
This makes the model explicit about uncertainty.
Strategy 2: Request Sources
"Research [topic]. For every factual claim, cite your source
with format: [claim](source URL). If you're not certain about
something, explicitly say 'I'm uncertain about [X] because [reason]'
rather than guessing."
Strategy 3: Cross-Validation
For critical information:
- Get response from AI
- Ask: "What about this answer might be wrong?"
- Verify key claims independently
- Feed corrections back if needed
Problem 3: Not Following Complex Instructions
The Symptom
You write detailed instructions, but the AI ignores parts of them or gets confused.
Example:
You asked for: Analysis with 5 sections, each including examples and data
You got: 3 sections, no examples, generic statements
The Diagnosis
Root causes:
- Instructions too long/complex (working memory overload)
- Ambiguous instructions
- Conflicting requirements
- Model lost track of requirements mid-generation
The Fix
Strategy 1: Chunking + Confirmation
We're going to complete a task in 3 steps. First, let me describe all steps,
then you'll confirm your understanding. Then we execute one at a time.
STEP 1: Analyze the data and identify 5 key trends
STEP 2: For each trend, find 2 supporting examples
STEP 3: Create a summary table with trends, examples, and implications
Please confirm: What are you going to do in each step?
[Wait for confirmation]
Good. Now let's start with Step 1.
Strategy 2: Checklist Method
Create a market analysis report.
Requirements checklist (confirm you'll include all of these):
- [ ] Executive summary (2-3 paragraphs)
- [ ] Market size with specific numbers
- [ ] 3-5 key trends with evidence
- [ ] Competitive landscape (at least 4 companies)
- [ ] SWOT analysis in table format
- [ ] 3 strategic recommendations
- [ ] Sources cited throughout
Confirm the checklist, then begin.
Strategy 3: Simplify + Iterate
Instead of: "Do A, B, C, D, E all in one response"
Try: "Let's start with A. Once that's done, we'll move to B."
Problem 4: Wrong Tone or Style
The Symptom
The output is too formal when you want casual, or too casual when you want professional.
The Diagnosis
Root cause: Without explicit guidance, models default to a "safe" professional tone.
The Fix
Strategy 1: Tone Specification
Write an email to my colleague Jake about the project delay.
Tone requirements:
- Write like you're talking to a friend, not a formal business contact
- Use "I" and "you," not "we" or "one"
- Contractions are fine (we're, didn't, etc.)
- Be direct—no corporate jargon
- Conversational but not unprofessional
Think: how you'd explain this over coffee, not in a quarterly report.
Strategy 2: Provide a Reference
Write [content].
Here's an example of the tone I want (for a different topic):
"Hey Sarah—quick update on the design mockups. I pushed them to your
review stack, but heads up: the mobile version still needs work. I'm
thinking we might need to simplify the nav. Thoughts?"
Match this: casual, direct, brief, human.
Strategy 3: Iterative Tone Adjustment
Turn 1: [Get initial output]
Turn 2: "Too formal. Rewrite like you're texting a colleague, not writing a memo."
Turn 3: "Better, but still a bit stiff. Make it sound more like how people actually talk."
Problem 5: Too Long or Too Short
The Symptom
You get a wall of text when you wanted a summary, or bullet points when you needed depth.
The Diagnosis
Root cause: No explicit length guidance.
The Fix
Strategy 1: Explicit Length Constraints
Explain [topic].
Length: Exactly 3 paragraphs, approximately 150 words total.
Structure:
- Para 1: Core concept (2-3 sentences)
- Para 2: Why it matters (2-3 sentences)
- Para 3: Common misconception (2-3 sentences)
Strategy 2: Compression/Expansion Rounds
Turn 1: "Explain [topic]"
[Review output]
Turn 2: "Cut this by 60% without losing the key insights."
OR
Turn 2: "Expand this with 3 concrete examples and more detail."
Strategy 3: Reference Length
"Write this in about the same length as the following example:
[paste example of desired length]"
Problem 6: Lacks Specific Examples
The Symptom
All theory, no practical examples.
The Diagnosis
Root cause: Default to abstract explanations unless specifically prompted for concrete examples.
The Fix
Strategy 1: Explicit Example Requirements
Explain [concept].
Requirements:
- For every principle, provide a concrete example
- Examples must include specific numbers, names, or scenarios
- No abstract examples (bad: "like a company might do X")
- Good examples: "like how Slack uses X to achieve Y result"
Minimum 3 examples total.
Strategy 2: Example-First Structure
Explain [concept].
Structure:
1. Start with a specific, concrete example
2. Extract the general principle from the example
3. Show 2-3 variations of how this applies
4. End with a counter-example (when this doesn't work)
Problem 7: Doesn't Challenge Your Thinking
The Symptom
The AI just agrees with everything you say or gives you what you asked for without pushback.
The Diagnosis
Root cause: Models are trained to be helpful and agreeable. They won't naturally play devil's advocate.
The Fix
Strategy 1: Explicitly Request Challenge
Here's my plan: [describe plan]
Your job:
1. Identify the 3 weakest parts of this plan
2. Present the strongest counter-argument you can
3. What am I probably not considering?
4. Play devil's advocate—why might this fail?
Be direct. I want critique, not validation.
Strategy 2: Red Team Approach
I'm proposing [decision/strategy].
Act as three skeptics:
1. The pessimist: What's wrong with this?
2. The competitor: How would they exploit weaknesses?
3. The experienced advisor: What mistake am I repeating?
Each provides their perspective, then I'll revise based on valid concerns.
Strategy 3: Build It Into Your Constitution
Operating principle for this conversation:
- Challenge my assumptions when you spot logical gaps
- Point out when I'm oversimplifying
- Present counter-arguments before conclusions
- Flag risks I might be underestimating
Now let's discuss: [topic]
The Expert Troubleshooting Checklist
When output quality is poor, ask:
Content Issues:
- ☐ Is it too generic? → Add specific context and constraints
- ☐ Is it wrong? → Add verification loops and source requirements
- ☐ Does it lack examples? → Require concrete, specific examples
- ☐ Is it too abstract? → Force practical application
Structure Issues:
- ☐ Is it ignoring instructions? → Chunk the task, use confirmation
- ☐ Is it wrong length? → Set explicit length constraints
- ☐ Is it poorly organized? → Provide a clear structure to follow
Style Issues:
- ☐ Is the tone wrong? → Specify tone explicitly with examples
- ☐ Is it too agreeable? → Request challenge and critique
- ☐ Is it too cautious? → Reframe the request to avoid triggers
✍️ PRACTICE EXERCISE 4.1: Diagnostic Practice
Here are 3 problematic outputs. For each:
- Diagnose the root cause
- Write a fix
- Test it
Problem A:
You asked: "How should I price my SaaS product?"
You got: "Consider value-based pricing, competitive pricing, and cost-plus pricing. Research your market and test different price points."
Problem B:
You asked for a 150-word explanation of machine learning
You got: 450 words of dense technical content
Problem C:
You asked: "Critique my go-to-market strategy"
You got: "This is a solid strategy with clear market focus and good channel selection. The timeline seems reasonable and the budget allocation makes sense."
Write the improved prompts for each, then test them.
🎯 MODULE 4 CHECKPOINT
You now have systematic approaches to:
- Diagnose why outputs are poor
- Apply specific fixes for common problems
- Iterate efficiently to get quality outputs
- Build quality controls into your prompts
Integration Exercise:
Take a prompt you use regularly that doesn't always produce great results. Apply the troubleshooting framework:
- What specifically goes wrong?
- What's the root cause?
- What fix would address it?
- Test and validate
Document your before/after results.
MODULE 6: PRODUCTION WORKFLOWS
How Experts Actually Work Day-to-Day
Theory is one thing. Daily practice is another. Let's look at real workflows experts use to consistently produce high-quality outputs.
Workflow 1: The Rapid Iteration Protocol
Goal: Get from idea to polished output in 15-20 minutes
The Process
MINUTE 0-2: Quick Baseline
- Write minimal prompt
- Get initial output
- Identify obvious gaps
MINUTE 2-5: Constraint Layer
- Add 3-4 specific constraints
- Regenerate
- Evaluate improvement
MINUTE 5-10: Alternative Exploration
- Request 2-3 different approaches
- Compare strengths/weaknesses
- Choose best or synthesize
MINUTE 10-15: Refinement
- Sculpt the output with specific edits
- "Make X shorter"
- "Change Y tone"
- "Add Z element"
MINUTE 15-20: Polish
- Final quality check
- One more refinement pass
- Done
Example: Writing Product Copy
Minute 0-2 (Baseline):
"Write product description for noise-canceling headphones"
[Get generic output]
Minute 2-5 (Add constraints):
"Rewrite:
- Focus on the specific problem: open office noise
- Include one surprising benefit
- Keep under 50 words
- End with emotional outcome, not product features"
[Get better output]
Minute 5-10 (Explore alternatives):
"Give me 2 completely different approaches to this description:
1. Lead with a question
2. Lead with a bold claim"
[Compare options]
Minute 10-15 (Refine best option):
"Take version 2. Make the claim more specific and add one concrete detail
that makes it believable"
[Get refined version]
Minute 15-20 (Polish):
"Perfect. Now tighten the language—every word should earn its place"
[Final version]
Total time: 18 minutes
Result: Professional-quality copy
Workflow 2: The Research & Synthesis Protocol
Goal: Deep research on complex topic → actionable insights
The Process
PHASE 1: Landscape (10 min)
- Use Perplexity to map the territory
- Identify key sources, experts, frameworks
- Note what's consensus vs. debated
PHASE 2: Deep Dives (20-30 min)
- Use web_fetch to read primary sources
- Use Claude to analyze and synthesize
- Extract key insights and evidence
PHASE 3: Synthesis (15 min)
- Use ChatGPT or Claude to create structured output
- Focus on actionable insights
- Include confidence levels and caveats
PHASE 4: Validation (10 min)
- Cross-check key claims
- Identify gaps or uncertainties
- Add source citations
Example: Researching Market Entry Strategy
PHASE 1 - Landscape (Perplexity):
"What are the key considerations for a B2B SaaS company entering
the European market from the US? Focus on:
- Regulatory requirements (GDPR, etc.)
- Payment processing
- Sales/support localization
- Common mistakes American companies make
Prioritize recent sources (2023-2024) and primary sources."
[Review results, identify 3-4 key areas to go deeper]
PHASE 2 - Deep Dives (Claude with web_fetch):
"I'm researching GDPR compliance for a B2B SaaS. I've found these 3 articles:
[URLs from Perplexity]
Fetch each and synthesize:
1. What are the actual requirements (not just buzzwords)?
2. What does compliance cost (time & money)?
3. What's the enforcement reality (not just theoretical risk)?
4. What do companies commonly get wrong?"
[Repeat for other key areas]
PHASE 3 - Synthesis (Claude):
"Based on all this research, create a decision framework for US SaaS
companies evaluating European expansion. Include:
- Key readiness indicators
- Estimated costs and timeline
- Major risk factors
- Go/no-go criteria"
PHASE 4 - Validation:
"Identify any claims in this framework where:
- Evidence is thin
- Sources might be biased
- Conclusions might be premature
Rate confidence 1-10 for each major recommendation."
Total time: ~60 minutes
Result: Comprehensive, validated research
Workflow 3: The Creative Development Protocol
Goal: Original creative work (articles, campaigns, content)
The Process
PHASE 1: Constraint Definition (5 min)
- What's the goal?
- Who's the audience?
- What's off-limits?
- What must be included?
PHASE 2: Divergent Generation (10 min)
- Generate many options
- Prioritize variety over quality
- Explore unusual angles
PHASE 3: Evaluation & Selection (5 min)
- Compare options
- Identify strongest elements
- Choose direction or synthesize
PHASE 4: Development (15 min)
- Develop chosen direction
- Add detail and polish
- Maintain originality
PHASE 5: Refinement (10 min)
- Adversarial critique
- Address weaknesses
- Final polish
Example: Creating Blog Post
PHASE 1 - Constraints:
"I need a blog post about remote work productivity.
Constraints:
- Audience: Engineering managers at startups
- Must NOT include: generic tips, app recommendations, morning routines
- Must include: counter-intuitive insight, specific example, actionable framework
- Tone: Peer-to-peer, not guru-to-follower
- Length: 800-1000 words
Acknowledged?"
PHASE 2 - Divergent:
"Generate 10 different angles for this post. Mix:
- 3 counter-intuitive claims
- 3 specific frameworks or mental models
- 4 unusual perspectives
For each, one-sentence description."
PHASE 3 - Selection:
"I like #3 (the idea about async-first being about trust, not timezone)
and #7 (the framework for deciding what should be sync vs async).
How could we combine these into one cohesive post?"
PHASE 4 - Development:
"Write the full post combining those angles. Remember all constraints
from Phase 1."
PHASE 5 - Refinement:
"Act as an engineering manager reading this. What's weak or unconvincing?
[Review critique]
Revise to address the valid criticisms."
Total time: 45 minutes
Result: Original, thoughtful content
Workflow 4: The Code Development Protocol
Goal: Produce working, quality code efficiently
The Process
PHASE 1: Specification (5 min)
- What should it do?
- What are edge cases?
- What are non-functional requirements?
PHASE 2: Generation (10 min)
- Generate initial code
- Request explanation of approach
- Understand trade-offs
PHASE 3: Refinement (10 min)
- Request improvements for specific concerns
- Add error handling
- Improve readability
PHASE 4: Validation (5 min)
- Request test cases
- Think through edge cases
- Document usage
Example: Creating API Integration
PHASE 1 - Specification:
"I need a Python function to interact with Stripe's API to create
a subscription.
Requirements:
- Handle authentication
- Create customer if doesn't exist
- Subscribe to specified plan
- Return subscription object or error
- Handle common failure cases (payment fails, invalid plan, etc.)
- Type hints required
- Proper error handling
What approach would you recommend?"
PHASE 2 - Generation:
[AI explains approach]
"Sounds good. Write the code."
[Review code]
PHASE 3 - Refinement:
"Good start. Improvements:
1. Add retry logic for network failures
2. Make the logging more useful
3. Extract the API key handling to be more testable
Update the code."
PHASE 4 - Validation:
"Provide:
1. Example usage
2. Test cases covering: success, customer exists, payment fails,
network error, invalid plan
3. Brief documentation on how to use this"
Total time: 30 minutes
Result: Production-ready code with tests and documentation
Building Your Template Library
Experts don't start from scratch every time. They build reusable templates.
Template Structure
### [TEMPLATE NAME]
**Use for:** [when to use this]
**Prompt:**
[The actual prompt with placeholders]
**Variables:**
{VARIABLE1}: [description]
{VARIABLE2}: [description]
**Tips:**
- [tip 1]
- [tip 2]
**Example:**
[filled-in example]
Starter Template Library
Research Template:
### Deep Research
**Use for:** Comprehensive research on complex topics
**Prompt:**
Research {TOPIC} with depth level: {SURFACE/STANDARD/DEEP}
Requirements:
- Focus on {PRIMARY_QUESTION}
- Source types: {ACADEMIC/INDUSTRY/NEWS/MIXED}
- Date range: {DATE_RANGE}
- Perspective: {NEUTRAL/CRITICAL/SUPPORTIVE}
Deliverables:
1. Key findings (3-5 with evidence)
2. Consensus vs. debate (what's agreed, what's contested)
3. Confidence assessment (rate each finding 1-10)
4. Gaps in knowledge (what's unclear or needs more research)
**Variables:**
{TOPIC}: The research subject
{PRIMARY_QUESTION}: The specific question you need answered
{DATE_RANGE}: Time period for sources
**Tips:**
- Be specific about what you actually need to know
- Request confidence levels to identify uncertain areas
- Use Perplexity for initial research, Claude for synthesis
Expert Habits: The Daily Practice
Habit 1: Start with Quick Tests
Never write the "perfect" prompt first. Always:
- Write minimal prompt (30 seconds)
- See what you get (baseline)
- Then optimize
Time saved: 70% (most "perfect" prompts waste time over-engineering)
Habit 2: Keep a Failure Log
When something doesn't work:
- Document what you tried
- Note why it failed
- Record what fixed it
Example entry:
Date: 2024-03-15
Task: Product comparison
Failed approach: "Compare X vs Y"
Why it failed: Too generic, got surface-level comparison
What worked: "Compare X vs Y for [specific use case]. Focus on:
[specific criteria]. Include: [specific data points]."
Lesson: Specificity in criteria matters more than task description
Habit 3: Cross-Model Validation
For important outputs:
- Generate with primary model
- Ask different model to critique
- Use critique to improve
Example:
Generate with ChatGPT → Critique with Claude → Refine with ChatGPT
Or:
Research with Perplexity → Synthesize with Claude → Validate with ChatGPT
Habit 4: Version Your Prompts
Track what changes improved results:
Prompt v1: "Explain machine learning"
Result: Generic textbook explanation
Rating: 3/10
Prompt v2: "Explain machine learning using only examples from cooking"
Result: Creative but imprecise
Rating: 5/10
Prompt v3: "Explain machine learning. Use cooking analogy for core
concept, then 2 real business applications"
Result: Clear, practical, memorable
Rating: 8/10
Keep your best versions for reuse.
Habit 5: Build Chains for Complexity
When stuck on complex task:
- Break into steps
- Execute one step at a time
- Review between steps
Never try to do everything in one prompt if task is complex.
MODULE 7: LIVE CHALLENGES
Putting Everything Together
Time to apply everything you've learned. These challenges simulate real-world scenarios experts face.
Challenge 1: The Vague Executive Request
The Scenario
Your CEO says: "Our website copy isn't working. Make it more engaging."
That's all you get. No specifics. No examples. No metrics.
Your Task
Phase 1: Requirements Extraction (10 min)
Design prompts that extract:
- What "not working" actually means
- What "engaging" means in this context
- Success criteria
- Constraints
Phase 2: Strategy Development (15 min)
Based on extracted requirements:
- Propose 3 different approaches
- Show examples of each
- Explain trade-offs
Phase 3: Execution (20 min)
- Execute best approach
- Create revised copy
- Provide A/B testing plan
Expert Approach
PHASE 1 - Extract Requirements:
"I need to improve website copy, but first I need to understand the current
situation. Act as a website optimization consultant and ask me the 10 most
important questions you'd need answered to give good recommendations.
Organize by priority: Must-know vs. Nice-to-know"
[Use AI's questions to think through requirements]
Then provide context: [Your actual situation]
PHASE 2 - Strategy:
"Based on this context, propose 3 different copy approaches:
Approach 1: [Focus on specific angle]
Approach 2: [Different angle]
Approach 3: [Third angle]
For each:
- Core strategy in 1-2 sentences
- Example of how homepage headline would change
- Pros and cons
- Best for [what situation]"
PHASE 3 - Execute:
"Let's go with Approach [X]. Rewrite:
- Homepage headline and subheading
- Main value proposition section
- CTA text
Requirements:
[Include all your constraints from phase 1]
After writing, explain what makes this more engaging than the original."
✍️ YOUR TURN: Challenge 1
Execute this challenge with a real or simulated scenario from your work.
Document:
- How you extracted requirements
- What strategy you chose and why
- Your final deliverables
- Time taken
Challenge 2: The Research Rabbit Hole
The Scenario
Your sales team says: "Our enterprise sales cycles are taking 9 months. Industry standard is 6 months. Figure out what we're doing wrong."
You need to:
- Research enterprise sales best practices
- Analyze common bottlenecks
- Identify what might be wrong specifically
- Recommend solutions
Time limit: 60 minutes
Expert Approach
STEP 1 - Industry Research (Perplexity, 15 min):
"Research B2B enterprise sales cycles in SaaS (2023-2024):
Focus on:
- Average length by company size and deal size
- Common bottlenecks that extend cycles
- What distinguishes 6-month from 9-month cycles
- Recent best practices or changes
Prioritize: Research reports, sales expert insights, company case studies"
STEP 2 - Diagnostic Framework (Claude, 15 min):
[Paste research findings]
"Based on this research, create a diagnostic framework to identify
why our sales cycles are longer than industry standard.
Structure as:
1. Common bottleneck categories (5-7)
2. For each: how to identify if this is our issue
3. Diagnostic questions to ask our sales team
4. Data we should analyze"
✍️ YOUR TURN: Challenge 2
Pick a complex business problem you're actually facing.
Execute the research → analysis → recommendations workflow.
Time yourself. Evaluate the quality of your output.
Challenge 3: The Crisis Response
The Scenario
It's 9 AM. Your phone is buzzing.
A major customer just tweeted: "We've been unable to access [your product] for 3 hours. No response from support. This is unacceptable for enterprise software. Evaluating alternatives."
They have 50,000 followers. Other customers are starting to reply with their own complaints.
You need to:
- Craft an immediate public response
- Draft a direct message to the customer
- Create an internal communication for your team
- Outline a broader crisis communication plan
Time limit: 20 minutes (because crises don't wait)
Expert Approach
IMMEDIATE (First 5 minutes):
Prompt to ChatGPT:
"URGENT crisis response needed.
Situation: Major customer posted public complaint about 3-hour outage.
Escalating on social media.
I need 3 immediate responses (write all 3 simultaneously):
1. PUBLIC TWEET REPLY (60 characters max):
- Acknowledge immediately
- Show we're on it
- Move to DM
2. DIRECT MESSAGE to customer:
- Apologize sincerely (but no excuses)
- Explain what happened (brief, honest)
- What we're doing to fix
- What we're doing to prevent
- Direct contact offer
3. INTERNAL TEAM MESSAGE:
- Situation summary
- Action items for team
- Communication protocol during resolution
Tone: Urgent but calm, human not corporate, taking responsibility.
WRITE ALL THREE NOW."
Why This Approach Works:
- Separates urgent from important - Handles immediate need first
- Uses AI for speed - Gets responses in seconds, not minutes
- Requests multiple outputs simultaneously - More efficient than sequential
- Maintains human judgment - Review before posting, but quickly
✍️ YOUR TURN: Challenge 3
Simulate a crisis in your domain. Set a timer for 20 minutes.
Execute:
- Immediate response (5 min)
- Broader communication plan (15 min)
Evaluate: Was it fast enough? Was quality maintained? What would you improve?
🎯 MODULE 7 CHECKPOINT
You've now tackled real-world challenges:
- Vague requirements → Extract and execute
- Research projects → Deep analysis and recommendations
- Crisis response → Fast, high-quality under pressure
Final Integration:
Identify the 3 most common challenging tasks in your work.
For each:
- Map which techniques from the course apply
- Design a complete workflow
- Create a template for reuse
- Execute once to validate
- Document lessons learned
MODULE 8: THE EXPERT'S MINDSET
How to Think Like a Master Prompter
We've covered techniques, tools, and workflows. Now let's talk about the mindset that makes everything work.
Principle 1: You Are Programming, Not Asking
Shift your mental model:
❌ Don't think: "I'm asking the AI a question"
✅ Do think: "I'm writing code that generates the output I need"
Every prompt is a program:
- Variables: The context you provide
- Functions: The structures you create
- Conditionals: The constraints you set
- Loops: The iterations you run
This shift makes you more systematic and less frustrated when things don't work immediately.
Principle 2: Embrace Productive Failure
Most people:
- Write prompt
- Get poor output
- Get frustrated
- Give up or lower standards
Experts:
- Write prompt
- Get poor output
- Diagnose what specifically is wrong
- Apply targeted fix
- Iterate until excellent
Every failure is data. Use it.
Principle 3: Model-Appropriate Expectations
Don't expect:
- Perfect factual accuracy without verification
- Creative brilliance on first try
- Reading your mind
- Perfect voice matching without examples
- Complex reasoning without scaffolding
Do expect:
- Excellent outputs with proper prompting
- Rapid iteration
- Creative combinations
- Pattern matching
- Tireless refinement
Principle 4: Time Investment Philosophy
The expert's time calculation:
Spending 20 minutes on prompt design that saves 3 hours of work = EXCELLENT investment
Spending 45 minutes trying to get perfect prompt instead of good-enough in 10 minutes = POOR investment
When to invest in perfection:
- Reusable templates you'll use 20+ times
- High-stakes outputs
- Learning exercises
When "good enough" is better:
- One-off tasks
- Drafts you'll manually revise anyway
- Exploratory work
Principle 5: The Collaboration Mindset
You + AI is better than You or AI alone
AI strengths:
- Processing large amounts of information
- Generating many options quickly
- Recognizing patterns
- Consistent formatting
- Tireless iteration
Your strengths:
- Judgment and discernment
- Understanding context and nuance
- Knowing what "good" looks like
- Strategic thinking
- Emotional intelligence
The expert approach:
Use AI for: Generation, research, analysis, formatting, variation
Use You for: Direction, evaluation, refinement, judgment, synthesis
The Expert Self-Assessment
You're thinking like an expert when you:
- ☐ See prompts as programs you're constructing
- ☐ Diagnose failures systematically instead of getting frustrated
- ☐ Know which tool to use for which task
- ☐ Iterate quickly rather than seeking perfection immediately
- ☐ Combine your judgment with AI's capabilities
- ☐ Stay current with new models and techniques
- ☐ Use AI ethically and responsibly
- ☐ Make your prompts simpler as you get better
Your Development Path Forward
Month 1: Master the basics
- Use all frameworks from Module 2
- Build your first template library (10 templates)
- Track what works and what doesn't
Month 2: Develop workflows
- Create standard workflows for common tasks
- Optimize for speed without sacrificing quality
- Start using prompt chains regularly
Month 3: Model expertise
- Become expert in your primary tool
- Develop model-specific optimization techniques
- Cross-validate important work across models
Month 4: Advanced techniques
- Master persona synthesis and adversarial prompting
- Build complex prompt chains
- Create your own advanced techniques
Month 5: Scale and systematize
- Document all your learnings
- Build comprehensive template library
- Train others on your techniques
Month 6 and beyond: Stay cutting-edge
- Follow model releases and updates
- Experiment with new capabilities
- Contribute to the prompting community
🎯 FINAL COURSE CHECKPOINT
You've completed the Expert Prompting Masterclass. You now understand:
Fundamentals:
- The expert mindset (programming, not asking)
- System 2 thinking activation
- Context engineering principles
Frameworks:
- Layered prompting
- Constitutional prompting
- Chain-of-thought scaffolding
- Few-shot mastery
- Constraint-based creativity
Model Mastery:
- ChatGPT optimization
- Claude optimization
- Perplexity optimization
Advanced Techniques:
- Persona synthesis
- Adversarial prompting
- Constraint stacking
- Prompt chaining
- Meta-prompting
Congratulations, Expert! 🎓
You've completed the Expert Prompting Masterclass.
You now possess:
- A comprehensive framework for approaching any prompting challenge
- Specific techniques that work across all major AI models
- Workflows you can use immediately
- The mindset to continue improving
What separates you from intermediate users:
- You understand WHY techniques work, not just WHAT to do
- You diagnose problems systematically
- You choose tools strategically
- You iterate intelligently
- You continuously improve
Your next steps:
- Practice daily - Use these techniques in real work
- Build your library - Document what works for you
- Share your learnings - Teach others
- Stay current - Follow AI developments
- Keep experimenting - The field evolves rapidly
Remember: Expert prompting is not about knowing every trick. It's about understanding the underlying principles and applying them systematically.
You're now equipped to get dramatically better outputs from any AI tool, solve complex problems with AI assistance, and work faster without sacrificing quality.
Welcome to expert-level prompting. Now go create something remarkable.