Skip to main content

Chapter 1: Clarity and Specificity

🎯 Learning Objectives​

After completing this chapter, you will be able to:

  • Identify and eliminate ambiguity in prompts
  • Write specific, actionable instructions
  • Use concrete language instead of vague terms
  • Apply the "5 W's" framework for completeness

📚 Introduction​

The difference between a prompt that works and one that doesn't often comes down to clarity. Vague prompts lead to vague responses. Specific prompts lead to useful outputs.

The Problem with Vague Prompts​

Consider this prompt:

Write something about technology.

Problems:

  • What kind of technology?
  • What format (article, list, poem)?
  • How long?
  • For what audience?
  • What tone?

The AI has to guess all of these, leading to unpredictable results.

The Clarity Framework​

The 5 W's + H​

Apply journalism's classic framework to every prompt:

QuestionWhat It AddressesExample
WhatThe task/topic"Write a blog post about AI in healthcare"
WhoAudience/persona"For hospital administrators"
WhereContext/platform"For our LinkedIn company page"
WhenTime reference"Covering 2024 developments"
WhyPurpose/goal"To establish thought leadership"
HowFormat/method"In 500 words with 3 subheadings"

Before and After Examples​

Vague:

Tell me about marketing.

Clear:

Explain 5 digital marketing strategies that small retail
businesses (under $1M revenue) can implement within a
month on a budget of $500 or less. For each strategy,
include expected ROI and one specific action step.

Vague:

Write code for a website.

Clear:

Write a Python Flask application that:
1. Displays a homepage with a contact form
2. Validates email format on the form
3. Stores submissions in a SQLite database
4. Shows a success message after submission

Include comments explaining each section.

Eliminating Ambiguity​

Common Ambiguous Words to Avoid​

Vague TermBetter Alternative
"Good""Meeting criteria X, Y, Z"
"Short""Under 100 words"
"Detailed""Including X, Y, Z components"
"Professional""Formal tone, no contractions, industry terminology"
"Quick""Within 2 paragraphs"
"Better""More efficient in terms of time/memory"
"Simple""Understandable by someone with no technical background"
"Soon""Within the next 24 hours"

Quantify When Possible​

❌ "Give me a few examples"
✓ "Give me exactly 5 examples"

❌ "Write a short paragraph"
✓ "Write a paragraph of 50-75 words"

❌ "List the main points"
✓ "List the 3 most important points"

❌ "Make it detailed"
✓ "Include statistics, quotes, and specific case studies"

Concrete vs. Abstract Instructions​

The Abstraction Ladder​

MOST ABSTRACT (avoid)
↑
"Write well"
|
"Write professionally"
|
"Write in business style"
|
"Write formally with no contractions"
|
"Use active voice, formal vocabulary,
no contractions, industry terminology"
↓
MOST CONCRETE (prefer)

Examples​

Abstract:

Make it sound smart.

Concrete:

Use technical vocabulary appropriate for software engineers.
Include specific metrics and data points.
Reference industry standards (e.g., ISO, IEEE).
Cite recent research where applicable.

Abstract:

Write persuasively.

Concrete:

Use the AIDA framework:
- Attention: Start with a surprising statistic
- Interest: Present the problem vividly
- Desire: Show benefits with social proof
- Action: End with a clear call-to-action

Specificity Techniques​

1. Define Scope Explicitly​

❌ "Explain machine learning"

✓ "Explain supervised machine learning, specifically:
- Definition in 2 sentences
- 3 common algorithms with one-sentence descriptions
- One real-world application example

Do NOT cover: unsupervised learning, deep learning,
or neural networks."

2. Specify What to Include AND Exclude​

"Write about the benefits of electric vehicles.

INCLUDE:
- Environmental benefits
- Cost savings
- Performance advantages

EXCLUDE:
- Detailed technical specifications
- Charging infrastructure challenges
- Brand comparisons"

3. Use Precise Numbers​

❌ "Write some bullet points"
✓ "Write exactly 5 bullet points, each 15-20 words"

❌ "Give examples"
✓ "Provide 3 examples: one from healthcare, one from
finance, one from education"

4. Define Success Criteria​

"Write a product description that:
✓ Is between 150-200 words
✓ Mentions the product name at least 3 times
✓ Includes 4 key features
✓ Has a clear call-to-action
✓ Uses no more than one exclamation mark"

Testing for Clarity​

Ask yourself these questions:

CLARITY CHECKLIST

☐ Could another person interpret this prompt differently?
☐ Am I using any vague words (good, nice, better, some)?
☐ Have I specified the format I want?
☐ Have I specified the length?
☐ Is the audience clear?
☐ Would I know if the output is "correct"?
☐ Can the AI complete this without asking questions?

Practical Exercise​

Exercise: Clarity Improvement​

Transform these vague prompts into clear, specific ones:

Prompt 1 (Vague):

Write about dogs.

Prompt 2 (Vague):

Help me with my presentation.
View Solutions

Prompt 1 (Clear):

Write a 200-word informational paragraph about Golden
Retrievers for a family considering dog adoption. Include:
- Temperament and personality traits
- Exercise requirements (hours per day)
- Suitability for households with children under 10
- One potential challenge for first-time owners

Tone: Friendly and helpful, avoid technical breeding terminology.

Prompt 2 (Clear):

I'm preparing a 10-minute presentation for my company's
quarterly sales meeting. The audience is 30 sales
representatives.

Help me by:
1. Suggesting 5 engaging opening lines related to Q3 performance
2. Creating an outline with 4 main sections
3. Recommending 3 data visualizations that would support
key points
4. Writing a memorable closing statement

Topic: We exceeded our Q3 sales target by 15%, and I want
to celebrate the team's success while motivating them for Q4.

Key Takeaways​

Summary
  1. Vague prompts = vague responses; clarity drives quality
  2. Use the 5 W's + H framework for completeness
  3. Quantify everything possible (length, number, time)
  4. Replace abstract words with concrete descriptions
  5. Define both what to include AND exclude
  6. Test with the clarity checklist before sending

← Back to Module 3 | Next: Context and Constraints →