Prompting Master Class Wizard

Work through the full master class in a guided path that keeps the content complete and easier to consume.

Jump to course

1. Foundations

Understand the real problem, core principles, and prompt rules.

2. Prompt Improvement

See weak prompts, add structure, and strengthen business context.

3. Output Design

Design the real deliverable, refine drafts, and avoid assumptions.

4. Projects

Build a repeatable project and living instruction system.

5. Quality & Growth

Review output, manage risks, and keep the system improving.

Business Context

Most people start using AI by asking one quick question and hoping for a good answer. That usually produces something generic because the AI does not know the real goal, the audience, or the business context.

This master class teaches prompting in a practical way. It walks through one real use case from the first weak prompt all the way to a reusable project with living instructions, standardized language, and quality controls.

The starting scenario is simple enough for anyone to understand: a customer needs an update about a delay. Then the scenario grows into a dealership service communication workflow where tone, consistency, clarity, and standard language matter.

Core Principles

AI is not intelligence.
AI is not truth.
AI is an amplifier.
Better prompting starts with better thinking.
Do not let AI guess when information is missing.
When something is unclear, ask for clarification first.

Core Prompting Rules

Do not make assumptions when important information is missing.
Ask questions when the request is unclear or more detail is needed.
Identify missing facts before asking AI to produce a final answer.
State the audience, tone, and format whenever they matter.
Treat the first output as a draft that can be improved.
Review the result before sharing it with a customer, coworker, or leader.

Section 1: The Real-World Problem

A customer is expecting their unit to be ready. The repair is delayed. The employee needs to send an update quickly, but the wording matters because poor communication can damage trust.

Exercise

Write the first prompt you would naturally give AI for this situation.

What to Notice

Most first prompts are vague. That is normal. The goal is not to be perfect on the first try. The goal is to improve the request step by step.

Section 2: What Happens with a Weak Prompt

Here is the kind of prompt many people start with:

Write a message to a customer about a repair delay.

Exercise

List what is missing from that request before asking AI to answer it.

How long is the delay?
Why is the delay happening?
Is this an email, text, or phone script?
What tone should be used?
What should the customer know or not be told?

What to Notice

AI did not fail here. The request was incomplete. When the request is incomplete, the output is at risk.

Section 3: Use a Simple Prompt Formula

Start with a simple structure:

Task + Audience + Tone + Format

Exercise

Rewrite the weak prompt using that formula.

Example

Write a short email to a customer explaining their repair will take one extra day. Keep it polite, professional, and easy to understand.

What to Notice

This is already much better because the AI now knows:

what to write
who it is for
how it should sound
what format to use

Clear structure usually improves output fast.

Section 4: Add Business Context

Now we move from a universal example into a dealership-specific one.

We are a powersports dealership. A customer's repair is delayed one day because a part did not arrive on time. The customer expected pickup tomorrow. Write a clear, polite email that maintains trust, avoids technical jargon, and does not sound defensive.

Exercise

Add one more useful detail to the prompt, such as whether the team is offering another update, rescheduling pickup, or apologizing for the inconvenience.

What to Notice

Context is what turns AI from generic to useful. The more relevant business reality you provide, the more usable the output becomes.

Section 5: Ask for the Output You Actually Need

Many users ask for "a message" when they really need several formats for different channels.

Exercise

Expand the request so AI creates all of the pieces needed for the workflow.

Create a subject line, a customer email, a short SMS version, and three talking points for a service advisor phone call. Keep everything polite, clear, and consistent.

What to Notice

Output design matters. You are not just asking AI to write. You are asking it to support a real business process.

When you ask for the right structure, the output becomes easier to use immediately.

Section 6: Refine the First Draft

The first answer is usually a draft. Strong users improve it with focused follow-up prompts.

Exercise

Take the first result and improve it with one or more of these follow-up prompts.

Make this warmer and more empathetic. Reduce the formality. Shorten the email to five sentences. Add one reassuring sentence that shows we are prioritizing the repair. Remove any wording that sounds defensive or vague.

What to Notice

You do not need to start over every time. Refinement is one of the most important prompting habits to build.

Section 7: Do Not Let AI Guess

This is one of the most important rules in the whole master class.

If the request is missing information, AI should not silently fill the gap with assumptions when accuracy matters.

Exercise

Look at this prompt and decide what AI should ask before writing the final answer.

Write a message to a customer about their delayed unit.

Example Clarifying Questions

What is causing the delay?
How long is the delay expected to be?
Is this email, SMS, or both?
Should the message sound formal, friendly, or highly apologetic?
Are there any details that should not be shared?

What to Notice

If AI is guessing, the output is at risk. If AI is asking, the output improves.

Section 8: Expand the Scenario

Now the learner sees why one good prompt is not enough for repeated work. In a dealership, the same communication pattern may happen under several different conditions.

one-day part delay
backordered part with no clear ETA
technician unexpectedly out
additional repair discovered during diagnosis
warranty approval delay

Exercise

Pick two of these scenarios and compare how the wording should stay consistent and how it should change.

What to Notice

This is where one-off prompting begins to break down and where projects become useful.

Benefits of Projects and Instructions

They save time by reducing repeated setup.
They improve consistency across similar tasks.
They preserve useful business context.
They help standardize language and definitions.
They make it easier to solve repeat issues faster.

What Belongs in the Instructions

who the audience is
what tone should be used
what good output looks like
what to avoid
approved or preferred language
important business rules
core prompting rules such as not making assumptions

Step-by-Step: Create the Project

Create a new project. Give it a clear name such as Service Customer Communication or Dealership Service Messaging.
Write the project purpose. State what the project is for and the type of work it should help produce.
Define the audience. For this use case, the audience is customers receiving service updates.
Define the tone and style. Clear, polite, calm, professional, and easy to understand.
Add the core rules. Do not make assumptions. Ask for clarification when something is unclear or more information is needed. Avoid jargon. Protect customer trust.
Add approved language and things to avoid. This helps standardize output and reduce weak wording.
Test the project with a real scenario. Run a service delay example and review the output.
Refine the instructions. If the output is off, update the instruction set instead of only fixing that one answer.

Exercise: Draft the First Instruction Set

You are helping create customer service communications for a powersports dealership service department. Write in a tone that is polite, clear, calm, professional, and easy to understand. Avoid technical jargon and avoid language that sounds defensive or blames others. Do not make assumptions when important information is missing. If the request is unclear or more detail is needed, ask clarifying questions before producing a final response. Protect customer trust and focus on clear next steps. Use approved, consistent language whenever possible.

Exercise

Have the learner edit this draft by adding one dealership-specific rule, one approved phrase, and one phrase to avoid.

What to Notice

A project is not magic. It is simply a stronger working environment for AI.

Maintaining a Living Instruction Set

Instructions should change when the team learns something important. This is what makes the system stronger over time.

Add a rule when the same mistake repeats.
Add a definition when language needs to be standardized.
Add an example when a format should stay consistent.
Add a business note when AI keeps missing the same real-world context.
Add a phrase to avoid when it creates confusion, blame, or weak communication.

Approved Language

Create three approved phrases for service delay communication.

We wanted to keep you updated on the current repair timeline.
We are actively working to complete your repair as soon as possible.
We will update you again as soon as we have the next confirmed step.

Phrases to Avoid

Create three phrases that should be avoided.

We have no idea when it will be done.
It is not our fault.
You will just have to wait.

Exercise: Standard Definitions

Build standard definitions so outputs stay consistent.

Term	Simple Definition	Why It Matters
Service delay	A repair timeline change that affects the expected completion date.	Keeps the wording consistent across emails, texts, and scripts.
Backordered part	A needed part that is not currently available from the supplier.	Helps avoid vague or misleading explanations.
Next update	The next confirmed communication point the customer can expect.	Helps prevent open-ended promises.

Quality Check Before Sharing

Did the AI use the correct audience and tone?
Did it invent or assume any facts that were not provided?
Did it follow the project rules and approved language?
Is the next step clear for the customer?
Should anything learned from this output be added to the instructions?

Risks and Mitigation

Risk	What It Looks Like	Impact
Assumption risk	AI fills in missing details instead of asking for clarification.	Wrong or misleading messages get created.
Inconsistent language	Different prompts produce different wording for the same situation.	Customer experience becomes uneven.
Knowledge loss	Lessons learned stay in one person's head instead of going into the instructions.	The team repeats mistakes and rework continues.
Blind trust	The AI output sounds polished so nobody checks it.	Bad information gets shared with confidence.

Mitigation

Require clarification when important information is missing.
Store approved language and definitions in the project.
Update the instructions after repeat issues or lessons learned.
Review every customer-facing output before sending.

Role Optimization

What AI Does Creates drafts, rewrites language, formats outputs, applies stored standards, and helps scale repeatable communication work faster.

What the Human Does Provides judgment, confirms facts, decides what matters, reviews tone, and updates the instruction system based on real experience.

Best Fit Repeatable communication, SOP writing, training materials, summaries, first drafts, and structured work that benefits from consistency.

Do Not Expect Perfect truth, perfect judgment, or safe communication when key facts are missing and nobody reviews the output.

Core Principle: The goal is not just to write better prompts. The goal is to build a better system where AI helps with repeatable work, humans provide judgment, and business knowledge gets captured so performance improves over time.

Return to Course Overview