Lesson Content

Weak prompts usually fail because they ask for an answer without giving the AI a job, audience, context, or definition of success. A business owner needs repeatable prompts that produce usable work, not lucky one-off answers.

Plain-language view: a prompt is a work order. If you hand a person a sticky note that says "fix this," they have to guess. If you hand them a work order that says who the customer is, what happened, what result you need, what rules apply, and what to check before sending, they can do useful work. AI is similar. It can help, but it needs a clear work order.

Use this base structure:

Role: Act as a practical business assistant for a small business owner.

Goal: Help me complete [business task].

Context: Use only this background: [facts, audience, product, deadline, constraints].

Output: Return [email, checklist, plan, table, script, brief, SOP].

Review: List assumptions, risks, missing facts, and what I should verify before using it.

This structure works across current AI assistants because it does not depend on a secret model feature. It tells the assistant how to work and gives the business owner a way to inspect the result.

Think of the six parts like this:

Role means "what hat should the AI wear?"

Goal means "what job should it finish?"

Context means "what facts should it use?"

Constraints mean "what rules, limits, tone, deadline, policy, or budget apply?"

Output means "what should the answer look like?"

Review means "what could be wrong, missing, risky, or worth checking?"