🎯 Give Clear Instructions — Stop Being Vague with AI

What Is It?

Give Clear Instructions is the most fundamental prompt engineering technique — and the one most people skip.

Think of it like ordering food. “Give me food” gets you something random. “I’d like a medium-rare ribeye steak with mashed potatoes and grilled asparagus” gets exactly what you want. AI works the same way.

💡 The #1 rule: If your prompt could mean 10 different things, the AI will pick one randomly. Be specific enough that there’s only one reasonable interpretation.

When to Use It

• ✅ Every single time you use AI (this is the foundation)
• ✅ When you’re getting vague or unhelpful responses
• ✅ When you need a specific format, length, or style
• ✅ When the task has particular requirements

Before & After

❌ Before (Vague)

Write me something about emails

What’s wrong: “Something” could be an article, a template, a poem, tips, history… the AI has to guess.

✅ After (Clear)

Draft a professional follow-up email to a client who hasn’t responded to our project proposal in 2 weeks. Keep it polite but include a soft deadline. Maximum 150 words.

What’s better: Specific task (draft), specific type (follow-up email), specific context (client, proposal, 2 weeks), specific tone (polite), specific constraint (150 words).

Platform Tips

Microsoft 365 Copilot

• In Word, be specific about the document type: “Write a project status update” not “Write something”
• In Outlook, Copilot works best with clear instructions like “Reply saying I’ll attend but can’t stay past 3pm”
• In Teams, be specific: “Summarise action items from this meeting” not “What happened?”

ChatGPT

• ChatGPT handles longer, more detailed instructions well — don’t hold back on specifics
• Use the Custom Instructions feature to set baseline context so individual prompts can focus on the task

Claude

• Claude excels at following detailed instructions — the more specific, the better the output
• Use XML-style tags to structure complex instructions: <task>, <context>, <format>

Gemini

• Gemini works well with numbered instructions for multi-part tasks
• Be explicit about format — Gemini sometimes defaults to bullet lists

Real Examples from the Prompt Library

These prompts from our Prompt Library demonstrate clear instructions:

1. Professional Email Reply — Notice how it specifies the type of reply, tone, and constraints
2. Meeting Summary — Clear about what to include (decisions, action items, owners)
3. Data Analysis — Specific about the type of analysis and output format

Related Techniques

After mastering clear instructions, level up with:

• 🎭 Set a Role — Tell AI who to be for even better results
• 📋 Add Context — Give AI the background it needs
• 📐 Define the Format — Control the shape of the output