Examples

Examples are one of the fastest ways to make a prompt better. Instead of hoping the AI reads your mind, you show it what “good” looks like. It is less magic, more show-and-tell, which is usually a better strategy than vague optimism.

What this term means

In prompt engineering, examples are sample inputs, sample outputs, or both, included in a prompt to guide the AI’s response. They act like a pattern the AI can follow.

If you want a certain tone, format, level of detail, or writing style, examples make that visible. You are not just saying, “Write it professionally.” You are giving the model a concrete reference point.

Examples can be:

• A sample answer
• A sample format
• A before-and-after rewrite
• A short style reference
• A few labeled input-output pairs

This is why examples appear in prompt frameworks like MEOW, TREAT, and POWER. In each one, the idea is the same: give the AI a target to imitate so the result is closer to what you actually want.

Why it improves a prompt

Examples improve prompts because they reduce ambiguity.

Without examples, the AI has to guess things like:

• How formal the response should be
• How long it should be
• What structure to use
• What counts as a “good” answer
• What kind of wording you prefer

With examples, you make those choices visible. That usually leads to output that is:

• More consistent
• Closer to your preferred style
• Easier to review or reuse
• Less likely to wander off into weird territory

Think of it like giving directions. “Head somewhere nice for dinner” is not very helpful. “Go somewhere like that little Italian place with the simple menu and giant portions” is much clearer. One invites chaos. The other gets you lasagna.

In prompt engineering, examples are especially useful when you care about tone, formatting, summaries, categorization, rewriting, customer support replies, and any task where “close enough” is not actually close enough.

Similar words you might see

You may see examples described with similar terms like:

• Samples
• References
• Model answer
• Style example

These are all pointing to the same basic idea: show the AI a pattern worth copying.

3 sample prompts

Sample prompt 1

Rewrite the following email so it sounds professional but friendly.

Use this style example:
“Hi Jordan, just a quick update, the report is ready and attached. Let me know if you would like a shorter summary for the meeting. Thanks.”

Rewrite this email:
“Hey, I finished the thing. Sending it now. Tell me if you need more.”

This prompt uses a style example to show the tone the user wants. Instead of just asking for “professional but friendly,” it gives a clear sample of that tone in action.

Using the example improves the prompt because the AI does not have to guess what “professional but friendly” means. That phrase can mean ten different things depending on the person, the workplace, and how many passive-aggressive emails they have survived this week.

Sample prompt 2

Categorize each customer message as Billing, Technical Support, or General Question.

Examples:

• “I was charged twice this month” → Billing
• “The app crashes when I open settings” → Technical Support
• “Do you offer team discounts?” → General Question

Now categorize these messages:

1. “My invoice amount looks wrong.”
2. “I cannot log in after the update.”
3. “Can I switch plans later?”

This prompt uses samples as labeled input-output pairs. The AI sees the pattern, the categories, and how to apply them before doing the real task.

Using the examples improves the prompt because they define the categories more clearly than a long explanation usually would. In many cases, one good sample is worth a paragraph of instructions, which is good news for everyone who does not enjoy writing policy manuals for robots.

Sample prompt 3

Write a product description for a reusable water bottle.

Reference model answer:
“A compact desk lamp with adjustable brightness and a clean matte finish. Designed for small workspaces, it provides focused light without cluttering your desk.”

Follow the same structure: 2 sentences, clear language, practical benefits, no hype.

This prompt uses a reference or model answer to guide structure and writing style. The example shows concise wording, sentence count, and the kind of practical tone the user wants.

Using the example improves the prompt because it sets boundaries. The AI is more likely to produce a focused, usable product description instead of a dramatic sales pitch that sounds like it was written for a late-night infomercial.

Quick summary

Examples show the AI what good looks like, which is often far more effective than explaining it badly and hoping for the best.