Free mini-course · Lesson 1 / 6
How to write an effective prompt
A prompt is the request you give an AI system. Prompt engineering means designing, testing and improving those requests to get results that fit the task.
Editorial team: Take Two ·
The idea
Before opening a chat, finish this sentence: “I need this result, for this person, and it will be useful if…” An observable goal helps more than “do your best”. You do not need to code; you need to describe the work precisely.
A vague request
Write a professional email.
Why it is incomplete
The recipient, purpose and desired outcome are missing. The model will have to make choices for you and may add details you did not provide.
A prompt to try
Try the prompt in your usual AI tool. Select and copy the text; no coding is needed. Access, features and limits depend on the tool and plan you use.
Draft an email to an existing business client.
Goal: ask them to confirm our meeting on 15 October at 10:00.
Context: we will review the migration plan. The meeting will last 30 minutes.
Tone: polite and direct.
Format: subject and body, no more than 100 words.
Do not add locations, links or commitments I have not provided.
End with an explicit request for confirmation.What to look for
The draft must include the meeting date, time, duration and purpose; ask for confirmation; and stay within the word limit. It must not invent a video call link.
Now try it
Your exercise
Try the vague request and the complete request in separate new conversations. Compare both answers against the same criteria. Then change only the recipient: the email is now for a colleague and should use a more informal tone.
Check the result
- Does it include all four meeting details?
- Does it ask for confirmation without adding information?
- Does it follow the requested tone and word limit?
See the commented answer
An acceptable subject is “Meeting confirmation: 15 October, 10:00”. The body must mention reviewing the migration plan and the 30-minute duration. Many wordings work: assess the requirements, not an exact sentence.
Examples and commented answers are teaching material written for this course. They are not recorded model outputs; answers may vary.
Sources and further reading
The exercises are original Take Two teaching examples. For technical background on instructions, examples and evaluation criteria, see: