Start With the Real Situation
The quality of an AI-written email depends on the context you give it. Before asking for a draft, write down who will read the email, what you need from them, the tone you want, and any constraints such as deadlines or attachments.
A useful prompt includes the recipient, the relationship, the goal, the facts that must be included, and the desired length. This keeps the answer focused and reduces the bland language that makes AI drafts feel mass-produced.
Use a Two-Step Drafting Process
First ask for a plain draft. Then ask for a revision with a specific improvement: warmer tone, shorter sentences, stronger call to action, or more executive style. Treat the first answer as raw material, not the final email.
For important messages, ask the model to list possible misunderstandings before finalizing the draft. This is especially useful for pricing changes, client updates, project delays, and sensitive internal feedback.
Keep Your Voice
Replace generic phrases with details only you would know. A strong email usually has one clear ask, one reason the ask matters, and one next step. AI can help with structure, but credibility comes from your judgment.
Never paste confidential client data, trade secrets, private employee information, or sensitive financial details into a public AI tool unless your organization has approved that workflow.