An effective email assistant requires four components:
Communication style analysis: Understanding how you write
Writing instructions: A style guide the AI can follow
Automation flow: A system that applies the style guide to incoming emails and drafts responses
Continuous improvement: A process to feed your edits back to the style guide so the assistant's drafts improve the more you use it
The assistant handles routine drafting while you maintain control through review and approval of every email before sending.
Analyze your communication style
Before an AI can write like you, it needs to understand how you communicate. This analysis should capture:
Tone markers: Warm, formal, direct, enthusiastic
Structural patterns: How you open and close emails, paragraph length, use of bullet points
Language preferences: Phrases you use frequently, level of formality
Personalization habits: How you address recipients, references to previous conversations
How to gather this data:
Have your AI assistant of choice review your most recent sent emails
If your AI tool has email integration, you can ask it to fetch and analyze your recent emails directly
Example prompt for style analysis:
Analyze my email communication style. Use whichever of my integrations that make sense.
Use an artifact if you think that's helpful. If using an artifact, consider what kind of artifact (interactive, visual, checklist, etc.) might be most helpful to this specific task.
The output should be a summary you can review and validate. If the AI misses something important about your style, provide feedback before proceeding.
Create writing instructions
Transform the style analysis into actionable instructions the AI can follow when drafting emails.
Example prompt for creating instructions:
Based on this communication style analysis, create writing instructions for an AI email assistant. The instructions should:
- Be concise (under 500 words)
- Include specific examples
- Focus on reproducible patterns
Bonus: Maintain a single source of truth
Keep the master copy of your writing instructions in a project folder or version-controlled repository. Copy to other locations (Google Workspace Studio, etc.) only when ready to deploy updates.
Set up continuous improvement
Your email assistant improves over time through a feedback loop:
Collect examples: When you modify an AI draft before sending, save both versions
Identify patterns: After collecting 3-5 examples, analyze what you consistently change
Update instructions: Refine the writing instructions based on the patterns
Deploy updates: Push the improved instructions to your flow
This iterative process trains your assistant to match your preferences more closely with each cycle.