Automation With Authenticity
The goal of AI email automation is not to remove the human from the equation, but to handle the repetitive 60-70% of messages that follow predictable patterns. This includes routine queries, scheduling confirmations, and simple follow-ups. By automating
these, you free up mental energy for complex, high-value communication. The key is to make the automated responses feel just as personal and authentic as if you typed them yourself. A generic, AI-generated draft is easily spotted, especially by people who know you well. The aim is to create a system where the AI acts as a true assistant, adopting your unique style to maintain consistency and trust.
Step 1: Define Your Personal Voice
Before you can teach an AI your voice, you must first understand it yourself. This process involves analysing your own writing to create a "style blueprint." Start by gathering 5-10 examples of emails you've written that best represent your authentic self. Look for patterns in your greetings and sign-offs, your typical sentence length, your use of emojis or exclamation points, and any common phrases or idioms you favour. Ask yourself: Are you formal or conversational? Direct or diplomatic? Detailed or concise? You can even use AI tools like ChatGPT to analyse your samples and provide a summary of your tone, style, and structure. This analysis forms the foundation for training your AI.
Step 2: Choose the Right AI Tool
The AI email landscape is crowded, but the right tool for you depends on your specific needs. Look for features that specifically allow for voice customisation. Many modern platforms have moved beyond simple templates. Tools like Superhuman, now part of Grammarly, allow AI assistants to draft emails in your voice. Others, like Microsoft Copilot in Outlook and Gemini in Gmail, are increasingly integrated into the email clients themselves, learning from your behaviour. Key features to look for include the ability to use 'custom instructions' or upload 'knowledge' files, which are essentially documents that teach the AI about you and your style. Some tools can even research a recipient on platforms like LinkedIn before drafting an email to ensure the tone is appropriate.
Step 3: Train and Prompt Your AI Assistant
Training is an active process. The most effective method is to provide the AI with your style blueprint and writing samples. In many tools, this is done by pasting your analysis into a 'custom instructions' field. This tells the AI to always respond in a manner consistent with your defined voice. When you need a specific email drafted, your prompt is crucial. Don't just say, "Write a follow-up email." Instead, provide context. For example: "Using my defined style, draft a polite but firm follow-up to John about the overdue project report. Mention that the new deadline is this Friday." This combination of a pre-trained voice and a specific, contextual prompt yields the best results.
Step 4: Supervise, Edit, and Refine
AI is a powerful assistant, but it should not work unsupervised. Always review and edit AI-generated drafts before sending them. This is your chance to add genuine opinions, personal anecdotes, or strategic imperfections that make the writing feel human. AI tends to be neutral, so injecting your actual perspective is critical. Think of the AI as generating the first 80% of the draft. Your job is to add the final 20% of nuance, personality, and strategic insight. This human-in-the-loop approach ensures quality and prevents the AI from sending something off-brand or inappropriate. Over time, as you correct and refine the AI's drafts, it will get better at mimicking your style, but the final sign-off should always be yours.
