Start by Giving the AI a Role
The most common mistake professionals make is asking an AI tool a question without any context. Instead of a generic query, start your prompt by assigning a persona. For example, don't just ask it to 'write a summary'. Instruct it to 'Act as a senior
project manager summarising the key decisions from a client meeting for an executive'. This simple step frames the task, helping the AI adopt the correct tone, vocabulary, and level of detail for the intended audience, which dramatically improves the quality of the first draft.
Provide Clear Context and Examples
AI models know nothing about your specific project, company jargon, or preferred formats unless you provide that information. For repetitive tasks, give the AI the necessary background. This could be a transcript of a meeting, raw notes, or a previous report. The rule is simple: the better the input, the better the output. For truly consistent results, provide an example of what a good final document looks like. This is called 'few-shot prompting' and it trains the AI to mimic a specific structure and style, reducing the need for heavy editing.
Master the Meeting Summary
Meetings are a primary source of time-consuming documentation. A powerful prompt can turn a messy transcript into a structured, actionable summary in seconds. Use a prompt that asks for specific sections. For instance: 'Here are the raw notes from our weekly check-in. Create a summary with these sections: 1. Key Decisions Made, 2. Action Items (with owner and deadline), and 3. Open Questions for Next Meeting'. This transforms a passive record into a tool for accountability and follow-through.
Create Reusable Prompt Templates
For tasks you perform daily or weekly, such as status updates or standard emails, don't reinvent the prompt every time. Create a reusable template with placeholders for the variable information. For example: 'Write a weekly status update for the [Project Name] team. Include these sections: Highlights, Blockers, and Next Steps. Here are the bullet points for this week: [paste your points]'. Store these templates in a shared document for your team to ensure consistency and save time across the board.
Use AI to Draft, Not to Finalise
Think of the AI's first output as a mediocre first draft, not a finished product. The goal is to eliminate the 'blank page' problem and get a structured document you can refine. Use follow-up prompts to iterate. You can ask it to 'make the tone more formal', 'cut this down by half', or 'rewrite this section to be more direct'. Always review and edit the final output. AI can make factual errors or sound unnaturally confident, so a human check for accuracy, tone, and context is essential before sharing.
Ask It to Ask You Questions
For more complex or ambiguous documentation, one of the most powerful but underused strategies is to have the AI help you build a better prompt. Start by saying: 'I need to create a project proposal. Before you draft it, ask me five questions to better understand what is required'. This forces the AI to seek out the critical context it needs, such as the target audience, key objectives, and specific constraints. It turns a one-way command into a collaborative process, leading to a much more relevant and useful final document.
