Build Your Deep Research Assistant
What does the video cover?
- What Deep Research is: delegating a research assignment to AI instead of asking it a question.
- When to use it: strategic decisions, synthesizing complex evidence and learning unfamiliar topics quickly.
- What to keep in mind: a citation does not make a claim correct, so you still check the key sources.
The exercise: run your industry intelligence brief
Most of what you need as a leader is a clearer picture of where your industry is heading and who is already moving. This prompt turns Deep Research into a structured brief tailored to your role and goals.
- Open a new chat in ChatGPT, Copilot (web), Copilot (work), Claude or Gemini.
- Turn on Deep Research. In ChatGPT, choose Deep Research from the tools menu. In Copilot, use the Researcher agent. In Claude, use the Research button. In Gemini, choose Deep Research.
- Paste the prompt below with your details filled in and run it.
- Read Part 3 first, because it connects the research to your next steps. Then pick one insight and decide whether it changes anything you are doing this quarter.
The prompt template
Character: You are the world's most respected analyst of industry transformation through AI. Objective: I'm [role] at [company] in [industry]. I want to understand where my industry is heading as generative AI matures, and who is already preparing well. Do: - Draw on respected analyst reports, industry research, and serious forecasting for the industry outlook - For trailblazers, source only from trade press, serious journalism, academic research, or executive interviews - Prioritize organizations that centered people in their approach: workforce change, new ways of working, cultural preparation Don't: - No vendor content, sponsored research, or listicles - No outcome-only stories: I need to understand how, not just what Output: Part 1: Where the industry is heading. What respected analysts forecast for generative AI's impact on this industry over the next 3–5 years. Include key shifts, the biggest risks, and where value will move. Flag where forecasters meaningfully disagree. Part 2: Who is leading the charge. Three to five organizations already preparing for that future. Per case: what they are doing, how they are handling people, source. One closing paragraph on what these trailblazers have in common. Part 3: How I start. Referencing my [90-day goal], tell me where I would start to build toward this future for myself and/or my organization, ideally starting with [3–5 things I'd hope AI can take off my plate].
How do you make this prompt yours?
The template works as it is, but it does not know you yet. Five changes turn a generic result into one you trust.
- Be precise about your industry. "Mid-size logistics in Europe" gets a far sharper brief than "logistics."
- Make your 90-day goal measurable. For example: "reduce manual demand forecasting by 40%" instead of "use more AI."
- List your real time drains. Name the three to five tasks you would hand to AI first, such as weekly capacity reports or competitor monitoring.
- Narrow the sources if you need to. Add the analysts, trade publications or regions you trust most, and the ones you want excluded.
- Ask for a second pass. When the report lands, ask: "Which of these claims rests on a single source? Check them again."
Comparing tools? Our guide to AI deep research tools shows how ChatGPT, Copilot, Claude and Gemini differ.
Personalizing a prompt by hand works, but it takes a few rounds. In AI Leader Advanced, LeadOS interviews you before the course starts and builds a personalized library of exercises and prompts around your role, your workflows and your goals. Your first brief already knows your industry, your goal for the quarter and what you want off your plate.
See AI Leader Advanced