How to Use ChatGPT
to Run Your Startup
Most founders use ChatGPT to write emails. The founders getting the real leverage are using it to think through decisions, prep for hard conversations, synthesise customer research, and automate the work that used to eat their afternoons. Here are the exact use cases and prompts — organised by where you’ll save the most time.
- Where AI gives founders the highest leverage
- Using ChatGPT as a thinking partner
- Decision support — how to prompt for better judgment
- Communication: emails, updates, and hard conversations
- Hiring: JDs, scorecards, and interview prep
- Fundraising: narrative, investor updates, and diligence prep
- Building your personal prompt library
Where AI Gives Founders the Highest Leverage
The founders getting the most from AI aren’t using it as a fancy autocomplete. They’re using it to compress the time between “I need to think through this” and “I have a clear enough view to act.” That’s the real leverage — not faster writing, but faster thinking.
The highest-leverage use cases, in order: structured thinking about decisions (AI as a thinking partner), first drafts of anything written (saving the blank-page problem), synthesising large amounts of information (customer research, market analysis, investor questions), and preparing for conversations that require nuance (hard feedback, investor questions, sales objections).
The lowest-leverage use case: generating content that sounds like AI. If the output reads like it was written by a competent but uninspired committee, it needs your voice added before it goes anywhere.
The single prompt that changes how founders use AI: “What am I not thinking about here?” After you’ve described a problem or decision to ChatGPT, asking this question consistently surfaces the blind spots and assumptions you’ve been carrying that you couldn’t see from inside your own thinking.
Using ChatGPT as a Thinking Partner
AI as a thinking partner is different from AI as a writing tool. You’re not asking it to produce something — you’re asking it to challenge, extend, or structure your own thinking. This requires different prompts and a different mindset: the output isn’t the deliverable, your clearer thinking is.
“I’m working through a strategic decision about [describe the situation]. Here’s my current thinking: [describe where you are]. Here’s what I’m uncertain about: [list the uncertainties]. Act as a smart, skeptical advisor. Don’t tell me what to do — ask me the 3 questions that, if I could answer them clearly, would make this decision obvious. Then tell me what I’m probably underweighting in my current thinking.”
“I’ve decided to [describe your decision]. I’m fairly confident this is right. Make the strongest possible case for the opposite decision — not to change my mind, but to make sure I’ve genuinely considered the alternatives. What would someone who strongly disagreed with me say? What evidence would support their view? What am I most likely to be wrong about?”
Decision Support
One of the most underused applications of AI for founders: preparing for a decision by exploring it systematically before making it. Most decisions that feel difficult aren’t actually difficult — they just haven’t been thought through completely. AI helps complete the thinking faster.
“I’m about to make this decision: [describe it clearly]. Assume it’s 18 months from now and this turned out to be a significant mistake. What are the 3 most likely reasons it failed? For each: how likely is it on a scale of 1-10, what are the early warning signs I should watch for in the first 90 days, and is there anything I can do now to reduce that risk?”
Communication: Emails, Updates, and Hard Conversations
The communication tasks that consume the most founder time — investor updates, hard feedback conversations, customer escalations, team updates — are exactly the ones where AI provides the most time savings with the least quality loss, because they follow predictable formats and benefit from the same structural thinking applied repeatedly.
“I need to have a difficult conversation with [describe the person and their role] about [describe the issue]. I’m nervous about this conversation because [describe what makes it hard]. Help me: (1) Identify the core thing I need to communicate clearly, (2) Anticipate how they’re likely to respond and how to handle each response, (3) Write the opening 3 sentences I should use to start the conversation, (4) Tell me what I should not say that I might be tempted to say.”
Hiring: JDs, Scorecards, and Interview Prep
Hiring is one of the most time-consuming and highest-stakes activities for a founder. AI compresses the time spent on the admin and prep work — job descriptions, scorecards, interview questions, offer letters — so the founder’s time goes to the judgment calls that actually require human judgment.
“I just interviewed a candidate for [role]. Here are my notes from the interview: [paste notes]. Here’s what I was looking for in this hire: [describe the key criteria — skills, behaviours, cultural alignment]. Based on this: (1) What are the strongest signals from this interview — positive and negative? (2) What questions do I still have that I should ask in a follow-up? (3) What’s the risk I’m most likely to be overlooking — either in hiring this person or in not hiring them? (4) On a scale of 1-5, how would you rate this candidate against each criterion and why?”
Fundraising: Narrative, Investor Updates, and Diligence Prep
Fundraising produces a predictable set of documents — the narrative, the investor update, the data room request responses — that follow patterns across every raise. AI handles the structure and first draft; the founder adds the specificity and the honest numbers that make it credible.
“I have an investor meeting next week for my Series [seed/A/B] raise. My company: [describe briefly]. My metrics: [key numbers]. Help me prepare for the Q&A. Generate the 10 hardest questions this investor is likely to ask about my business — including the ones I most want to avoid. For each question: write the honest, confident answer I should give, and flag any answer where I’m on thin ground and need to either improve the substance or be more honest about uncertainty.”
Building Your Personal Prompt Library
The founders getting the most from AI aren’t starting from scratch every time. They’ve built a personal prompt library — a collection of prompts that have worked, organised by use case, that they refine over time.
Start with the prompts in this playbook. When one produces a great output, note what made it work. When one falls short, note what was missing. After 30 days of active use, you’ll have a personalised library that reflects the specific ways your startup operates — far more valuable than any generic prompt collection.
Store it in Notion, a text file, or wherever you capture working knowledge. The goal is that any prompt you’ve used before takes 30 seconds to find and adapt, not 5 minutes to recreate from memory.
Get 200+ more tested prompts across hiring, ops, fundraising, and growth — free.