25 ChatGPT Prompts for Founders and Entrepreneurs (2026)

Most lists of ChatGPT prompts for founders are written by people who have never had to write a Sunday-night investor update after a bad week. They give you "write a business plan" prompts and pretend that is the work that matters.

The work that matters is narrower. It is the investor update that explains a missed quarter without burning the relationship. It is the rejection email to the candidate you spent six weeks recruiting. It is the customer interview synthesis that tells you whether the wedge you bet the company on is actually pulling. ChatGPT is useful for slices of that work, and useless for the rest. This list covers prompts for the slices that are actually useful.

It is written for someone running a real venture (pre-seed through Series B, 1-50 employees) who sends 20-40 emails a week and writes a board or investor update once a month. At higher scale the work changes and you delegate most of it. The prompt patterns still apply when you do not.

How to use these ChatGPT prompts for founders

A few things to know before you open a chat.

ChatGPT is good at the structured-writing tasks once you have done the thinking: investor updates from your own metrics, rejection emails, board pre-reads, candidate-spec writeups, customer interview synthesis, decision-framing memos. It is bad at the parts of founder work that require knowing your specific company: which investor will react how to a missed quarter, what your CTO is actually worried about, whether your wedge is real or whether you are convincing yourself. The free tier will confidently invent benchmarks and "industry data" if you ask. Use your own data and your own judgment for those.

Do not paste cap tables, ARR figures tied to a named company, customer names, or any MNPI into the free tier. The free tier may use your inputs for model training, and most investor agreements (and customer NDAs) require you to keep that information off third-party tools that train on inputs. For real work, use a paid plan with chat history disabled, an enterprise-licensed tool your firm has approved, or anonymise the inputs (rename customers, scale revenue by a constant factor, strip named employees).

The prompts below assume specific, real inputs. Vague briefs give you generic outputs that read like every startup essay on the internet. The bracketed examples are not optional.

Investor pitch and fundraising

Five prompts for the work that happens between fundraises.

1. Draft a 1-line company description for outreach

Prompt: "I am writing 1-line company descriptions for investor outreach. Company: [name]. What we do in plain English: [1-2 sentences]. Who pays us: [buyer role and company type]. Our category, in the words an investor would use: [paste]. The closest 2-3 comparable companies an investor would recognize: [list]. Write 5 variants of a 1-line description: (1) outcome-led (the change we create for the buyer), (2) category-led (the established market), (3) wedge-led (the specific entry point), (4) named-comparable-led ('X for Y'), (5) plain-pitch (one sentence with no analogy). Constraints: each under 18 words, no buzzwords, no 'AI-powered,' no 'platform' unless we genuinely are one. Output as a numbered list."

The "no AI-powered" line matters in 2026 because investors have spent 30 hours that week reading decks with "AI-powered" in the first sentence. Stripping it is the easiest way to read as a real company.

2. Generate 8 investor outreach subject lines

Prompt: "Generate 8 cold-investor email subject lines. Company: [1 line]. Stage: [seed / Series A / etc.]. The investor's specific thesis or recent investment that gives us standing: [paste]. The most defensible number we have to lead with: [paste, e.g., '$1.2M ARR growing 15% MoM']. Constraints: each under 45 chars, no questions, no 'quick intro,' no 'wanted to share,' no all-caps. Mix the angles: (1) number-led, (2) reference-to-investor-thesis, (3) named-customer-led, (4) curiosity without bait, (5) contrarian. Pick top 2 for the A/B test. Output: numbered list of 8, then top 2."

The "no quick intro" line is the single biggest open-rate fix for cold investor emails. Investors get 40 "quick intro" emails a week and most are unread.

3. Write the company-slide narrative for a deck

Prompt: "Draft the narrative for the 'what we do' slide in a pitch deck. Audience: [stage: seed / Series A investors at consumer / enterprise / fintech firms]. The slide has: a 6-word headline, 2-3 bullets, one product screenshot. Write the headline (under 6 words, names the outcome), the 3 supporting bullets (each starting with a verb or a number, each under 15 words), and a 1-line caption for the screenshot. Constraint: no 'revolutionizing,' no 'transforming,' no 'next-generation.' Output the 5 components labeled."

The verb-or-number-first cap on bullets is what separates a deck a partner can flip through quickly from one that needs the founder to narrate every slide.

4. Translate a recent metric into a 2-sentence investor talking point

Prompt: "Below is a recent metric we are proud of: [paste with the number and the time window]. The context that makes it meaningful: [paste 1 line]. Write a 2-sentence version a founder could drop into a partner call: sentence 1 names the metric and the comparison that makes it impressive, sentence 2 names the underlying reason it happened. Constraint: no 'we are excited,' no 'incredible growth.' Output the 2 sentences only."

This is the prompt for the moment in a partner meeting when the GP asks "what changed last quarter." Founders who answer in 2 sentences raise faster.

5. Draft a fundraise-process update for existing investors

Prompt: "Draft an email to existing investors about an active fundraise. The round: [paste, $X at $Y target close]. The status: [paste, e.g., '3 term sheets, lead candidate is firm X, we are running diligence through end of month']. The ask: [paste, e.g., 'please intro to firm Y if you have a relationship']. Structure: (1) headline in 1 sentence, (2) status in 3 bullets with the specific names and numbers, (3) the ask in 1 sentence, (4) timeline in 1 sentence. Constraint: under 150 words. No 'as you know.' No 'we are thrilled.' Output the email only."

Existing-investor process emails do the most useful work in a fundraise. Most founders write one at the start of the round and skip the rest.

Hiring and team building

Five prompts for the work that compounds across every team you build.

6. Write a role spec a senior candidate will actually read

Prompt: "Write a role spec for [role title] at [company stage and 1-line description]. The 3 problems this person will own in their first 90 days: [bullet list, with the specific metric or output each one moves]. The 2-3 things the role is explicitly NOT (so candidates self-select): [list]. Audience: senior candidates currently employed somewhere good. Structure: (1) the role in 1 sentence, (2) the 3 first-90-day problems with the metric each one moves, (3) the boundary list (what it is not), (4) the compensation range and any negotiable elements, (5) the way to apply (specific, not 'send your resume'). Constraint: no 'rockstar,' no 'ninja,' no 'wear many hats,' no 'fast-paced environment.' Output as 5 labeled sections."

The "boundary list" section is the single biggest improvement to founder hiring quality. Strong candidates self-select harder when the role tells them what it is not.

7. Draft a candidate rejection email that does not burn the relationship

Prompt: "Draft a rejection email to a senior candidate after a final-round interview. Candidate: [name, current role]. The specific reason for the rejection: [paste, be honest internally, we will decide what to share]. What we WILL share with them: [paste 1 line that is honest but not damaging]. Whether we want to keep them in our network: [yes/no]. Write the email: (1) brief opener that names the decision in the first 2 sentences, (2) 2 sentences of specific positive observation from the interview, (3) honest framing of the reason we did not move forward, (4) a clear path or no-path for future contact. Constraint: under 130 words. No 'unfortunately at this time,' no 'we regret to inform.' Output the email only."

Most founder rejection emails are written by autopilot and read by senior candidates as low-effort. The 2-sentence specific positive observation costs nothing and changes how candidates talk about the company afterward.

8. Generate 5 hiring interview questions for a specific role

Prompt: "Generate 5 behavioral interview questions for [role] at [company stage]. The 3 capabilities this role needs to demonstrate: [bullet list]. For each question: the capability it tests, the question itself, the kind of answer that signals 'yes,' the kind of answer that signals 'no,' and one follow-up question to ask if the first answer is too polished. Constraint: no 'tell me about yourself,' no 'where do you see yourself in 5 years,' no questions that can be answered from a resume. Output as 5 labeled blocks."

The "kind of answer that signals no" framing is what separates interview questions that filter from interview questions that are theater.

9. Write a reference-check question set

Prompt: "Draft a reference-check call script for a senior hire. Candidate: [name, role offered]. The 3 specific concerns I want references to address (these are the soft signals from interviews that need a third-party check): [bullet list]. Reference person's relationship to candidate: [manager / peer / direct report]. Generate 6 questions: (1) opens with low-stakes corroboration of basic facts, (2) probes for the specific concern most likely to surface, (3) asks for a recent example of the candidate's biggest weakness, (4) asks for one specific example of the candidate's biggest strength, (5) asks 'would you hire them again, and for what,' (6) closes with the cleanest possible 'is there anything I should know that I did not think to ask.' Output as a numbered list."

The "for what" qualifier on question 5 catches the polite-yes that masks a soft no.

10. Draft a board-deck slide on team

Prompt: "Draft the 'team' slide content for a board update. Company stage: [paste]. Team size: [X total, Y new this quarter]. Most important hires this quarter and the role each one fills: [paste]. Open roles still being recruited: [paste]. Attrition this quarter: [paste with reason category, not name]. Structure: (1) headline in 1 sentence that names the strategic shift in team composition, (2) 'new hires' as a table with name/role/from, (3) 'open roles' as a list with status (pipeline, finalist, paused), (4) 'attrition' as a 1-line honest note. Constraint: no 'world-class,' no 'top-tier,' no 'incredibly talented.' Output as 4 labeled blocks."

The "attrition as a 1-line honest note" rule prevents the slide from omitting the most important signal on it.

Product and customer development

Five prompts for the work between launches.

11. Synthesize 8 customer interviews into themes

Prompt: "Below are notes from [N] customer discovery interviews on [topic]. Cluster the observations into 4-6 themes. For each theme: a 1-line label, the 2-3 strongest pieces of evidence with the interviewee initials only, one tension or disagreement across speakers, and what this theme implies for the product roadmap. Constraint: only use facts in the notes. Do not invent quotes. If a theme has only one speaker behind it, label it 'single-source' so I flag it for the team. Notes: [paste]."

The "do not invent quotes" rule is load-bearing. The model will paraphrase plausible-sounding interview quotes if you do not pin this explicitly.

12. Write a feature-cut announcement to existing users

Prompt: "Draft an email to existing users announcing we are sunsetting [feature]. Reason: [paste honest reason]. What we will say publicly: [paste 1-line honest framing]. Timeline: [paste, when it stops working, when their data is exported]. Replacement or workaround: [paste]. Structure: (1) names the change in the first sentence, (2) gives the honest reason in plain English, (3) names the timeline with the specific dates, (4) names the replacement or alternative, (5) offers a path to talk to us if they have questions. Constraint: no 'in order to better serve you,' no 'we are excited to announce' (we are not), no 'streamlining our offering.' Output the email only."

The "in order to better serve you" line is the single most insulting phrase in feature-cut emails. Users read it as confirmation that we are pretending they will benefit from our cost-cutting.

13. Generate 6 product positioning angles

Prompt: "Below is what our product does in plain English: [paste 2-3 sentences]. The 3 closest competitors and what they emphasise: [bullet list, competitor name and the 1-line angle each one leads with]. Generate 6 positioning angles we could lead with: (1) the angle that competes head-on with the leader's claim, (2) the angle that addresses what the leader explicitly avoids, (3) the angle that names a specific use case the leader does not serve, (4) the contrarian angle that contradicts conventional wisdom in this category, (5) the buyer-segment angle (who specifically we are for), (6) the workflow angle (where in their day we fit). For each: a 1-line statement of the position and the risk of leading with it. Output as a numbered list."

Most early-stage companies pick the wrong positioning by default. The 6-angle generation makes it visible that the default was a choice and not a conclusion.

14. Write a roadmap announcement to a small enterprise customer

Prompt: "We just decided to deprioritize a feature that a specific enterprise customer was counting on. Customer: [name and ARR]. The feature they wanted: [paste]. Why we are deprioritizing: [paste honest reason]. What we ARE going to ship instead in that window: [paste]. Draft an email I (the founder) would send to their executive sponsor that: (1) names the change in the first sentence, (2) does not pretend it is good news for them, (3) explains the underlying tradeoff in 2 sentences, (4) offers a specific make-good (extended trial, professional services credit, named priority on a different feature), (5) commits to a 15-minute call this week. Constraint: under 200 words. No 'aligning with our strategy,' no 'we appreciate your patience.' Output the email only."

Founders write this kind of email three times a year. The cost of writing it badly is one churn event.

15. Translate a usage-data report into 3 product priorities

Prompt: "Below is product usage data for the last 30 days: [paste, DAU, retention curves by cohort, feature-use rates, top exit points, NPS comments if any]. Audience for the analysis: my product team and engineering lead. Write a 1-page synthesis: (1) the single most important pattern in 25 words, (2) the 3 supporting findings with the specific numbers, (3) the 3 product priorities for the next 30 days, ranked, with the named owner and the success criterion for each, (4) the 1 thing we should deliberately NOT do in response to this data. Constraint: do not invent benchmarks. Do not say 'industry average is X.' Output as 4 labeled sections."

The "deliberately not do" section is what stops founders from over-reacting to a single month of data.

Board reports and investor updates

Five prompts for the writing work that signals you can run the company.

16. Draft a monthly investor update

Prompt: "Draft a monthly investor update for [month]. Stage: [paste]. The 3 numbers that lead: [paste, MRR/ARR, growth rate, runway]. The 3 things that went well this month: [bullet list]. The 2 things that did not go well and what we learned: [bullet list]. The 1 specific ask for investors: [paste]. Hiring update in 1 line: [paste]. Structure: (1) headline that names the month's defining moment in 12 words, (2) the 3 numbers as a small table, (3) 'wins' as a 3-bullet list with the specific data point behind each, (4) 'lessons' as a 2-bullet honest list, (5) the ask in 1 line, (6) hiring in 1 line. Constraint: under 400 words. No 'crushing it,' no 'really excited,' no 'amazing momentum.' Tone: confident but honest. Output as 6 labeled sections."

This is the prompt I tested for this post. The verbatim ChatGPT response and editor's notes are below.

What this prompt actually returns (tested 2026-05-28)

We ran Prompt 16 on free ChatGPT with a concrete brief baked in: [Series A SaaS payments-infrastructure company; ARR $1.45M up from $1.28M (+13% MoM), 96% gross retention, 22 months runway; wins: closed first 6-figure customer Northwind Logistics ($145k ACV, 24-month term, partner referral from Stripe), shipped multi-currency reconciliation feature requested by 4 of 5 largest customers, hired Maya Reyes as senior engineering manager from Plaid; lessons: missed self-serve activation target by 14% from March funnel changes adding 36 hours to first-meaningful-payment (reverting worst by May 15), two mid-market deals slipped to Q2 from unanticipated procurement security reviews (starting security questionnaires 4 weeks earlier); ask: warm intros to VP-Finance contacts at Series A-C fintech; hiring: Head of Sales at final round, Product Designer pipeline thin]. Reproduced verbatim:

April 2026 Investor Update

1. Headline

First 6-Figure Customer Validates Enterprise Expansion and Multi-Currency Product Strategy

2. Key Numbers

Metric	April 2026
ARR	$1.45M (+13% MoM from$ 1.28M)
Gross Retention	96%
Runway	22 months at current burn

3. Wins

Closed our first 6-figure customer: Northwind Logistics signed a $145k ACV contract on a 24-month term. The deal was sourced through a partner referral from Stripe and is our largest contract to date.
Shipped the multi-currency reconciliation feature requested by 4 of our 5 largest customers during QBRs. We expect this to expand NRR potential within the existing customer base and strengthen our position with finance teams operating across multiple regions.
Hired Maya Reyes as Senior Engineering Manager after a four-month search. Maya previously served as a tech lead at Plaid and will own engineering execution across onboarding, ledger infrastructure, and platform reliability.

4. Lessons

We missed our self-serve activation target by 14% this month. Funnel changes shipped in March increased the time-to-first-meaningful-payment by an average of 36 hours. The team identified the largest friction points and is reverting the worst-performing changes by May 15.
Two mid-market deals expected to close in March slipped into Q2 after procurement introduced vendor-security reviews we had not anticipated early enough in the cycle. We are now initiating security questionnaires approximately four weeks earlier during sales engagement.

5. Investor Ask

We are looking for warm introductions to VP-of-Finance leaders at Series A–Series C fintech companies, particularly teams managing cross-border or multi-currency payment operations.

6. Hiring

We currently have two senior roles open: Head of Sales and Product Designer. The Head of Sales search is at final-round stage with a lead candidate; Product Designer sourcing remains the bigger gap in the pipeline.

Editor's notes. The model held the load-bearing structure exactly as the brief asked: all 6 sections present and labeled, the 12-word headline cap respected (the model's headline is 10 words), and the banned phrases ("crushing it," "really excited," "amazing momentum") cleanly absent. Every one of the 11 numbers from the brief is preserved intact ( $1.45M,$ 1.28M, 13%, 96%, 22 months, $145k, 24 months, 14%, 36 hours, May 15, 4 weeks). The honest framing in Lessons reads like a real Series A founder wrote it, not a model: the "the team identified the largest friction points and is reverting the worst-performing changes by May 15" sentence is sharper than what most founders write on their first try. The four things to fix before sending: (1) the Investor Ask section runs to 2 lines / 30 words against the prompt's explicit "ask in 1 line" constraint, so the model violated a stated rule and the ask needs trimming to "We need warm intros to VP-Finance contacts at Series A-C fintech companies" or similar; (2) the Hiring section runs 2 sentences / 36 words against the same "1 line" rule, needs trimming to "Two senior roles open: Head of Sales (final-round, lead candidate) and Product Designer (sourcing remains thin)"; (3) the second Wins bullet adds "strengthen our position with finance teams operating across multiple regions" which is editorial expansion the brief did not authorize, easy to cut; (4) the headline names "Multi-Currency Product Strategy" as a defining framing, which is one of two reasonable readings (the other being the Northwind deal itself), so a founder might want the alternative "First 6-Figure Customer Closed via Partner Referral" headline as the variant for an A/B with their angel-list. Also worth noting: the model added an unrequested preamble ("Here's a draft formatted for a typical monthly investor update.") which is a recurring pattern across all prompts that ask for output-only and should be explicitly banned in the prompt template ("Output the 6 sections only, no preamble"). Net: shippable as an 85% first draft, ~5 minutes of editing to tighten the 1-line constraints in sections 5 and 6 and to cut the editorial line in Wins.

17. Write a quarterly board pre-read

Prompt: "Draft a 1-page board pre-read for the upcoming board meeting on [date]. Company stage: [paste]. The 3 most important strategic questions on the table this quarter: [bullet list]. The 3 numbers that matter most: [paste]. The decision we need the board to make: [paste]. Structure: (1) headline (the meeting's defining question in 1 sentence), (2) state of the business in 3 numbers, (3) the 3 strategic questions with the founder's recommendation on each, (4) the named decision the board is being asked to make, (5) the 1 thing the board could reasonably push back on, (6) the 'what I am worried about' section in 2 sentences. Constraint: under 500 words. No 'aligning with our vision.' No 'taking ownership.' Output as 6 labeled sections."

The 'what I am worried about' section is the part most board pre-reads omit and the part most experienced board members want to see.

18. Draft a 1-sentence quarterly headline

Prompt: "Below is our quarterly business data: [paste, top-line metrics, growth rate, customer count, pipeline]. The 3 things the quarter actually changed strategically: [bullet list]. Write 5 candidate 1-sentence headlines for the quarter, varying angle: (1) the number that most changed, (2) the named customer or named segment that emerged, (3) the strategic shift, (4) the lesson learned, (5) the new question now in front of us. Constraint: each under 22 words. No 'a quarter of strong execution.' Output as a numbered list, then pick the top one and explain in 1 sentence."

This is the headline that opens every board update and investor email for the quarter. Generating 5 and picking 1 takes 4 minutes and matters more than the next 3 hours of slide-polishing.

19. Write a difficult investor email about a missed quarter

Prompt: "We missed our growth target this quarter. Actual: [X]. Plan: [Y]. The honest reason: [paste]. The fix: [paste plan]. The runway impact: [paste, how this changes the cash plan]. Draft a 200-word email to lead investors that: (1) names the miss in the first sentence with the exact number, (2) gives the honest reason in 2 sentences, (3) names the specific fix and the timeline, (4) names the cash-plan impact honestly, (5) does not over-promise the recovery. Constraint: no 'a tough quarter,' no 'headwinds,' no 'lots of learnings.' Tone: direct, confident in the diagnosis, honest about the uncertainty. Output the email only."

The "no headwinds" guardrail is necessary. The word "headwinds" tells investors you do not yet know what actually happened.

20. Generate 3 candidate openers for a sensitive update

Prompt: "We need to deliver an update that includes news investors will not love: [paste headline]. The reason it is uncomfortable: [paste]. Generate 3 candidate first sentences for the update. Each must (1) name the news in the first 12 words, (2) signal we have done the diagnostic work, (3) avoid passive voice and weasel words. The 3 should vary across angles: (a) lead with the cause, (b) lead with the plan, (c) lead with the number. Constraint: each under 25 words. No 'we want to be transparent.' No 'I want to share.' Output as a numbered list."

Three options is the right number. With one option the investor reads it as fait accompli; with three the founder picks the framing that fits the specific investor's temperament.

Personal effectiveness and decision-making

Five prompts for the founder work that has no team to delegate to.

21. Write a 1-page decision memo before a hard call

Prompt: "I am about to make a decision: [paste the decision in 1 sentence]. The 3 most plausible options: [bullet list]. The data I have: [paste]. The data I do NOT have but wish I did: [paste]. The deadline: [paste]. The reversibility: [reversible / hard to reverse / one-way door]. Draft a 1-page memo: (1) the decision in 1 sentence, (2) the options with the strongest argument for each (no straw men), (3) the named recommendation with the specific reasons, (4) the 2 risks of the recommendation, (5) the 1 piece of information that would change my mind. Constraint: do not hedge. Output as 5 labeled sections."

The "1 piece of information that would change my mind" rule is the single biggest improvement to founder decision quality. It forces a falsification condition that most decisions skip.

22. Draft a weekly priority list from a chaotic week

Prompt: "Below is my brain dump of everything on my plate going into next week: [paste, in any order]. My role: [founder/CEO of stage X]. The 3 outcomes I am supposed to drive this quarter: [bullet list]. Cluster the brain dump into 4 buckets: (1) Must do personally this week (decisions or external relationships only I can hold), (2) Must be moved this week but I can delegate, (3) Reactive only (I will handle if it comes to me), (4) Not this week (I will park). For each item: which bucket and a 1-sentence reason. Constraint: bucket 1 has a max of 5 items. Output as a labeled table."

The 5-item cap on bucket 1 is the load-bearing constraint. Without it the model gives back everything in bucket 1 and you are no better off.

23. Write a 'no' to a partnership request that does not fit

Prompt: "I received a partnership inquiry from [name, company]. The proposal: [paste]. Why it does not fit for us right now: [paste honest reason]. The version that would fit if anything would: [paste, if exists]. Draft a reply that: (1) declines the specific proposal in the first sentence, (2) explains the underlying reason in 1 sentence (not in 3 paragraphs), (3) names the version that would fit if any does, (4) keeps the door open without committing to a future revisit. Constraint: under 90 words. No 'unfortunately at this time,' no 'we love what you are doing.' Output the reply only."

Most partnership-no emails take 20 minutes to write and read worse than a 90-second version. The constraint is the value.

24. Synthesize a difficult conversation into 3 action items

Prompt: "I just had a difficult conversation with [role, e.g., 'my COO']. The conversation was about: [paste]. What they said: [paste, anonymised]. What I said: [paste]. What I left unsaid: [paste]. Draft a 1-page synthesis for my own notes: (1) the actual disagreement in 1 sentence, (2) the part I think they were right about, (3) the part I think I was right about, (4) the 3 specific things I will change in response, (5) the 1 conversation I owe them by Friday. Constraint: be honest about the parts I do not yet know. Output as 5 labeled sections."

Founders avoid this synthesis exercise more than they avoid the conversation itself. The prompt removes the friction.

25. Draft a personal weekly retro

Prompt: "Draft my weekly founder retro for [week]. The 3 things I am proudest of: [paste, even if small]. The 2 things that went badly: [paste honestly]. The 1 conversation I avoided this week: [paste]. The 1 hire I should be making and am not: [paste]. Structure as 5 short sections, 2-3 sentences each. Constraint: do not let me off the hook on the avoided conversation or the missing hire. End with one specific thing I will do differently next week. Output as 5 labeled sections."

The "do not let me off the hook" instruction works on the model in a way that pure self-discipline does not.

Tips for getting better results

Three things that matter more than any specific prompt.

Be honest about the unstated objective. Every founder task has the stated objective on paper and a different one in your head. Naming the unstated one in your prompts is what separates outputs that get used from outputs that get ignored. The model cannot read your situation, but it can use the information you choose to give it.

Ask for 5, pick 1. ChatGPT is faster at generating variants than picking the best one. Use it for the volume, use your taste for the selection. The reverse does not work.

Treat the first answer as the safe answer. ChatGPT's first version is always the safe, balanced, vendor-neutral version. The useful version comes from the follow-up: "make the recommendation sharper," "remove the hedge," "what is the version a senior board member would push back on." Budget for the second cycle.

FAQ

Should I paste my company data into ChatGPT?

Not into the free tier under any circumstances. Cap tables, ARR figures, named customers, hiring decisions, and board discussion notes are confidential by default. The free tier may use your inputs for model training. Use a paid plan with chat history disabled (ChatGPT Team or Enterprise), an enterprise-licensed tool your firm has approved, or anonymise inputs aggressively before pasting. For MNPI the answer is always the firm-approved tool, never the free tier.

Can ChatGPT replace a chief of staff?

No, and the framing misses the actual question. ChatGPT can take an hour off a half-day task for a founder who already knows what to ask. It cannot replace the chief of staff's role (knowing which stakeholders matter, calling the political risk of a decision, holding the founder accountable to a calendar). Founders who try to skip the chief of staff with AI end up writing all the synthesis themselves, just faster.

Are there founder tasks ChatGPT should never touch?

Three I would name: (1) anything that requires you to make a specific hiring decision, because the model will rationalise the candidate you already favored; (2) anything involving a personal performance issue with a named employee, because the model's diplomatic register makes hard conversations softer than they should be; (3) anything you would not paste into your own group chat with your investors, because the data-handling question is real.

What is the single most useful prompt for a first-time founder?

Prompt 21: the 1-page decision memo. Most first-time founders make decisions in their heads and then defend them post-hoc. Writing the memo before deciding fixes more bad decisions than any single piece of advice in any book.

How do I tell whether a ChatGPT-drafted investor update is good?

Read the headline out loud. If your lead investor could send it back as "this could be any startup's update," it is not specific enough. The job of the headline is to tell the reader what this month was actually about. If the headline does that, the body almost always works.

What to do next

Pick one decision you are putting off this week. Run Prompt 21 (1-page decision memo) on it tonight. The exercise takes 12 minutes and removes more procrastination than any time-management technique.

If you have an investor update due this month, draft it using Prompt 16. Compare the result to the update you would have written. Notice the specific places the model added or removed something useful.

Send one rejection email using Prompt 7 this week. Watch how the candidate replies. The 2-sentence specific positive observation changes the relationship after the no.