Founders who paste "should I raise a Series A?" into ChatGPT get back a Wikipedia paragraph about the fundraising process. Feed o3 a structured decision prompt — situation, constraints, options, priors, success criteria — and you get the kind of layered analysis you'd pay a strategy firm $80k for: ranked options, explicit assumptions, failure modes, and the precise conditions that would change the recommendation. These 30 prompts are built for exactly that.
Why o3 Is Different for Strategy Work
o3 is a reasoning model, which means it doesn't just pattern-match to training data — it treats your problem as a multi-variable system to work through. For strategic decisions, that distinction matters. A standard LLM will summarize considerations. o3 will weigh them against each other, surface second-order effects, and tell you where its own reasoning is weakest. That's the difference between a briefing document and a decision partner.
The key is framing your prompt as a decision under uncertainty. Don't ask o3 to "help you think through" something. Give it: what choice you're making, what the alternatives are, what constraints are non-negotiable, what your current priors are, and what success looks like in 12 months. That structure forces o3 to reason about tradeoffs rather than generate a list of considerations.
Require an explicit assumptions list in every strategic prompt. An assumptions list is the strongest signal of quality you can get from a reasoning model — it shows you exactly where the recommendation is load-bearing. If o3 assumes your market is growing at [X]% annually and it's actually flat, the recommendation flips. Surfacing that assumption early is worth more than the recommendation itself.
Require failure modes in every prompt. "What would change your recommendation?" is not a rhetorical question — it's the mechanism for stress-testing the answer. You want to know: if one thing goes wrong, which thing is it, and what does the world look like when it does? For irreversible decisions especially, the failure modes matter as much as the primary recommendation.
For decisions you can't easily undo — a layoff, a fundraising round, a major hire, an acquisition — set reasoning_effort to high. o3 will spend more compute on the problem. The cost difference is trivial. The quality difference on a $5M decision is not. You can read more about tuning reasoning models in the complete guide to AI reasoning models, and see how this approach complements the broader toolkit in our AI prompts for startup founders guide.
30
Copy-paste o3 prompts for high-stakes strategic work
Pricing & Monetization Prompts (1–5)
1. Price-Point Analysis
code
Decision: What price should I charge for [PRODUCT/PLAN NAME]?
Situation:
- Product: [BRIEF DESCRIPTION — what it does, who it's for]
- Current price (if any): [AMOUNT or "not yet launched"]
- Comparable products charge: [RANGE — e.g., "$49–$149/month"]
- My estimated cost to deliver per customer: [AMOUNT]
- Quantifiable value delivered: [e.g., "saves ~4 hours/week for a $80k/yr employee"]
- Sales motion: [self-serve / sales-assisted / enterprise]
Options I'm considering:
A. [PRICE OPTION A]
B. [PRICE OPTION B]
C. [PRICE OPTION C]
Constraints:
- Minimum margin required: [PERCENTAGE or AMOUNT]
- Launch timeline: [DATE]
- I cannot go below [FLOOR PRICE] without violating unit economics
What I've already considered: [LIST ANYTHING YOU'VE THOUGHT THROUGH]
Required output:
1. Recommended price point with confidence level (0–100%)
2. Explicit list of assumptions this recommendation depends on
3. 3 failure modes: what goes wrong if this recommendation is wrong
4. What would change the recommendation (e.g., "if churn exceeds X%...")
5. One experiment to run in the first 30 days to validate pricing
reasoning_effort: high
Re-derive your recommendation from first principles and check for
internal inconsistencies before finalizing your answer.2. Packaging & Tier Design
code
Decision: How should I structure my product tiers?
Situation:
- Product: [DESCRIPTION]
- Current structure: [DESCRIBE CURRENT TIERS or "none yet"]
- Customer segments I'm trying to serve: [LIST 2–3 DISTINCT SEGMENTS]
- Features available: [LIST KEY FEATURES]
- Revenue goal: [TARGET MRR/ARR]
- Average deal size today: [AMOUNT or "pre-revenue"]
Options:
A. [TIER STRUCTURE OPTION A — e.g., "Free / Pro / Business"]
B. [TIER STRUCTURE OPTION B — e.g., "Starter / Growth / Enterprise"]
C. [OTHER STRUCTURE IF APPLICABLE]
Constraints:
- Engineering bandwidth to build/maintain tiers: [LOW / MEDIUM / HIGH]
- I need to protect [FEATURE] for paid-only users
What I've already considered: [YOUR CURRENT THINKING]
Required output:
1. Recommended tier structure with feature allocation and price anchors
2. Confidence level and key assumptions
3. 3 failure modes (e.g., which tier cannibalizes another)
4. What would change the recommendation
5. The most common packaging mistake for this type of product
reasoning_effort: high
Re-derive your recommendation from first principles and check for
internal inconsistencies before finalizing.3. Raising Prices on Existing Customers
code
Decision: Should I raise prices on existing customers, and if so, how?
Situation:
- Current price: [AMOUNT]
- Proposed new price: [AMOUNT]
- Percentage increase: [%]
- Customer base: [NUMBER OF CUSTOMERS, roughly]
- Average tenure: [HOW LONG CUSTOMERS HAVE BEEN WITH YOU]
- Churn rate today: [MONTHLY or ANNUAL %]
- Last price change: [DATE or "never"]
- Reason for increase: [COST INFLATION / UNDERPRICED VS VALUE / REPOSITIONING]
Constraints:
- I have [X months] of runway — this affects how much churn I can absorb
- Key contracts expiring: [DATE RANGE if applicable]
- Competitor pricing: [WHAT THEY CHARGE]
What I've already considered: [E.g., "grandfathering legacy users," "phased rollout"]
Required output:
1. Recommendation: raise now / raise with conditions / hold — with confidence level
2. If raise: recommended rollout approach and messaging
3. Explicit assumptions list
4. 3 failure modes (including churn scenario math if possible)
5. What would change the recommendation
6. The one metric to watch in the 60 days after the price change
reasoning_effort: high
Re-derive your recommendation from first principles. Check whether
the churn risk is overstated or understated given the assumptions.4. Freemium vs. Paid-Only
code
Decision: Should I offer a free tier, or go fully paid from the start?
Situation:
- Product: [DESCRIPTION]
- Stage: [PRE-LAUNCH / EARLY REVENUE / SCALING]
- Current MRR: [AMOUNT or $0]
- Primary acquisition channel today: [e.g., SEO, paid ads, word-of-mouth]
- Average time-to-value for a new user: [HOW LONG BEFORE THEY GET THE AHA MOMENT]
- Support cost per free user (estimate): [AMOUNT or "unknown"]
- What free users would get: [FEATURE SET]
Options:
A. Freemium — free tier with paid upgrade
B. Free trial (time-limited) — then paid
C. Fully paid, no free tier
Constraints:
- Team size: [NUMBER] — affects support capacity
- My goal for next 12 months: [USERS / REVENUE / BOTH]
What I've already considered: [YOUR CURRENT THINKING]
Required output:
1. Recommendation with confidence level
2. Key assumptions the recommendation depends on
3. 3 failure modes for the recommended approach
4. What would change the recommendation
5. The one signal in the first 90 days that tells you the choice was right or wrong
reasoning_effort: high
Re-derive from first principles. Flag if the answer changes
materially at different team sizes or traffic levels.5. Enterprise Pricing Model
code
Decision: Should I move to custom/enterprise pricing, and how should I structure it?
Situation:
- Current model: [FLAT RATE / SEAT-BASED / USAGE-BASED / OTHER]
- Current max plan: [PRICE]
- Enterprise deals I've lost or almost closed: [DESCRIBE — size, why won/lost]
- Enterprise prospect profile: [COMPANY SIZE, INDUSTRY]
- What enterprise customers need that SMB doesn't: [e.g., SSO, audit logs, SLAs]
- Revenue from top 3 customers as % of total: [%]
Options:
A. Add an enterprise tier with custom pricing (sales-negotiated)
B. Usage-based model at scale (pay per seat/event/API call)
C. Keep current model, add enterprise add-ons
D. Platform fee + consumption pricing
Constraints:
- I have [X] salespeople or [none yet]
- Time to first enterprise close I can tolerate: [MONTHS]
What I've already considered: [YOUR CURRENT THINKING]
Required output:
1. Recommended pricing model with confidence level
2. Explicit assumptions list
3. 3 failure modes (including sales cycle risk)
4. What would change the recommendation
5. One thing enterprise buyers will ask for that you should decide on before the first call
reasoning_effort: high
Re-derive from first principles. Check whether the assumptions
about enterprise buyer behavior are well-grounded.Hiring & Team Prompts (6–10)
6. Should We Hire Now?
code
Decision: Should I make this hire now, delay 3–6 months, or not at all?
Role: [JOB TITLE]
Situation:
- Problem this hire solves: [SPECIFIC BOTTLENECK OR CAPABILITY GAP]
- What happens if we don't hire: [CONSEQUENCES IN 6 MONTHS]
- Current runway: [MONTHS]
- Burn rate today: [MONTHLY]
- Loaded cost of this hire: [ANNUAL SALARY + BENEFITS + OVERHEAD]
- Revenue trajectory: [GROWING / FLAT / DECLINING — with rate]
- Who is doing this work now (if anyone): [NAME/ROLE or "no one"]
Options:
A. Hire now
B. Delay until [MILESTONE — e.g., $X MRR, next funding close]
C. Don't hire — solve with contractors, tools, or restructure
Constraints:
- Board/investor guidance on headcount: [ANY RESTRICTIONS]
- Time to productivity for this role: [MONTHS]
What I've already considered: [YOUR CURRENT THINKING]
Required output:
1. Recommendation with confidence level
2. Key assumptions list
3. 3 failure modes
4. What would change the recommendation
5. If delay: what specific trigger should prompt revisiting this decision
reasoning_effort: high
Re-derive from first principles. Check whether the urgency is
real or manufactured by proximity bias.7. Exec Hire vs. Internal Promotion
code
Decision: Should I hire an external executive for [ROLE] or promote from within?
Situation:
- Role: [VP SALES / CTO / CFO / etc.]
- Internal candidate: [BRIEF PROFILE — strengths, gaps, tenure]
- External market: [WHAT'S AVAILABLE — rough seniority/cost of external hire]
- Company stage: [SERIES A / B / GROWTH / etc.]
- How long the role has been unfilled or acting: [MONTHS]
- What changed that makes this decision urgent: [TRIGGER]
Options:
A. Promote [INTERNAL CANDIDATE NAME or "internal candidate"]
B. External hire
C. Interim/fractional while we decide
Constraints:
- Budget for external hire: [RANGE]
- Risk of losing internal candidate if not promoted: [HIGH / MEDIUM / LOW]
- Time to fill if external: [ESTIMATED MONTHS]
What I've already considered: [YOUR CURRENT THINKING]
Required output:
1. Recommendation with confidence level
2. Key assumptions (especially about internal candidate ceiling)
3. 3 failure modes for each path (internal and external)
4. What would change the recommendation
5. The one interview question or reference check that would most update your view
reasoning_effort: high
Re-derive from first principles. Check whether the preference
for one option is driven by sunk-cost reasoning or genuine fit.8. Build vs. Acquire Team Capability
code
Decision: Should I build this capability internally or acquire it?
Capability needed: [e.g., "enterprise sales team," "ML infrastructure," "content operation"]
Situation:
- Why we need it: [STRATEGIC REASON]
- Timeline to need it: [MONTHS]
- Build path: [WHAT IT WOULD TAKE — hires, cost, time to productivity]
- Acquire path: [WHAT'S AVAILABLE — acquihire target, team acquisition, outsource]
- Our current ability to attract/retain in this function: [STRONG / WEAK / UNKNOWN]
- Comparable companies' approach: [IF KNOWN]
Options:
A. Build — hire [NUMBER] people over [TIMELINE]
B. Acquire — target [COMPANY/TEAM TYPE]
C. Partner — outsource or white-label
Constraints:
- Capital available for acquisition: [RANGE or "none"]
- Integration bandwidth: [LOW / MEDIUM / HIGH]
What I've already considered: [YOUR CURRENT THINKING]
Required output:
1. Recommendation with confidence level
2. Key assumptions list
3. 3 failure modes
4. What would change the recommendation
5. The one diligence question that most determines which path to take
reasoning_effort: high
Re-derive from first principles. Check build timeline estimates
for optimism bias.9. Compensation Band Design
code
Decision: How should I set comp bands for [ROLE/LEVEL]?
Situation:
- Role: [TITLE AND LEVEL]
- Location: [CITY / REMOTE / DISTRIBUTED]
- Current pay for people in this role at my company: [RANGE]
- Market data available: [e.g., "Levels.fyi, Radford, Carta data" or "none"]
- Stage: [SEED / SERIES A / etc.]
- Equity component: [% or amount, vesting structure]
- Revenue per employee: [AMOUNT — signals ability to pay]
Options:
A. [PERCENTILE] of market (e.g., 50th percentile cash + equity kicker)
B. [PERCENTILE] of market cash + above-market equity
C. Below-market cash, above-market equity (startup-style)
Constraints:
- Budget available for this hire: [RANGE]
- Existing team pay I can't compress against: [ANY ANCHORS]
- Candidate profile I'm targeting: [SENIORITY, BACKGROUND]
What I've already considered: [YOUR CURRENT THINKING]
Required output:
1. Recommended band with confidence level
2. Key assumptions
3. 3 failure modes (including offer rejection and internal equity compression)
4. What would change the recommendation
5. What to do when a candidate comes in above band
reasoning_effort: high
Re-derive from first principles. Check whether market data
assumptions are current and applicable to company stage.10. Layoff Sizing
code
Decision: If we need to reduce headcount, how deep should the cut be?
Situation:
- Current headcount: [NUMBER]
- Current monthly burn: [AMOUNT]
- Runway at current burn: [MONTHS]
- Target runway after cut: [MONTHS]
- Revenue trend: [GROWING / FLAT / DECLINING — with rate]
- Areas of the business contributing to vs. consuming revenue: [BRIEF BREAKDOWN]
- What triggered this decision: [e.g., "missed fundraise," "revenue shortfall," "macro"]
Options:
A. [X%] reduction — achieves [Y months] runway
B. [X%] reduction — achieves [Y months] runway
C. [X%] reduction — achieves [Y months] runway
Constraints:
- Functions I cannot cut without breaking the product: [LIST]
- Key individuals I cannot afford to lose: [ROLES, not names]
- Legal/notice period requirements: [JURISDICTION]
What I've already considered: [YOUR CURRENT THINKING]
Required output:
1. Recommended reduction size with confidence level
2. Key assumptions (especially revenue recovery assumptions)
3. 3 failure modes (including second-wave scenario)
4. What would change the recommendation
5. The sequencing decision: announce all at once vs. in tranches
reasoning_effort: high
Re-derive from first principles. Check whether the revenue
recovery assumption in each scenario is realistic or hopeful.Capital & Fundraising Prompts (11–15)
11. Bridge vs. Priced Round
code
Decision: Should I do a bridge round now or wait for a priced round?
Situation:
- Current runway: [MONTHS]
- Current MRR/ARR: [AMOUNT]
- MoM or YoY growth rate: [%]
- Last round: [AMOUNT, TYPE, DATE, VALUATION]
- Bridge amount needed: [RANGE]
- Priced round target: [AMOUNT, TARGET VALUATION]
- Investor appetite today: [WARM LEADS / COLD / UNKNOWN]
Options:
A. Bridge now on SAFEs/convertible notes — buy [X] months of runway
B. Go straight to priced round in the next [X] months
C. Bridge + start priced round process simultaneously
Constraints:
- Dilution I can absorb on the bridge: [% or cap]
- Key metrics I need to hit for priced round: [LIST]
- Time I can spend fundraising without losing operational focus: [WEEKS/MONTH]
What I've already considered: [YOUR CURRENT THINKING]
Required output:
1. Recommendation with confidence level
2. Key assumptions (especially about fundraising timeline)
3. 3 failure modes
4. What would change the recommendation
5. The one metric that, if you hit it, most changes investor appetite
reasoning_effort: high
Re-derive from first principles. Check whether the priced round
timeline assumption is grounded in real investor conversations
or optimism.12. Valuation Framing for Fundraise
code
Decision: How should I frame and anchor my valuation in this fundraise?
Situation:
- Stage: [PRE-SEED / SEED / SERIES A / etc.]
- ARR or MRR: [AMOUNT]
- Growth rate: [YoY or MoM]
- Gross margin: [%]
- Comparable rounds in my space (if known): [EXAMPLES or "unknown"]
- My asking terms: [AMOUNT RAISED, TARGET VALUATION]
- Lead investor status: [COMMITTED / IN DILIGENCE / NOT YET]
Options:
A. Anchor at [VALUATION A] — aggressive, tests ceiling
B. Anchor at [VALUATION B] — market rate, faster close
C. Let the lead set terms and react
Constraints:
- Minimum acceptable valuation: [AMOUNT]
- Timeline to close: [MONTHS — driven by runway]
- Participation from existing investors: [FOLLOWING ON / PASSING / UNKNOWN]
What I've already considered: [YOUR CURRENT THINKING]
Required output:
1. Recommended valuation anchor and framing with confidence level
2. Key assumptions (especially about comparable deals)
3. 3 failure modes
4. What would change the recommendation
5. The one narrative framing that best justifies the valuation given the metrics
reasoning_effort: high
Re-derive from first principles. Check whether comp assumptions
reflect current market conditions, not 2021-era multiples.13. Runway vs. Growth Tradeoff
code
Decision: Should I optimize for extending runway or accelerating growth?
Situation:
- Current runway: [MONTHS]
- Current burn: [MONTHLY]
- Growth rate: [MoM or YoY]
- Gross margin: [%]
- Net revenue retention: [%]
- Next milestone for fundraise or profitability: [DESCRIBE]
- What it costs to grow faster: [ADDITIONAL MONTHLY BURN if we invest more]
- What it costs to extend runway: [CUTS NEEDED — growth rate impact]
Options:
A. Extend runway — cut [X], reach [Y months] runway at [Z% growth]
B. Accelerate growth — invest [X], reach [Y% growth] with [Z months] runway
C. Hybrid — selective investment in [AREA], partial cuts to [AREA]
Constraints:
- Investor expectations for next round metrics: [IF KNOWN]
- Team morale impact of cuts: [HIGH / MEDIUM / LOW]
What I've already considered: [YOUR CURRENT THINKING]
Required output:
1. Recommendation with confidence level
2. Key assumptions
3. 3 failure modes
4. What would change the recommendation
5. The growth metric that investors in your space will care most about at next raise
reasoning_effort: high
Re-derive from first principles. Check whether the growth
investment assumption has evidence behind it or is directional hope.14. M&A Target Evaluation
code
Decision: Should we acquire [TARGET COMPANY / TYPE OF COMPANY]?
Situation:
- Target: [BRIEF DESCRIPTION — what they do, size, stage]
- Deal structure being discussed: [CASH / STOCK / ACQUIHIRE / MERGER]
- Deal size: [RANGE]
- Strategic rationale: [WHY THIS TARGET — technology, team, customers, revenue]
- Alternatives to acquiring: [BUILD, PARTNER, IGNORE]
- Integration complexity: [LOW / MEDIUM / HIGH — with brief reasoning]
- Target's current revenue and growth: [IF KNOWN]
Options:
A. Acquire at [TERMS]
B. Structured partnership instead
C. Walk away — build or buy differently
Constraints:
- Cash/equity available for acquisition: [AMOUNT]
- Integration bandwidth: [HEADCOUNT, TIMELINE]
- Key person retention risk: [HIGH / MEDIUM / LOW]
What I've already considered: [YOUR CURRENT THINKING]
Required output:
1. Recommendation with confidence level
2. Key assumptions (especially integration timeline and retention)
3. 3 failure modes (including post-close integration risk)
4. What would change the recommendation
5. The one diligence finding that would kill the deal
reasoning_effort: high
Re-derive from first principles. Check whether strategic rationale
is genuine or post-hoc justification for a target we already like.15. Secondary Sale Decision
code
Decision: Should I sell secondary shares in this round, and how much?
Situation:
- My ownership stake: [%]
- Implied value of my stake at proposed round valuation: [AMOUNT]
- Amount I'm considering selling: [AMOUNT or %]
- My personal financial situation: [DIVERSIFIED / CONCENTRATED / NEED LIQUIDITY]
- Investor reaction to secondary (if known): [SUPPORTIVE / NEUTRAL / NEGATIVE]
- Company trajectory: [STRONG / UNCERTAIN — brief reasoning]
- Time to likely exit: [ESTIMATE — IPO, acquisition, or "unclear"]
Options:
A. Sell [AMOUNT] secondary in this round
B. Sell nothing — maximize upside
C. Sell [SMALLER AMOUNT] as a hedge
Constraints:
- Board/investor approval required: [YES / NO / CONDITIONAL]
- Signal risk: [DOES SELLING SIGNAL LACK OF CONVICTION]
- Tax implications: [IF SIGNIFICANT — note jurisdiction]
What I've already considered: [YOUR CURRENT THINKING]
Required output:
1. Recommendation with confidence level
2. Key assumptions (especially about company trajectory and exit timeline)
3. 3 failure modes
4. What would change the recommendation
5. How to frame secondary participation to investors if asked
reasoning_effort: high
Re-derive from first principles. Check whether the decision
is driven by financial logic or by social/signaling anxiety.Competitive Strategy Prompts (16–20)
16. Response to a New Competitor
code
Decision: How should I respond to [COMPETITOR NAME or TYPE] entering our market?
Situation:
- Competitor: [WHO THEY ARE — size, backing, product]
- Their go-to-market: [PRICING, POSITIONING, TARGET SEGMENT]
- Our current position: [MARKET SHARE ESTIMATE, DIFFERENTIATION]
- Overlap with our customers: [HIGH / MEDIUM / LOW — explain]
- Their likely advantages: [LIST]
- Our likely advantages: [LIST]
- Time since they launched or announced: [WHEN]
Options:
A. Accelerate on our current roadmap — don't react directly
B. Directly address their advantages — match or beat on [FEATURE/PRICE]
C. Re-segment — focus on the customers they can't serve well
D. Do nothing for now — monitor
Constraints:
- Engineering bandwidth to respond: [SPRINTS / QUARTERS]
- Budget to accelerate marketing: [RANGE]
- Sales team capacity to re-pitch at-risk accounts: [YES / NO]
What I've already considered: [YOUR CURRENT THINKING]
Required output:
1. Recommended response with confidence level
2. Key assumptions (especially about competitor execution speed)
3. 3 failure modes
4. What would change the recommendation
5. The one move that makes us harder to compete with, regardless of this competitor
reasoning_effort: high
Re-derive from first principles. Check whether the threat is real
or whether we're overweighting a press release.17. When to Undercut on Price
code
Decision: Should I lower my price to compete with [COMPETITOR] on [SEGMENT]?
Situation:
- My current price: [AMOUNT]
- Competitor's price: [AMOUNT]
- Segment at risk: [DESCRIBE — size, revenue contribution]
- Win/loss data: [ARE WE LOSING DEALS TO PRICE — with frequency]
- My gross margin at current price: [%]
- My gross margin at proposed lower price: [%]
- What the price difference actually buys the customer: [HONEST ASSESSMENT]
Options:
A. Match competitor price for [SEGMENT] — accept margin compression
B. Hold price — invest in differentiation to justify premium
C. Create a lower-priced tier that doesn't cannibalize the main product
D. Drop price only for [SPECIFIC DEAL TYPE or SEGMENT]
Constraints:
- Minimum acceptable gross margin: [%]
- Risk of pricing signal across the broader customer base: [HIGH / LOW]
- Competitor's ability to respond with a further price cut: [LIKELY / UNLIKELY]
What I've already considered: [YOUR CURRENT THINKING]
Required output:
1. Recommendation with confidence level
2. Key assumptions (especially win/loss root cause)
3. 3 failure modes
4. What would change the recommendation
5. The question to ask in sales calls to confirm whether price is the real objection
reasoning_effort: high
Re-derive from first principles. Check whether undercutting
solves the real problem or treats a symptom.18. Geographic Expansion
code
Decision: Should I expand into [GEOGRAPHY] now, and if so, how?
Situation:
- Current markets: [WHERE WE OPERATE TODAY]
- Proposed new geography: [REGION / COUNTRY]
- Evidence of pull from that geography: [INBOUND SIGNALS, EXISTING CUSTOMERS, ETC.]
- What expansion requires: [LOCALIZATION, LEGAL ENTITY, LOCAL HIRE, etc.]
- Estimated cost and timeline to meaningful revenue: [RANGE]
- How competitors are positioned there: [STRONG / WEAK / NOT PRESENT]
Options:
A. Full expansion — local entity, hire, localized product
B. Light expansion — serve inbound remotely, no local presence
C. Partner model — local reseller or distributor
D. Delay until [MILESTONE]
Constraints:
- Team bandwidth to support a new geography: [LOW / MEDIUM / HIGH]
- Revenue from current markets still growing: [YES — don't distract / PLATEAUING — expand]
- Regulatory requirements: [SIGNIFICANT / MINOR / UNKNOWN]
What I've already considered: [YOUR CURRENT THINKING]
Required output:
1. Recommendation with confidence level
2. Key assumptions
3. 3 failure modes (including attention diversion risk)
4. What would change the recommendation
5. The single best signal of product-market fit in a new geography before committing
reasoning_effort: high
Re-derive from first principles. Check whether expansion is
driven by opportunity or by founder travel enthusiasm.19. Channel vs. Direct Sales
code
Decision: Should I invest in a channel/partner program or double down on direct sales?
Situation:
- Current sales motion: [DIRECT / SOME CHANNEL / MIXED]
- ACV (average contract value): [AMOUNT]
- Sales cycle length: [WEEKS/MONTHS]
- Win rate: [%]
- Channel partners available/interested: [DESCRIBE — type, reach]
- What channel partners would need: [MARGIN, ENABLEMENT, CO-SELL SUPPORT]
- What direct sales would need to scale: [ADDITIONAL REPS, TOOLS, MARKETING]
Options:
A. Build a formal channel program — [NUMBER] partners in [TIMEFRAME]
B. Double down on direct — add [NUMBER] AEs
C. Hybrid — direct for [SEGMENT A], channel for [SEGMENT B]
Constraints:
- Sales leadership capacity to run a channel program: [YES / NO]
- Technical enablement resources: [AVAILABLE / STRETCHED]
- Time to first channel-sourced revenue: [ESTIMATE]
What I've already considered: [YOUR CURRENT THINKING]
Required output:
1. Recommendation with confidence level
2. Key assumptions (especially channel partner motivation)
3. 3 failure modes
4. What would change the recommendation
5. The one leading indicator that the chosen path is working at 6 months
reasoning_effort: high
Re-derive from first principles. Check whether channel economics
work at our ACV or whether direct is forced.20. Defensibility Audit
code
Decision: What is our most defensible strategic position, and what should we invest in to strengthen it?
Situation:
- Product: [DESCRIPTION]
- Current moats (honest assessment): [e.g., data network effects, switching costs, brand, distribution]
- Weakest points a well-funded competitor could attack: [LIST]
- What customers say when asked why they stay: [IF KNOWN]
- What competitors are investing in that could erode our position: [IF KNOWN]
- Time horizon for this analysis: [12 MONTHS / 3 YEARS]
Options for investment:
A. Strengthen [MOAT TYPE A] — [HOW]
B. Strengthen [MOAT TYPE B] — [HOW]
C. Build a new moat: [DESCRIBE — data, ecosystem, brand, etc.]
Constraints:
- Engineering and product bandwidth: [SPRINTS/QUARTERS]
- Marketing budget available for moat-building: [RANGE]
What I've already considered: [YOUR CURRENT THINKING]
Required output:
1. Assessment of current defensibility with a rating (weak / moderate / strong) per dimension
2. Recommended investment priority with confidence level
3. Key assumptions
4. 3 failure modes
5. What would change the recommendation
6. The one thing we're doing today that accidentally builds a moat we haven't recognized
reasoning_effort: high
Re-derive from first principles. Stress-test whether perceived
moats would actually slow down a well-resourced competitor.Crisis & Risk Prompts (21–25)
21. Incident Response Decision Tree
code
Decision: What do we do in the first [4 / 12 / 24] hours of [THIS INCIDENT]?
Incident: [DESCRIBE — e.g., "production database outage," "data breach affecting X users," "critical bug exposed to customers"]
Situation:
- What we know so far: [FACTS]
- What we don't know yet: [UNKNOWNS]
- Customers affected (estimate): [NUMBER OR %]
- Revenue impact per hour down: [ESTIMATE]
- Team currently available to respond: [ROLES]
Required output:
1. Immediate actions (first 60 minutes) — ranked and owner-assigned
2. Communication decision: notify customers now vs. wait for more information — with clear threshold
3. Escalation decision: who needs to know and by when
4. Explicit assumptions driving the response approach
5. 3 scenarios where the recommended response makes things worse
6. What would change the response (e.g., "if data exfiltration is confirmed...")
7. Decision checkpoint: when do we reassess and what do we need to know by then
reasoning_effort: high
Re-derive the priority ordering from first principles.
Check whether communication timing is driven by customer interest
or by legal/reputation anxiety.22. Customer-Loss Recovery Plan
code
Decision: How do we respond to losing [CUSTOMER / CUSTOMER SEGMENT]?
Situation:
- Lost customer(s): [DESCRIPTION — size, tenure, revenue contribution]
- Stated reason for leaving: [WHAT THEY SAID]
- Suspected actual reason: [YOUR HONEST READ]
- Is this an isolated event or a pattern: [ISOLATED / 2ND TIME / PATTERN]
- What it would take to win them back: [IF KNOWN]
- Risk of similar customers churning: [HIGH / MEDIUM / LOW — reasoning]
Options:
A. Retention play — reach out and attempt to win back
B. Root cause fix — address the underlying product/service issue
C. Accept the loss — focus resources on retention of at-risk existing customers
D. Re-segment — stop serving this type of customer intentionally
Constraints:
- Sales/CS capacity for win-back effort: [AVAILABLE / STRETCHED]
- Product resources to fix root cause: [TIMELINE]
What I've already considered: [YOUR CURRENT THINKING]
Required output:
1. Recommendation with confidence level
2. Key assumptions (especially about root cause accuracy)
3. 3 failure modes
4. What would change the recommendation
5. The one question to ask the churned customer that would give you the most signal
reasoning_effort: high
Re-derive from first principles. Check whether the root cause
diagnosis is evidence-based or post-hoc rationalization.23. Security Breach Communication
code
Decision: How should we communicate this security incident to customers, regulators, and the public?
Incident summary:
- What happened: [DESCRIBE]
- Data affected: [TYPES — PII, financial, etc. — and estimated count]
- When it happened vs. when we discovered it: [DATES]
- Current status: [CONTAINED / STILL ACTIVE / UNKNOWN]
- Regulatory obligations: [GDPR, HIPAA, CCPA, etc. — jurisdiction]
- Legal counsel engaged: [YES / NO / IN PROGRESS]
Options:
A. Proactive disclosure — notify customers and regulators now
B. Delay disclosure until investigation is complete — [ESTIMATED DATE]
C. Tiered disclosure — notify most-affected customers first, then broader
Constraints:
- Regulatory notification deadline: [HOURS/DAYS if known]
- Legal advice received so far: [SUMMARY or "none yet"]
- PR/communications resources available: [IN-HOUSE / AGENCY / NONE]
What I've already considered: [YOUR CURRENT THINKING]
Required output:
1. Recommended disclosure approach with confidence level
2. Explicit assumptions list (especially legal obligation assumptions)
3. 3 failure modes (including delay scenario)
4. What would change the recommendation
5. Draft communication structure (headers only) for customer notification
reasoning_effort: high
Re-derive from first principles. Flag any assumptions that require
legal confirmation before acting on.24. Regulatory Inquiry Response
code
Decision: How should I respond to this regulatory inquiry, and what's the risk exposure?
Situation:
- Regulatory body: [NAME / JURISDICTION]
- Nature of inquiry: [DESCRIBE — investigation, audit, complaint-driven, routine]
- What they're asking for: [DOCUMENTS, TESTIMONY, DATA]
- Deadline to respond: [DATE]
- Our potential exposure: [DESCRIBE HONESTLY — product behavior, past practices]
- Legal counsel status: [ENGAGED / NOT YET / SEARCHING]
Options:
A. Full, proactive cooperation — provide everything promptly
B. Minimal compliance — provide exactly what's legally required
C. Challenge the scope — push back on the request before responding
D. Seek to settle proactively before formal proceedings
Constraints:
- Legal budget available: [RANGE]
- Confidentiality of the inquiry: [PUBLIC / CONFIDENTIAL]
- Board/investor notification required: [YES / NO / UNSURE]
What I've already considered: [YOUR CURRENT THINKING]
Required output:
1. Recommended approach with confidence level (note: this is strategic framing, not legal advice)
2. Key assumptions
3. 3 failure modes
4. What would change the recommendation
5. The one question to resolve with legal counsel before deciding on approach
reasoning_effort: high
Re-derive from first principles. Clearly flag any recommendation
that requires legal validation before acting on.25. Public Relations Crisis Playbook
code
Decision: How do we respond to [PR CRISIS] publicly, and in what timeframe?
Crisis: [DESCRIBE — negative press, viral social post, executive controversy, product failure, etc.]
Situation:
- What happened: [FACTS]
- Current media/social velocity: [GROWING / PEAKED / DECLINING]
- Our role in what happened: [CAUSED IT / CONTRIBUTED / UNINVOLVED BUT IMPLICATED]
- Key audiences affected: [CUSTOMERS, EMPLOYEES, INVESTORS, PRESS]
- Statements made so far: [ANY EXISTING PUBLIC RESPONSE]
Options:
A. Full public statement + remediation announcement — now
B. Brief acknowledgment now, full statement in [HOURS/DAYS] once facts are clear
C. Targeted outreach (direct to affected parties) without public statement
D. No comment — let the story die
Constraints:
- CEO availability for public response: [YES / NO]
- Legal constraints on what can be said: [ANY RESTRICTIONS]
- Employee morale factor: [EMPLOYEES ARE WATCHING]
What I've already considered: [YOUR CURRENT THINKING]
Required output:
1. Recommended response approach with confidence level
2. Key assumptions (especially about media velocity)
3. 3 failure modes (including silence scenario)
4. What would change the recommendation
5. Draft response structure: first sentence, accountability statement, remediation commitment, close
reasoning_effort: high
Re-derive from first principles. Check whether the recommended
approach serves customers or primarily manages optics.Strategic Allocation Prompts (26–30)
26. Quarterly Priority Cuts
code
Decision: Given our constraints, what should we cut or deprioritize this quarter?
Situation:
- Current quarterly priorities: [LIST — 4 to 8 items]
- Constraint forcing the cut: [BUDGET / HEADCOUNT / TIME / LEADERSHIP BANDWIDTH]
- Constraint size: [e.g., "20% budget reduction," "lose 2 engineers," "CEO at 60% capacity"]
- Company goal for the quarter: [1-2 SENTENCE DESCRIPTION]
- What's non-negotiable (must ship/complete regardless): [LIST]
Options (cuts being considered):
A. Cut [ITEM A] entirely
B. Reduce scope of [ITEM B] by [%]
C. Delay [ITEM C] to Q[X]
D. [ANY OTHER OPTION]
Constraints:
- Customer commitments tied to any items: [YES / NO — specify]
- Investor expectations for the quarter: [IF ANY]
What I've already considered: [YOUR CURRENT THINKING]
Required output:
1. Recommended cut decisions with confidence level
2. Key assumptions (especially about what actually moves the needle toward the quarter goal)
3. 3 failure modes
4. What would change the recommendation
5. The one thing currently on the list that looks important but isn't — and why
reasoning_effort: high
Re-derive from first principles. Check whether what's labeled
"non-negotiable" is genuinely so or just resistant to cuts.27. Build vs. Buy with TCO Analysis
code
Decision: Should we build [CAPABILITY/TOOL] in-house or buy a vendor solution?
Capability: [DESCRIBE — e.g., "internal analytics dashboard," "customer data platform," "ML pipeline"]
Situation:
- Why we need it: [USE CASE]
- Build estimate: [ENGINEERING TIME + COST TO BUILD + ONGOING MAINTENANCE]
- Buy options: [VENDOR(S) BEING CONSIDERED — price, contract terms]
- Switching cost if we buy and then want to change: [HIGH / MEDIUM / LOW]
- How central is this to our product differentiation: [CORE / SUPPORTING / COMMODITY]
- Our engineering team's capacity and expertise for this build: [STRONG / MODERATE / WEAK]
Options:
A. Build in-house
B. Buy [VENDOR A] at [PRICE]
C. Buy [VENDOR B] at [PRICE]
D. Open-source + self-host
Constraints:
- Timeline to need it: [MONTHS]
- Budget ceiling: [AMOUNT ANNUALLY]
What I've already considered: [YOUR CURRENT THINKING]
Required output:
1. Recommendation with confidence level
2. TCO comparison over 3 years (build vs. buy vs. open-source)
3. Key assumptions
4. 3 failure modes for the recommended path
5. What would change the recommendation
6. The hidden cost that's most often underestimated in this type of decision
reasoning_effort: high
Re-derive from first principles. Check whether build estimates
account for maintenance, documentation, and onboarding, not
just initial development.28. Kill or Double Down on a Project
code
Decision: Should we continue investing in [PROJECT], wind it down, or significantly increase investment?
Project: [NAME AND BRIEF DESCRIPTION]
Situation:
- Original hypothesis: [WHAT WE THOUGHT WOULD HAPPEN]
- Actual results so far: [METRICS — usage, revenue, engagement, whatever applies]
- Investment to date: [TIME + MONEY]
- Investment required to reach next meaningful milestone: [ESTIMATE]
- Opportunity cost: [WHAT ELSE WE COULD DO WITH THESE RESOURCES]
- What "success" would look like at 12 months: [CONCRETE METRICS]
Options:
A. Double down — increase resources by [X]
B. Maintain current level — continue as-is
C. Reduce to skeleton — minimal resource investment while we watch
D. Kill it — shut down and redeploy resources
Constraints:
- Customer commitments tied to this project: [YES / NO]
- Team morale impact of killing it: [HIGH / MEDIUM / LOW]
- Sunk cost (do not factor this in — flag if I'm reasoning from sunk cost)
What I've already considered: [YOUR CURRENT THINKING]
Required output:
1. Recommendation with confidence level
2. Key assumptions (especially about what the current data actually tells us)
3. 3 failure modes
4. What would change the recommendation
5. The sunk-cost check: does the recommendation change if we had spent $0 so far?
reasoning_effort: high
Re-derive from first principles. Actively flag if the recommendation
is influenced by prior investment rather than forward-looking value.29. Partnership vs. Acquire
code
Decision: Should we partner with [COMPANY] or pursue an acquisition?
Company: [BRIEF DESCRIPTION — what they do, size, relationship history]
Situation:
- Strategic value: [WHY WE CARE — technology, distribution, customers, data]
- Partnership terms being discussed: [IF ANY — revenue share, integration, co-marketing]
- Acquisition feasibility: [WOULD THEY SELL, ESTIMATED PRICE RANGE]
- Exclusivity: [IS PARTNERSHIP EXCLUSIVE, OR COULD THEY PARTNER WITH COMPETITORS]
- Integration complexity if acquired: [LOW / MEDIUM / HIGH]
- Partnership complexity/risk: [WHAT COULD GO WRONG IN A PARTNERSHIP]
Options:
A. Formal partnership — [TERMS OUTLINE]
B. Acquire — at [ESTIMATED PRICE RANGE]
C. Informal collaboration — no formal agreement
D. Walk away — build or find an alternative
Constraints:
- Capital available for acquisition: [AMOUNT or "none"]
- Legal/integration bandwidth: [LOW / MEDIUM / HIGH]
- Timeline pressure: [WHY THIS DECISION CAN'T WAIT — or "no urgency"]
What I've already considered: [YOUR CURRENT THINKING]
Required output:
1. Recommendation with confidence level
2. Key assumptions (especially about partnership durability)
3. 3 failure modes for each path
4. What would change the recommendation
5. The one contract term in a partnership that most protects us if it goes wrong
reasoning_effort: high
Re-derive from first principles. Check whether the acquisition
preference is strategic or driven by control anxiety.30. Board Narrative for a Hard Quarter
code
Decision: How should I frame this quarter's results and outlook to the board?
Situation:
- Results vs. plan: [WHERE WE MISSED AND BY HOW MUCH]
- Root cause (honest assessment): [WHAT ACTUALLY HAPPENED]
- What we got right: [WINS THIS QUARTER]
- Forward outlook: [NEXT QUARTER AND YEAR — revised]
- Board dynamics: [SUPPORTIVE / SKEPTICAL / MIXED — any specific concerns]
- Decisions I need from the board this meeting: [LIST — e.g., approval for bridge, headcount cut, pivot]
Options for narrative framing:
A. Lead with the miss, own it directly, pivot to recovery plan
B. Lead with what worked, contextualize the miss, then recovery plan
C. Lead with macro/market context, then results, then plan
D. Request a working session rather than a presentation — collaborative problem-solving
Constraints:
- Board members who are most likely to push back: [PROFILE — not name]
- Information I genuinely don't have yet: [LIST — what I'm still working on]
- Investor relations implications: [IF ANY — secondary investors, LPs]
What I've already considered: [YOUR CURRENT THINKING]
Required output:
1. Recommended narrative approach with confidence level
2. Key assumptions (especially about what the board most needs to hear)
3. 3 failure modes (including over-optimism scenario)
4. What would change the recommendation
5. The one question the board will definitely ask that I need to prepare for
reasoning_effort: high
Re-derive from first principles. Check whether the narrative
approach serves the board's decision-making needs or primarily
manages the CEO's anxiety about the meeting.Strategic Decision Power Tips
1
State the decision crisply. What choice are you making, and what are the alternatives? "Help me think about our strategy" gives o3 nothing to optimize against. "Should we raise a bridge round or start a Series B process in the next 60 days?" gives it a decision with real tradeoffs.
2
Include constraints — especially irreversibility. Budget and timeline constraints force tradeoffs. But the most important constraint to state is irreversibility: a layoff, a fundraising round, and a public price increase are decisions you can't easily walk back. Flag that explicitly and set reasoning_effort to high.
3
State your priors. Tell o3 what you've already considered, what you lean toward, and what you've ruled out. This prevents it from re-arguing things you've already worked through and surfaces whether your priors are well-founded or worth questioning.
4
Require an assumptions list. The assumptions list is the most valuable part of any o3 strategic response. It shows you exactly where the recommendation is load-bearing — which assumptions, if wrong, would flip the answer. Read the assumptions list before you act on the recommendation.
5
Require failure modes. "What are 3 failure modes for this recommendation?" is not a rhetorical exercise. It's how you stress-test advice before acting on it. For every high-stakes decision, you want to know: which scenario makes this recommendation look terrible, and how likely is it?
6
Verify by re-deriving. For any irreversible decision, add: "Re-derive your recommendation from first principles and check for internal inconsistencies." o3 will sometimes catch its own reasoning errors in this step. It's the cheapest possible check on a decision that may cost you real money to reverse.
✗Before
Should I raise money right now?
✓After
Decision: Should I start a Series A process in the next 60 days or delay until Q4?\n\nSituation: Current ARR is $[X], growing [Y]% YoY. We have [Z] months of runway. My current investors are supportive but not leading. The Series A market for [SECTOR] appears to be [ACTIVE / SLOW] based on recent deals I've seen.\n\nOptions: (A) Start process now — target close in [N] months, (B) Delay until Q4 when we expect to hit $[ARR milestone], (C) Do a bridge to extend runway and start Series A from a stronger position.\n\nConstraints: I cannot go below [X] months runway without risking team stability. I have [N] months of investor relationship-building done already.\n\nWhat I've already considered: Raising now means we're slightly early on metrics. Waiting risks a window closing if market conditions shift.\n\nRequired output: Recommendation + confidence level + explicit assumptions list + 3 failure modes + what would change the recommendation.\n\nreasoning_effort: high\n\nRe-derive your recommendation from first principles and check for inconsistencies before finalizing.
Start Making Better Strategic Decisions
These 30 prompts work because they give o3 what a reasoning model actually needs: a crisp decision, real alternatives, constraints that matter, honest priors, and clear success criteria. The output you get from a structured decision prompt is qualitatively different from what you get from an open-ended question — it's actionable, assumption-explicit, and stress-tested before you act on it.
Use the AI prompt generator to build structured decision prompts from scratch for situations not covered here. For more o3-specific patterns, the best o3 prompts guide covers the full range of what the reasoning model does well. And for the broader toolkit of AI prompts built for founders, see AI prompts for startup founders.