Each prompt below is designed to feed into Claude or ChatGPT. Run them in order — outputs of earlier prompts feed into later ones. Don't skip the steel-mans.

1. Structure shortlist. "My M2 verdict and ICP are: [paste]. From the 7 standard revenue structures (one-off transaction, subscription, cohort/seasonal, retainer, productised service, marketplace fee, hybrid), shortlist the 3 that best match this buyer's purchase psychology and cash rhythm. For each: why it fits, the cash-flow shape over 12 months, and the operational load on me as a solo founder. Be specific — no theory."
2. Steel-man each structure. "For each of the 3 shortlisted structures above, write the strongest argument against it for my specific case. What breaks at 10 customers? At 100? Where does the model trap me? Which one is the founder choosing because it's familiar vs the one the buyer actually wants?"
3. Willingness-to-pay reread. "Here are my 15 M2 interview synthesis sheets [paste]. Extract every signal about price: (a) what they currently pay for workarounds and substitutes, (b) what budget category this would come from (line item, discretionary, personal), (c) verbatim phrases about cost ('cheap', 'expensive', 'no-brainer at $X'). Rank by signal strength. Reject anything that's a future-intent guess, not past-behaviour evidence."
4. Price ladder. "Given the willingness-to-pay synthesis [paste] and my offer [paste], draft three price points: a low anchor (entry to capture volume), a target (where I want most buyers), and a high anchor (premium that protects perceived value of the target). For each: who buys, what they get extra, what's deliberately missing. Show the trade-off, don't just inflate."

   Live example: see how I architected the three EXITR tiers (Toolkit, Accelerator, Founder Lab) on the pricing page. I teach what I did this week.

5. Competitor pricing scan. "Find me 10 competitors and adjacent alternatives for my ICP [paste]. For each, document: visible price, what's included, what the upsell path looks like, and the implicit positioning (premium, accessible, freemium, value). Output a 10-row table. Highlight the gap nobody is pricing into."
6. Unit economics. "My target price is $X. Walk me through the unit economics: cost of delivery (my time × hourly cost, plus tools/people), gross margin per unit, customer acquisition cost assumption (use my realistic channel mix [paste]), payback period. State every assumption explicitly. If the assumption is a guess, label it [GUESS]. If gross margin is under 60%, flag what specifically I should cut or repackage."
7. Breakeven and runway. "My personal monthly runway is $X [paste]. My target price is $Y, gross margin Z%. Calculate: how many customers/month I need to cover personal runway alone, to cover that plus reinvestment, and to pay myself $10K/month. Show the curve for the first 12 months under two scenarios — slow ramp (10% growth/mo) and fast (25%/mo). Tell me which feels honest given my channel reality."
8. Cash-flow stress test. "Given the revenue structure [paste] and the breakeven curve above, simulate what happens to my cash position in months 1-12 under each: (a) hits 60% of forecast, (b) hits forecast but pays in 60-day terms not 30, (c) churns 20% of recurring customers in month 4. Show the lowest cash point and when. If I'd go negative on any scenario, name the structural fix — not 'find more customers'."
9. Pricing-page reverse engineer. "Here are my three price points [paste prompt 4 output] and my unit economics [paste prompt 6]. Write the pricing page section copy. For each tier: who it's for (one sentence), the outcome (not the features), the deliberate omission ('this is not for you if...'), and the buying-friction (one-click / application / call). No 'most popular' badge unless I tell you which one and why. Brand voice: direct, founder, no hype."
10. The 6-month verdict gate. "Given the model I just built, give me the three numbers I should commit to hitting by end of month 6 — beyond which, if I haven't hit them, the model itself is wrong, not just the marketing. Make the numbers narrow enough to actually fail. Then write the 'pivot the model' question I should ask myself if any of the three misses by more than 25%."