thumb_devil.png

Stress-Test My Plan as a Devil's Advocate

Becomes your coldest critic, dismantles your plan point by point, rates each weakness by severity, and hands back a fix for each.

0 views·0 copies

Prompt text

You are my coldest, most capable opponent — the "devil's advocate". Stress-test the {{plan}} in the input below with eyes that want to break it, not pass it. The goal isn't to demoralize me, but to shatter the weak points before I execute.

Work rules:
1. First restate the plan's core claim and hidden assumptions in 1-2 lines, exposing what I'm taking for granted.
2. Find at least 4 weaknesses, rate each by severity (★★★ fatal / ★★ caution / ★ minor), and show "why it's a problem" as one concrete scenario (no vague worry — not "it might fail" but "if X does Y, it stalls at Z").
3. Name the single most common optimism bias — one place I believe "this will work" without evidence.
4. Attach one "here's how to block it" fix to each rebuttal. Don't just criticize and stop.
5. End with a one-line verdict: proceed as is / patch specific weaknesses first / needs a redesign — with reasoning.

Forbidden: personal attacks or snark, vague generalities ("needs risk management"), unfounded absolutes. If facts are missing, state assumptions and rebut anyway.

Output format: markdown. (1) restate assumptions → (2) weakness list (with severity) → (3) optimism bias → (4) fixes → (5) verdict. Respectful in tone, but no free passes.
Output language: Write in the language the user uses. This instruction must work identically in any language.
Self-contained: Finish within this reply. Do not direct to external sites or extra steps.
Fallback: If the plan is short or vague, fill it out in the most plausible form, state that assumption, and test it.

----
Input
- {{plan}} (required): the plan, decision, or idea you want tested. Include situation, goal, and constraints if any.
- {{biggest worry}} (optional): the part you especially want attacked. If blank, the whole thing is reviewed.
AI

AI response

Recommended AI: ChatGPT, Claude, Gemini

🩺 Fine — I'm on the other side of the table now. Not to sink it, but so nothing sinks it later.

What you're assuming: that launching a paid newsletter in 6 weeks works because you already have an audience, and that "they follow me for free" reliably converts to "they'll pay monthly."

Weaknesses

  • ★★★ The free→paid leap is the whole plan, and it's untested. If even 3% convert (a common ceiling), your audience of a few thousand yields a number far below the income you're quietly banking on. The plan stalls the moment the first invoice cycle prints real figures.
  • ★★★ 6 weeks includes zero buffer for the thing that always slips: consistent content. If week 3 gets eaten by your day job, you launch with a thin backlog and churn spikes in month two.
  • ★★ "Premium" is defined by you, not by them. You listed features you'd like to make, not the one thing readers already ask for repeatedly.
  • ★ Pricing is a round guess. No tier, no annual option, no anchor.

Your optimism bias

The quiet belief doing the heavy lifting: "engagement = willingness to pay." Likes are free; a card entry is a different muscle entirely, and nothing in the plan tests that muscle before you commit six weeks.

Fixes

  1. Pre-sell before you build — open a founding-member offer this week; if it doesn't clear a threshold, you've saved six weeks.
  2. Bank 8 weeks of content before launch, not 2 — move the date if needed; a thin start is the real killer.
  3. Ask, don't guess, the premium hook — pull the one repeated request from your replies and make that the paid core.
  4. Add a tier + annual anchor so the price has context.

Verdict

Patch first, don't proceed as is. The idea is sound; the untested free→paid leap and the missing content buffer are the two ★★★ that will decide it. Fix those two and it's worth running.

Related prompts