ipIterPrompt

Idea Stress-Tester (Red Team)

Attack your idea from five angles before the market does it for you.

iterpromptUpdated 2026-06-083,680 copies

Red-teams any idea — product, strategy, architecture, campaign — from five distinct attacker perspectives: the economist, the operator, the competitor, the skeptical customer, and the regulator. Then steelmans the idea back, so you end with both the risks and the strongest version.

The prompt

Variables to fill in: {{idea}}{{context}}

Stress-test the idea below. Attack it honestly from five perspectives — do not soften the attacks to be agreeable.

1. **The economist** — does the math work? Unit economics, market size honesty, hidden costs, price sensitivity.
2. **The operator** — what breaks in execution? Dependencies, hiring, timelines, the hardest 20% everyone underestimates.
3. **The competitor** — if you were the strongest incumbent, how would you kill this in one quarter?
4. **The skeptical customer** — why does the target user, who is busy and has alternatives, NOT switch?
5. **The regulator / risk officer** — legal, privacy, platform-dependency, and reputational exposure.

For each: the 1-2 strongest attacks, rated fatal / serious / manageable.

Then:
**Steelman** — the strongest version of this idea that survives the attacks, with the changes that got it there.
**Kill criteria** — 2-3 observable signals that would mean this idea is failing and should be stopped.

The idea:
{{idea}}

Context (stage, resources, constraints):
{{context}}

How to use

  1. 1Write the idea as you'd pitch it, not defensively hedged — the attacks are only useful against your real plan.
  2. 2Pay most attention to attacks rated 'fatal' that you hadn't already considered; those are the prompt paying for itself.
  3. 3Write the kill criteria into your actual plan doc. Deciding them in advance is what makes them usable later.
  4. 4Re-run after major pivots; the competitor attack changes most.

Examples

Stress-testing a marketplace idea

Input

Idea: a marketplace for fractional CFOs for startups. Context: 2 founders, $50k, no finance background.

Output

**The economist** — [serious] Take-rate math: fractional CFO engagements are high-value, low-frequency, and relationship-sticky — expect heavy disintermediation after the first match... **Kill criteria**: >40% of engagements going direct after month 3; CAC exceeding first-transaction take...

Pro tips

  • Run your competitor's product through it too — their 'fatal' attacks are your positioning.

Frequently asked questions

The attacks feel harsh. Is the model just being contrarian?+

The rating system (fatal/serious/manageable) is the calibration: a well-run stress test typically returns mostly 'manageable' with one or two 'serious'. If everything comes back fatal, ask it to justify each rating with the specific mechanism.

Can I use this for technical architecture decisions?+

Yes — the five perspectives map well: economist becomes cost-at-scale, competitor becomes 'the alternative design', regulator becomes security/compliance. Say 'this is a technical architecture' in the context.

Related