One-sentence positioning: Use 4 steps to systematically bypass the 4 innate bugs of human decision-making — too narrow a frame, confirmation bias, short-term emotion, overconfidence.
When to use (TRIGGER)
- The user keeps going back and forth between two options ("should I quit my job," "React or Vue")
- The decision carries strong emotions (excitement, anxiety, urgency)
- The user already "has the answer in their heart" but is still looking for reasons to convince themselves (a signal of confirmation bias)
- Time still allows (at least a few hours to a few days), it's not firefighting-style emergency response
When not to use (ANTI-PATTERN)
- The decision is reversible and extremely low-cost (just try it, don't run a process)
- A true time-critical moment (must press before the elevator doors close), use OODA Loop
- The decision belongs to the Clear/Obvious quadrant of Cynefin (if there's a best practice, use it)
- The options are already locked down by external constraints (compliance/resources leave no choice)
4 Steps (STEPS)
W — Widen Your Options
Humans tend to simplify choices into "Yes/No" or "A/B." First force expansion:
- Ask yourself the "vanishing options test": if all current options disappeared, what would I do?
- Find an "opportunity cost" option: what else could I do with the same time/money/energy?
- Find an "AND rather than OR" solution: can I do both at the same time? Do them in phases?
- Find people who have made similar decisions and look at their option menu (outside view）
R — Reality-Test Your Assumptions
The answer intuition gives is often based on unverified assumptions. Attack the assumptions with evidence:
- Deliberately look for disconfirming evidence(to fight confirmation bias): list 3 facts that would overturn the current leaning
- Shrink the experiment: can a small-scale 1-week/1-month probe replace going all in? (Heath: "ooch" tentative small steps)
- Look from a distance: how did people in the same industry/same situation do it? What is the base rate?
A — Attain Distance Before Deciding
Short-term emotions overestimate immediate pressure and underestimate long-term impact. Use tools to create distance:
- Apply the 10/10/10 Rule: how will I view this decision in 10 minutes / 10 months / 10 years?
- Ask "If it were my best friend facing the same situation, what would I advise them to do?」
- Identify your "core priorities": what are the 3 things I value most in the next 5 years, and does this decision serve them?
P — Prepare to Be Wrong
The future is more uncertain than you think. A contingency plan isn't pessimism, it's responsibility:
- Do a Pre-Mortem: assume this decision fails 6 months from now, what are the most likely reasons?
- Set a "tripwire": what metric/event should make me re-evaluate?
- Think through the "escape route": if it's wrong, what's the fastest way out and the maximum loss?
Decision Checklist
- I listed at least 3 options (not 2)
- I proactively found ≥ 2 pieces of evidence contrary to the current inclination
- I applied at least one distancing method
- I set specific tripwire triggers and alerts
- I can state the maximum loss if it fails within 1 minute
Real Cases
Case 1: Whether to accept a new job offer
- W: Besides "accept / not accept," there are also "negotiate salary," "request a 2-week delayed start for a trial first," "simultaneously negotiate a counter-offer with the current company," "first do a 1-month consulting trial collaboration" — suddenly from 2 options to 5 options
- R: Talk with 2 former employees of the company who left within 3 years
- A: Looking back 10 years from now, what kind of person do I hope to become, and does this job push me toward that or push me away?
- P: tripwire = if within 90 days of joining the manager changes or the core product direction changes, immediately reassess
Case 2: Intel 1985 abandoning the memory chip business
Grove & Moore struggled amid huge losses over whether to cut the DRAM business. Grove used a distancing technique — asking Moore: "If the board replaced the two of us, what would the new CEO do first upon taking office?" Moore answered: "Cut DRAM." Grove: "Then why don't we cut it ourselves?" — a classic A (Attain Distance) application, and the result created Intel's transformation to CPUs.
Relationship with Other Frameworks
- WRAP includes pre-mortem as a sub-technique of the P step
- WRAP includes ten-ten-ten as a sub-technique of the A step
- WRAP and ooda-loop are complementary: WRAP is suitable for slow decisions (hours-days), OODA is suitable for fast decisions (seconds-minutes)
- Use cynefin to classify before deciding: Only the Complex/Complicated quadrants are worth running WRAP on
Common Pitfalls
- Going through the motions: The 4 steps become a form of self-comfort, and in reality you still decide by intuition — countermeasure: write down the output of each step, and do not allow "I thought about it"
- Only doing W and not P: After broadening the options you think you are done, without preparing for mistakes — countermeasure: tripwire is a hard requirement
- The "friend" you find when distancing is an echo chamber: Advice is filtered by preference — countermeasure: find someone who will oppose you, not someone who supports you
References
- Chip Heath & Dan Heath, 《Decisive》Crown Business 2013
- HBR summary: "The Process of Better Decision Making" 2013-03