decisive

name	decisive
description	Use when making important decisions, evaluating options, avoiding decision traps, or designing decision processes for teams. Covers the WRAP framework (Widen options, Reality-test, Attain distance, Prepare to be wrong), the four villains of decision-making, and structured methods for overcoming narrow framing, confirmation bias, short-term emotion, and overconfidence. Use when the user mentions "decision making", "should I", "which option", "narrow framing", "confirmation bias", "overconfidence", "pros and cons", "stuck deciding", "decision process", "premortem", "tripwire", or "base rates". For negotiation tactics, see negotiation. For product positioning decisions, see obviously-awesome.
license	MIT
metadata	{"author":"avery","source":"Decisive by Chip Heath & Dan Heath","version":"1.0.0"}

Decisive

Framework for making better choices by counteracting four cognitive villains that corrupt every stage of the natural decision process. Based on the Heath brothers' central finding: process matters more than analysis by a factor of six (Lovallo/Sibony study of 1,048 business decisions). The WRAP framework provides one countermeasure per villain.

Core Principle

The quality of the decision process determines the quality of the outcome -- not the brilliance of the analysis.

The foundation: Standard decision-making follows four steps: encounter a choice, analyze options, make a choice, live with it. Each step is corrupted by a specific cognitive villain: narrow framing, confirmation bias, short-term emotion, and overconfidence. The WRAP process wraps around your natural process to protect against each villain. "Whether or not" decisions fail 52% of the time; decisions with two or more alternatives fail only 32%. Multi-alternative decisions produce 6x more "very good" outcomes (40% vs. 6%).

Scoring

Goal: 10/10. When reviewing or designing a decision process, rate it 0-10 based on adherence to the four WRAP steps below. A 10/10 means options are genuinely widened, assumptions are reality-tested with outside-view data, distance is achieved before deciding, and failure/success scenarios are pre-planned; lower scores indicate vulnerability to one or more villains. Always provide the current score and specific improvements needed to reach 10/10.

The WRAP Framework

Each step counters one villain of decision-making.

Step	Villain It Counters	Core Question
W -- Widen Your Options	Narrow Framing	How can I expand my choice set?
R -- Reality-Test Your Assumptions	Confirmation Bias	How can I collect trustworthy, disconfirming information?
A -- Attain Distance Before Deciding	Short-Term Emotion	How can I overcome visceral feelings for my long-term values?
P -- Prepare to Be Wrong	Overconfidence	How can I plan for an uncertain future?

---

1. Widen Your Options (W)

Core concept: Most decisions are too narrow by default. "Whether or not" framing reduces complex choices to a single binary. The spotlight effect -- attending to what's visible and neglecting everything outside it -- causes us to miss better alternatives. The goal is to generate more and better options before analyzing.

Why it works: Paul Nutt's 30-year study: only 29% of organizations considered more than one alternative. Multi-alternative decisions produced 6x more "very good" outcomes. Multitracking (considering options simultaneously) doesn't slow decisions -- Eisenhardt's Silicon Valley research found it actually speeds them up because leaders understand the landscape and have fallback plans.

Key insights:

• Vanishing Options Test -- "You cannot choose any of your current options. What else could you do?" Forces the spotlight to move.
• AND Not OR -- "Is there a way I can do this AND that?" Rejects binary framing.
• Multitracking -- develop multiple alternatives simultaneously. Benefits: better landscape understanding, undercuts politics, built-in fallback. Banner-ad study: simultaneous designers produced superior work and felt less criticized.
• Opportunity Cost Prompt -- "What am I giving up by choosing this?" Reminder of alternatives increased pass rates from 25% to 45%.
• Bright Spots -- look internally for what's already working. Kaiser Permanente: one hospital had already solved the sepsis problem; scaling it reduced mortality 28% below the national average.
• Laddering Up -- extract the crucial features of your problem and search for analogies in distant domains. Speedo's Fairhurst: redefined "faster swimsuit" as "anything that moves fast" -- found that sharkskin roughness reduces drag. 83% of swimming medals at 2000 Sydney Olympics.
• Playlist -- a pre-built list of stimuli questions for recurring decision types. Not a checklist (stops errors) -- a playlist (opens possibilities).

Product applications:

Context	Application	Example
Binary decision	Apply AND Not OR	"Should we hire or not?" becomes "Could we hire part-time AND contract?"
Stuck brainstorming	Run Vanishing Options Test	Remove all current ideas; what would you try if none of them existed?
Strategy planning	Multitrack 2-3 approaches	Run parallel experiments rather than betting on one direction
Budget cuts	Use bright spots	Find the department already doing more with less; scale their approach
Innovation	Ladder up for distant analogies	Redefine the problem at a higher abstraction; who else has solved it?
Recurring decisions	Build a playlist	Pre-built questions that force broader thinking each time

Copy patterns:

• "What else could you do if none of these options existed?"
• "Is there a way to do this AND that?"
• "What are you giving up by choosing this?"
• "Who has already solved this problem?"
• "Fall in love at least twice before deciding."

Ethical boundary: Multitracking should produce real alternatives, not sham options. If the team easily agrees on which option is best, you may have sham options. Genuine disagreement is a sign of real choice.

See: references/widen-options.md for the full Vanishing Options Test, multitracking implementation, bright spots methodology, and laddering exercises.

---

2. Reality-Test Your Assumptions (R)

Core concept: Confirmation bias makes us seek information that confirms what we already believe. Even rigorous-looking data collection can be biased. The antidote: actively design ways to collect trustworthy, disconfirming information -- and separate the inside view (your optimistic assessment) from the outside view (base rates of how things like this usually go).

Why it works: Meta-analysis of 91 studies (8,000+ participants): people are more than twice as likely to favor confirming over disconfirming information. Tetlock's 20-year study: 284 expert forecasters making 82,361 predictions performed worse than simple base-rate extrapolation. Experts are reliable on base rates but unreliable on predictions.

Key insights:

• Consider the Opposite -- deliberately seek disconfirming arguments and evidence.
• "What Would Have to Be True?" (Roger Martin) -- shifts adversaries to collaborators. "What would have to be true for this option to be the very best choice?" Resolved 3-hour deadlock at Inmet Mining.
• Disconfirming Questions -- "What problems does it have?" (89% disclosure) vs. "It doesn't have any problems, does it?" (61%) vs. "What can you tell me about it?" (8%).
• Inside View vs. Outside View -- inside view is your optimistic self-assessment. Outside view is the base rates. Trust the outside view. Kahneman's curriculum group predicted 1.5-2.5 years; base rates said 7-10 years; it took 8.
• Base Rates -- aggregate data on how situations like yours turn out. Trust them like hotel reviews. Ask experts for frequencies, not predictions.
• Zoom Out, Zoom In -- outside view (base rates) + close-up investigation (texture that numbers miss). Both together.
• Ooch -- construct small experiments to test hypotheses before committing. CarsDirect: built a fake website, sold 3 cars overnight, founded the company. But don't ooch when you already know what's right and are just delaying.

Product applications:

Context	Application	Example
Team disagreement	Ask "What would have to be true?"	Transforms argument into collaborative assumption-testing
Evaluating a candidate	Use disconfirming questions	"Who were the last three to leave? What are they doing now?"
Market assessment	Zoom out (base rates) + zoom in (real users)	Industry failure rate + direct customer observation
Business plan confidence	Apply outside view	"What percentage of businesses like this succeed?"
New feature decision	Ooch before committing	Build a fake door test or concierge version before full development
Expert consultation	Ask for base rates, not predictions	"What percentage of cases like this succeed?" not "Do you think I'll win?"

Copy patterns:

• "What would have to be true for this to be the best choice?"
• "What problems does this have?"
• "What do the base rates say?"
• "Who has already tried this, and what happened?"
• "Let's test it before we commit."
• "What's the most likely way I could fail to get the right information here?"

Ethical boundary: Deliberate mistakes (Schoemaker) -- intentionally testing assumptions you've never questioned -- are a legitimate reality-testing tool. But overt deception of customers or stakeholders to "test" their reactions crosses the line. Reality-test with honest probing, not manipulation.

See: references/reality-test.md for the inside/outside view framework, disconfirming question templates, ooching methodology, and base-rate collection exercises.

---

3. Attain Distance Before Deciding (A)

Core concept: Short-term emotion is not the enemy -- visceral feelings contain real information. The problem is when they become the only voice at the table, crowding out long-term values. Loss aversion (2-4x stronger than equivalent gains), mere exposure (familiarity breeds preference), and status-quo bias conspire to make the present feel more important than the future.

Why it works: Study: 66% of students chose Job B for themselves; 83% recommended Job B for their best friend. Distance yields clarity because it naturally prioritizes the most important factor and downplays short-term emotion. Andy Grove's "What would our successors do?" broke Intel's paralysis about exiting memory chips -- the outsider perspective strips historical and political weight.

Key insights:

• 10/10/10 (Suzy Welch) -- "How will I feel about this 10 minutes from now? 10 months? 10 years?" Levels the emotional playing field between present intensity and future fuzziness.
• "What would I tell my best friend to do?" -- the single most effective question for personal decisions. We see the forest for friends; we get stuck in the trees for ourselves.
• "What would my successor do?" -- strips emotional attachment to history. Grove to Moore: "If the board brought in a new CEO, what would he do?" Moore: "Get us out of memories." Grove: "Why shouldn't we do it ourselves?"
• Core Priorities -- when distancing doesn't resolve a decision, it signals a conflict among long-term values. Name the conflict explicitly.
• Stop-Doing List (Jim Collins) -- urgencies always crowd out priorities. MIT study: managers' weekly activities had zero connection to their stated core priorities.
• Mere Exposure Principle -- repetition breeds comfort. "Institutional truths" may only feel true because they've been repeated, not proven. Three exposures to a false statement increased its rated truthfulness.

Product applications:

Context	Application	Example
Personal career decision	Run 10/10/10	"10 minutes: terrified. 10 months: proud. 10 years: won't matter."
Organizational inertia	Ask "What would our successors do?"	Strips the weight of sunk costs and political history
Competing priorities	Name core values explicitly	"Is the customer the volunteer surgeon or the patient?" settled years of board conflict
Time allocation	Build a stop-doing list	What would you stop doing if you had 10 years to live but $20M?
Status quo trap	Identify mere exposure bias	"Do we believe this because it's true or because we've said it 100 times?"
Hourly focus	Bregman's hourly beep	Timer every hour: "Am I doing what I most need to be doing right now?"

Copy patterns:

• "How will I feel about this in 10 months?"
• "What would I tell my best friend to do?"
• "If we got kicked out and the board brought in someone new, what would they do?"
• "What would I stop doing?"
• "Do we believe this because it's true or because we've repeated it?"

Ethical boundary: Distancing techniques are for gaining clarity, not for suppressing legitimate emotion. If every analysis says "proceed" but your gut screams "wrong," the gut may be detecting a pattern your conscious mind can't articulate. A "leemer" (pilot term for a vague sense something isn't right) deserves investigation, not dismissal.

See: references/attain-distance.md for the 10/10/10 framework, core priorities identification, stop-doing list methodology, and status-quo bias exercises.

---

4. Prepare to Be Wrong (P)

Core concept: The future is a range, not a point. Overconfidence makes us treat it as a point. When experts are "completely certain," they're wrong 40% of the time. Counteract by bookending possibilities (lower and upper), running premortems and preparades, setting tripwires to force reconsideration, and building safety factors.

Why it works: Soll/Klayman: when participants considered bookends explicitly, their prediction range accuracy improved from 45% to 96% of optimal. Prospective hindsight ("It happened. Why?") generates 25% more reasons than speculation ("Could it happen?"). LPCH rapid-response tripwires: code calls dropped 71%, hospital mortality dropped 18%, 33 lives saved in 18 months.

Key insights:

• Bookending -- estimate a lower bookend (dire scenario) and upper bookend (rosy scenario) instead of a single prediction. Invest when price is near the lower bookend with large upside.
• Premortem (Gary Klein) -- "It's 12 months from now. The project failed. What killed it?" Everyone writes reasons independently, then shares one at a time. Surfaces threats optimism suppresses.
• Preparade -- the upside premortem. "It succeeded wildly. Are we ready?" Softsoap locked up the world's entire pump supply before their national launch.
• Safety Factor -- compute the requirement, then multiply by a buffer. Elevator cables: 11x. Software schedules: +30-50%.
• Tripwires -- pre-defined signals that snap you out of autopilot. Types: dates/deadlines (without them, completion drops from 66% to 25%), budgets/thresholds, partitions (individually wrapped cookies last 24 days vs. 6 days unwrapped), and pattern recognition ("Call if you are worried about a patient").
• Realistic Job Preview -- expose people to the warts before they start. Meta-analysis of 40 studies: consistently reduces turnover through "vaccination" against shock.
• Procedural Justice -- people care about whether the process was fair, not just the outcome. Let people be heard, use accurate information, apply principles consistently, explain why.

Product applications:

Context	Application	Example
Project planning	Run a premortem	"It's one year from now and we failed utterly. Why?"
Launch preparation	Run a preparade	"We succeeded beyond expectations. Are we ready for the demand?"
Investment decision	Use bookending	Lower and upper scenarios; invest only near the lower bookend
Schedule estimation	Apply safety factor	Add 30-50% buffer to your best estimate
Ongoing projects	Set tripwires	"If we haven't reached X by date Y, we stop and reassess"
Hiring	Use realistic job preview	Show candidates the hardest parts before they accept

Copy patterns:

• "It's one year from now and we failed. What killed us?"
• "It's one year from now and we're heroes. Are we ready?"
• "What's our tripwire? When will we know to reconsider?"
• "What's the worst realistic scenario? The best?"
• "What's our safety factor?"
• "Sometimes the best way to defend a decision is to point out its flaws."

Ethical boundary: Tripwires are designed to enable boldness, not caution. Setting clear boundaries (budget cap, time limit, trigger condition) creates psychological permission to commit fully until that boundary is reached. Using tripwires to avoid commitment entirely defeats their purpose.

See: references/prepare-to-be-wrong.md for premortem/preparade facilitation guides, tripwire design templates, bookending methodology, and safety factor calibration.

---

Process Beats Analysis

The book's single most important empirical finding, applicable to every decision domain.

Factor	Impact on Decision Success
Process quality (WRAP adherence)	6x more influence than analysis quality
Analysis quality (depth of data, sophistication of models)	1x baseline influence

This means: a rough analysis with an excellent process outperforms a brilliant analysis with a poor process by a wide margin.

Common Mistakes

Mistake	Why It Fails	Fix
"Whether or not" framing	Reduces options to a binary; fails 52% of the time	Apply Vanishing Options Test; generate at least two real alternatives
Seeking confirming data	2x bias toward confirming information	Ask "What problems does this have?"; seek the opposite
Trusting expert predictions	Experts do worse than base-rate extrapolation	Ask experts for base rates, not predictions
Deciding in emotional heat	Short-term intensity crowds out long-term values	Run 10/10/10; ask "What would my best friend say?"
Single-point prediction	When "completely certain," wrong 40% of the time	Use bookends (lower and upper scenarios)
No tripwires	Autopilot leads to escalation of commitment	Set dates, budgets, and pattern triggers before starting
Sham options	Easy consensus may mean only one real option	Test: if your team disagrees about which is best, you have real options
Ooching when you should commit	Using experiments as emotional delay tactics	If you already know, commit. Ooch only when genuinely uncertain.

Quick Diagnostic

Question	If No...	Action
Are you considering at least two genuine alternatives?	Narrow framing -- "whether or not" trap	Run Vanishing Options Test or AND Not OR
Have you sought disconfirming information?	Confirmation bias -- "cooking the books"	Ask "What problems does this have?" and "What would have to be true?"
Do you know the base rates for situations like this?	Inside view dominating -- optimistic self-assessment	Find the outside view: "What percentage of cases like this succeed?"
Have you run a small experiment before committing?	Predicting instead of testing	Ooch: build a cheap test of your core assumption
Have you applied 10/10/10 or "best friend" test?	Short-term emotion may be distorting	Run the distance test before finalizing
Do you have a premortem documented?	Optimism is suppressing risk awareness	"It failed. Why?" -- everyone writes independently
Have you set a tripwire?	No mechanism to force reconsideration	Define the date, budget, or trigger that would make you reassess
Would a new CEO make the same decision?	Sunk cost and politics are distorting	Ask "What would our successors do?"

Reference Files

• widen-options.md: Vanishing Options Test, AND Not OR, multitracking, bright spots, laddering, playlist construction
• reality-test.md: Inside/outside view, base rates, disconfirming questions, ooching, "What would have to be true?"
• attain-distance.md: 10/10/10, best friend test, successor test, core priorities, stop-doing list, mere exposure
• prepare-to-be-wrong.md: Premortem, preparade, bookending, tripwires, safety factors, realistic job preview, procedural justice

About the Author

Chip Heath is a professor at Stanford Graduate School of Business. Dan Heath is a senior fellow at Duke University's CASE center. Together they are the authors of Made to Stick, Switch, Decisive, and The Power of Moments -- all New York Times bestsellers. Their work synthesizes behavioral science research into practical frameworks for business and personal decision-making. Decisive distills findings from decision science, psychology, and organizational behavior into the four-step WRAP process.

Further Reading

This skill is based on the Heath brothers' research into decision-making. For the complete case studies, research citations, and deeper frameworks:

• "Decisive: How to Make Better Choices in Life and Work" by Chip Heath and Dan Heath
• "Made to Stick: Why Some Ideas Survive and Others Die" by Chip Heath and Dan Heath