MINDSETOngoing practice

Create-Destroy Strategic Thinking

Generate better alternatives by rigorously attacking your own best ideas

Problem it solves

limiting beliefs

Best for

Strategists and leaders who want to overcome the natural tendency toward premature closure on first ideas and produce genuinely superior strategic alternatives

Not ideal for

Routine decisions where the first reasonable solution is adequate, or situations where speed of decision is more important than decision quality

Overview

Why this framework exists

Create-Destroy is a disciplined thinking process for overcoming the natural human tendency toward quick closure on first ideas. When facing a complex strategic situation, most people grab the first solution that comes to mind and then spend their effort justifying it rather than questioning it. Create-Destroy forces a deliberate cycle of generating alternatives and then rigorously attacking each one to expose fault lines and internal contradictions.

The process works because new alternatives should flow from a reconsideration of the facts of the situation and should address the weaknesses of already-developed alternatives. Simply adding shallow alternatives or straw men alongside the first idea is not Create-Destroy. Each new alternative must genuinely compete with the existing ones, and each attack must genuinely seek to destroy, not merely critique.

Rumelt recommends using a 'virtual panel of experts' to aid the destroy phase. This involves building an internal mental dialogue with well-remembered experts whose personalities and viewpoints you know well. Because humans are naturally adept at simulating other personalities, this panel can provide richer criticism than abstract frameworks. The panel helps overcome the natural reluctance to question one's own ideas.

Core principles

5 total
  1. Quick closure on first ideas is the enemy of good strategy; the first idea is rarely the best
  2. Genuine alternatives must emerge from reconsidering the facts, not from tweaking the first idea
  3. Destroying your own ideas is as important as creating them: you must expose fault lines before competitors or reality does
  4. A virtual panel of experts leverages the human brain's built-in capability for personality simulation
  5. Judgment can be improved through practice, but only if you commit to positions in advance and then evaluate your track record

Steps

4 steps
  1. Generate a First Insight
    Allow your initial analysis of the situation to produce a first idea, approach, or strategy. Do not suppress this instinct. Capture it clearly and completely, including the reasoning behind it.
    Pro tipThe first insight is valuable even if it is not the best one. It represents your instinctive read of the situation and often contains genuine insight. The problem is not the first idea itself but stopping there.
  2. Rigorously Attack the First Idea
    Subject your first idea to the most rigorous criticism you can muster. Invoke your virtual panel of experts and ask each one what is wrong with this approach. Look for internal contradictions, unstated assumptions, missing elements, and ways the idea could fail.
    Pro tipSteve Jobs on Rumelt's virtual panel would ask whether the strategy controls enough variables to deliver a truly excellent experience. Bruce Scott would demand to know the action implications. Alfred Chandler would question whether the strategy builds enduring capability.
    WarningDo not soften the attack to protect your ego. The goal is to genuinely destroy the idea, not to perform a ritual critique that leaves it intact.
  3. Generate a Genuinely Different Alternative
    Based on the weaknesses exposed in the destroy phase, develop a new alternative that addresses those weaknesses and flows from a genuine reconsideration of the facts. This is not a tweak of the first idea but a fundamentally different approach.
    Pro tipA useful heuristic: if your alternative and your original would look similar to an outside observer, you have not generated a genuine alternative. The new approach should start from a different diagnosis or guiding policy, not just adjust the actions.
    WarningBeware of adding straw-man alternatives (walk away, do more research, or a slightly modified version of the original) just to satisfy the requirement for alternatives. These do not constitute genuine Create-Destroy.
  4. Evaluate Alternatives Against Each Other
    Compare the alternatives in light of one another. Each alternative illuminates different aspects of the situation. The competition between alternatives often reveals insights that neither alternative alone would have produced.
    Pro tipGood strategies are usually 'corner solutions' that emphasize focus over compromise. If your evaluation produces a blended approach that tries to incorporate the best of each alternative, you may have lost the strategic focus that makes each alternative powerful.

Checklist

Saved in your browser

Examples

2 cases
TiVo Strategy Analysis

When executive MBA students analyzed TiVo's strategic situation, most quickly settled on their first insight: put smaller disk drives in TiVo machines, or focus on cable TV partnerships. When challenged to apply Create-Destroy, they reconsidered the fundamental diagnosis. Some saw TiVo as a consumer technology company; others diagnosed the real problem as cable companies' control of the set-top box, suggesting a legal strategy instead of a marketing one. Each genuinely different diagnosis produced a genuinely different strategy.

OutcomeThe Create-Destroy process revealed that the most promising strategic direction for TiVo was not better marketing of its current product but a fundamental reframing: attacking the cable monopoly's bundling practices through legal channels, where TiVo would have large and powerful allies. This insight would never have emerged from tweaking the first idea.
Virtual Panel of Experts Applied to Strategy

Rumelt maintains a mental panel of experts including Bruce Scott (who demands action implications), Alfred Chandler (who thinks in terms of historical trends and enduring scale), Steve Jobs (who asks whether you control enough variables for an excellent experience), and David Teece (who focuses on intellectual property, complementary assets, and licensing strategy). When evaluating a strategy, Rumelt mentally consults each expert and considers their specific critique.

OutcomeThe virtual panel consistently produces richer, more diverse criticism than abstract analytical frameworks because the human brain is optimized for personality simulation. Each panelist brings a different lens that exposes different weaknesses, collectively providing a more thorough critique than any single analytical approach.

Common mistakes

3 traps
Adding Straw-Man Alternatives
When asked to generate alternatives, most people add options like 'do nothing,' 'do more research,' or a slightly tweaked version of the original. These are not genuine alternatives and do not provide the critical perspective needed to improve strategy.
Justifying Rather Than Questioning
Human nature drives us to spend most of our effort justifying our first idea rather than questioning it. We seek confirming evidence and explain away disconfirming evidence. Create-Destroy requires actively seeking reasons why your idea is wrong.
Trusting Instinct in Complex Strategic Situations
While instinct works well for people situations and pattern matching, it is unreliable for complex strategic judgments involving likelihoods, cause-and-effect, and long-term competitive dynamics. These situations require disciplined analysis, not quick closure.

Origin story

How this framework came to be

Rumelt developed this technique through decades of teaching strategy to executive MBA students and consulting with organizations. He observed that when asked to generate alternatives, experienced executives consistently produced shallow straw men alongside their preferred option. The real challenge was not generating ideas but genuinely questioning them. The virtual panel of experts concept emerged from Rumelt's own practice of mentally consulting figures like Bruce Scott, Alfred Chandler, Steve Jobs, and David Teece when evaluating his own strategic recommendations.

Source

Traced to primary
Source · BOOK
Good Strategy/Bad Strategy: The difference and why it matters
Richard Rumelt · 2011
Open source →

Related frameworks

Browse all Mindset →