Roger Martin on Strategy
Roger Martin in conversation with Lenny Rachitsky on Lenny’s Podcast. Martin makes the case that strategy is a learnable skill, centres the conversation on his Strategy Choice Cascade, and argues that most companies fail at strategy because they play to play rather than play to win.
Key ideas
- The Strategy Choice Cascade. Five questions that must be answered together: (1) What is your winning aspiration? (2) Where will we play? (3) How will we win? (4) What capabilities must we have? (5) What management systems are required? Each answer constrains and reinforces the others; they cannot be solved independently.
- Two routes to winning. A company must either be the low-cost provider or be meaningfully differentiated in customers’ eyes. Any position in between leaves the company defenceless: competitors who have won on either axis can dictate terms at will.
- Capabilities and management systems are the moat. The how-to-win choice defines the moat; capabilities build it; management systems maintain it. The more multifaceted the capability stack, the less likely competitors are to attempt direct imitation — as Southwest Airlines demonstrates against every major US carrier.
- Counter-positioning and the fault line. Competitors are sometimes structurally prevented from responding: Estée Lauder could not take Clinique into mass-market channels without cannibalising its prestige portfolio, giving P&G’s Olay uncontested room to grow.
- Betterment over perfection. Rather than attempting a complete strategic overhaul, identify the single most painful gap between current and desired outcomes, make different choices to close it, then move to the next gap. Iterated over years, this compounds into genuine strategic mastery.
The Strategy Choice Cascade
| # | Question | What it does |
|---|---|---|
| 1 | Winning aspiration | Sets purpose and context; framed around customer benefit, not market share |
| 2 | Where to play | Defines the battlefield: customers, geography, product type, vertical stage, distribution channel |
| 3 | How to win | Specifies the competitive logic: low-cost leadership or differentiation customers recognise and value |
| 4 | Capabilities | Identifies activities and competencies required that competitors cannot or will not replicate |
| 5 | Management systems | Defines the hiring, training, incentives, and processes that build and sustain those capabilities |
On the resource-based view of the firm
Martin traces the academic dominance of the resource-based view (RBV) to institutional jealousy of Michael Porter. RBV argues strategy is about building resources, without specifying which resources or why. The framework produces tools such as VRIO analysis that practitioners ignore. Martin describes it as effectively mandatory doctrine at most top business schools and entirely unused in practice.
Case examples discussed
- Procter & Gamble: brand managers four levels below the CEO make strategy. P&G alumni outsized presence (10% of S&P 500 CEOs) attributed to training people at every level to make strategic choices.
- Southwest Airlines: pure low-cost model — single aircraft type, point-to-point routing, flexible unionised labour, direct booking — so structurally distinct that no competitor has successfully replicated it in 50 years.
- Vanguard: scale-dependent cost leadership; Jack Bogle resisted ETFs on principle but ultimately followed customers into them.
- Lego: differentiation so strong that children do not consider a toyshop without Lego to be a toyshop; captures 80–90% of category growth in most years.
- Four Seasons: exited real estate and hotel ownership to focus solely on management; redefined luxury as service that compensates for being away from home; achieved 10% staff turnover versus the industry’s 80%.
- Westlaw (Thomson Reuters): 100-year accumulation of lawyer-written case summaries using a proprietary keyword taxonomy; would require ~150,000 lawyer-years to replicate.
- Olay (P&G): exploited Estée Lauder’s counter-positioning constraint to become the largest skincare brand globally.