Set Strategy and Build Durable, Measurable Advantage

How to

1. 1Map the five forces for your specific business unit, not the whole 'industry': for each force write down the concrete evidence (Are buyers concentrated? Do suppliers control a critical input? How low are entry barriers? What substitutes exist?).
2. 2Identify which force is currently claiming the most profit — that is where your position is weakest and where positioning effort pays off most.
3. 3Ask the Magretta question: what would let us earn ROIC above our cost of capital here, structurally — a price premium, a cost advantage, or both?
4. 4Test whether you can shape rather than accept the structure: run the Blue Ocean six-paths scan for boundary reconstruction, or look for a counter-positioning move an incumbent cannot copy without damaging its existing business (Seven Powers).
5. 5Decide honestly whether the structure supports the returns you need. If it doesn't and can't be reshaped, the right strategic move may be to not play.

Watch out for

• —Confusing a hot industry (fast growth) with a profitable one — rapid growth with low entry barriers invites entrants who compete the margin away.
• —Analyzing at too high a level ('the software industry') so the forces blur; force analysis is sharp only for a specific business and buyer set.
• —Treating your five-forces read as a one-time exercise — Reeves and Big Data both note environments shift; the forces need re-reading as conditions change.

How to

1. 1Write the diagnosis in one paragraph: what is the single most important challenge blocking forward progress? Name the obstacle, not the aspiration.
2. 2Draft a guiding policy — the overall approach for grappling with that challenge — that rules out a vast array of possible actions (if it rules nothing out, it's not a policy).
3. 3List the coordinated actions and check coherence: does each action reinforce the others, or are they independent bets? Kill actions that don't fit.
4. 4Concentrate resources: pick the few pivotal objectives (Traction's Rocks, Leinwand's 3–6 capabilities) and starve the rest. Explicitly write down what you are choosing NOT to do.
5. 5Pressure-test with Porter's trade-offs: are your choices incompatible with rivals' approaches? If a competitor could do everything you do without giving anything up, you have no real trade-off and no durable position.

Watch out for

• —Fluffy diagnosis that restates the goal ('our challenge is to grow 30%') instead of the obstacle blocking it.
• —A guiding policy so broad it permits any action — a sign you avoided the hard choice.
• —Adding a 'differentiating capability' to the list because a competitor has it; coherence means the capabilities fit YOUR value proposition, not a benchmark.
• —Treating diagnosis as a one-time launch event; Rumelt frames strategy as a journey of discovery that gets refined.

How to

1. 1Set the winning aspiration: what does winning look like, in a particular place and a particular way (Playing to Win)? Not 'be a leader' — win at what, for whom.
2. 2Choose where to play: the specific segment, geography, channel, and customer job you will serve — and explicitly the ones you won't.
3. 3Choose how to win: pick differentiation or cost as your base of advantage (Porter), then define the single unique, provable attribute you'll own (Differentiate or Die, Obviously Awesome).
4. 4Anchor positioning on best-fit customers and their real competitive alternative: what would they use if you didn't exist? Position your unique attributes as the answer (Obviously Awesome's ordered method starting from competitive alternatives).
5. 5Pick a beachhead if you're crossing into a mainstream market: one narrow segment with a compelling, mission-critical reason to buy, dominate it, then expand (Crossing the Chasm).
6. 6Verify the cascade is coherent: do your capabilities and systems actually support this where/how? If not, adjust the choice or build the capability.

Watch out for

• —Choosing 'differentiation AND cost leadership' everywhere without the value innovation logic to support it — Porter warns this usually means stuck in the middle (Blue Ocean is the exception, and only when you genuinely break the value-cost trade-off).
• —Positioning for investors (grand future vision) instead of customers (immediate, provable value) — Obviously Awesome flags this directly.
• —Owning an attribute nobody cares about; the chosen difference must matter to the target market, backed by credentials.
• —Halo Effect caution: don't retrofit a 'positioning story' onto whatever is currently working — that's attribution, not choice.

How to

1. 1Run your position through the Seven Powers checklist: do you have (or can you build) scale economies, network economies, switching costs, branding, a cornered resource, process power, or counter-positioning? If none, you have no durable strategy yet.
2. 2For each candidate power, identify the Barrier explicitly — the concrete reason a rival can't or won't copy it (e.g., copying would cannibalize the incumbent's existing business = counter-positioning).
3. 3If your bet is internal-capability advantage (Leinwand, Davenport), verify the capability system is a mutually reinforcing set of 3–6 things competitors can't assemble piecemeal — imitation barrier comes from the fit, not any one capability.
4. 4Design the business model for the barrier from the start (Blitzscaling): can you engineer network effects, switching costs, or high gross margins into how the product works?
5. 5Estimate value the Helmer way: Market Size × Power. A big market with no power, or strong power in a tiny market, both underperform.
6. 6Continuity check (Porter/Magretta): a durable advantage requires a stable core value proposition over time so fit deepens and the barrier compounds — resist the urge to constantly reinvent the core.

Watch out for

• —Calling brand or culture a 'moat' without a barrier — Halo Effect warns these are often post-hoc attributions to success, not independent causes; be honest about whether the barrier is real.
• —Confusing operational effectiveness with strategic position: doing the same activities better is copyable (Porter); doing different activities, or the same ones in an incompatible configuration, is what lasts.
• —Betting entirely on one power in a shifting environment — Reeves and Big Data both note advantages erode; plan renewal.
• —Analytics-as-advantage requires more than tools: Davenport ties it to a DISTINCTIVE capability central to strategy — generic analytics everyone can buy is not a barrier.

How to

1. 1State each strategic bet as a falsifiable hypothesis with the riskiest assumption named first (Lean Startup).
2. 2Build the smallest test that produces real learning: an MVP, a façade prototype, or a five-day Sprint ending in interviews with five target customers (Sprint's magic number).
3. 3Work in small batches and raise tempo: many cheap tests beat one expensive one, and cadence — not any single experiment — drives the learning rate (Hacking Growth, The Lean Startup).
4. 4Use innovation accounting: track learning milestones and leading metrics, not vanity numbers, to judge whether the hypothesis held (Lean Startup, Hacking Growth's North Star).
5. 5Apply the barbell: keep the core safe and reversible while making bounded speculative bets; ensure any single failed experiment can't sink you (Antifragile).
6. 6Feed the pipeline with discovery behaviors — question assumptions, observe customers doing the job, network for outside perspectives (Innovator's DNA).

Watch out for

• —Running experiments without a decision attached — if no result would change what you do, you're gathering reassurance, not learning.
• —Overfitting to a tiny sample or to enthusiastic early adopters whose behavior won't generalize (Data Science for Business's warning on overfitting; Crossing the Chasm's early-market/mainstream gap).
• —Confusing motion with progress: high experiment volume on trivial questions isn't strategy learning. Prioritize high-impact tests (Hacking Growth).
• —The causation-vs-correlation trap: an experiment that shows an association isn't proof of a mechanism (see the tensions) — be clear which you have before betting big.

How to

1. 1Match the level of analysis to the decision — don't build a model for a reversible small call; do build one for a strategic, high-value bet (Analytics at Work).
2. 2Make assumptions explicit and test them; revisit models as conditions change (Analytics at Work).
3. 3Frame consequential decisions as expected value: enumerate outcomes, estimate probabilities from data, weight by business value (Data Science for Business).
4. 4Embed analytics into the process so insight leads directly to action — close the gap between analysis and decision (Analytics at Work; Competing on Analytics).
5. 5Install a weekly scorecard with a number owned by each person, and confront the brutal facts in review without blame (Traction; Good to Great).
6. 6Guard against overfitting and biases: hold out data, watch for base-rate neglect and overconfidence, and use structured checklists in high-stakes calls (Data Science for Business; Thinking, Fast and Slow).

Watch out for

• —Analysis paralysis on low-validity, one-off strategic calls where good data doesn't exist — Crossing the Chasm explicitly says use informed intuition, not numeric analysis, for high-risk low-data target decisions.
• —Vanity metrics that go up while the business doesn't — track meaningful outcome and leading indicators (Hacking Growth's North Star).
• —Halo-effect data: performance metrics collected after the fact get retrofitted into 'drivers.' Correlation with success is not causation of it (Halo Effect).
• —One version of the truth: fragmented, inconsistent data across silos produces confident wrong conclusions (Competing on Analytics).

How to

1. 1Practice transparency: communicate the true state of the company, including failures, rather than projecting unwarranted positivity (Horowitz's 'tell it like it is').
2. 2Diagnose wartime vs peacetime and set your style accordingly — directive and focused under existential threat, more delegative and expansive when dominant in a growing market (Horowitz).
3. 3Get the right people first, then decide direction — First Who, Then What — and hire for the specific world-class strength you need now (Good to Great; Horowitz).
4. 4Concentrate leadership attention on tipping-point levers: the few factors of disproportionate influence, not diffused effort (Blue Ocean).
5. 5Model fact-based decisions personally and invest resources in analytics if that's your capability — leaders who merely mandate it without modeling it fail (Competing on Analytics).
6. 6As the org scales, act as animator: diagnose the environment, segment where different strategic approaches apply, and orchestrate the portfolio (Reeves).

Watch out for

• —Confusing your ambition for the company with ambition for yourself — Level 5 is institution-first.
• —Avoiding hard decisions (layoffs, demotions) or doing them the wrong way, which destroys the trust that makes execution possible (Horowitz).
• —Over-crediting leadership for results and under-diagnosing structure or luck — Halo Effect; and beware charisma-as-competence, The Outsiders.
• —Mistaking constant reorganization for leadership; scaling should add structure and process incrementally ('give ground grudgingly'), not thrash the org.

How to

1. 1Define the few behaviors the strategy requires (e.g., surfacing bad news early, running experiments, acting on data) and make leaders model them visibly (Good to Great; Competing on Analytics).
2. 2Build a blameless climate for confronting facts and debriefing failures — autopsies without blame — so truth flows (Good to Great; OKRs from Mission to Metrics).
3. 3Use disciplined, transparent processes for politically sensitive decisions to minimize politics and preserve trust (Horowitz).
4. 4Identify the authentic strengths of your current culture and channel them toward the strategy instead of overwriting them (Leinwand).
5. 5Establish psychological safety and team trust so people voluntarily raise issues — Traction's high-trust issue-solving, and Blue Ocean's fair process (engagement, explanation, expectation clarity).
6. 6Evolve culture deliberately as you scale — Blitzscaling's Ship of Theseus — rather than letting it drift or ossify.

Watch out for

• —Declaring a culture instead of building one — culture is the accumulation of what leaders actually do and tolerate, not a statement.
• —Rewarding confident wrong answers over honest uncertainty, which quietly kills a fact-based culture.
• —Halo-effect self-congratulation: 'our great culture drove our results' is often the story we tell after winning; don't mistake it for a proven cause (Halo Effect).
• —Tying goals to compensation in a way that corrupts honesty — OKRs from Mission to Metrics warns detangling goals from pay keeps them a learning tool, not a gaming target.

How to

1. 1Translate the strategic choice into a shared, written vision everyone can repeat — who you are, where you're going, how you'll get there (Traction).
2. 2Get the right people in the right seats: share core values, operate in their strength, with clear roles (Traction; Good to Great; Horowitz).
3. 3Set 3–5 objectives with gradable key results per cycle; keep them transparent and reviewed weekly; use short nested cadences (OKRs; Goal Setting with OKR).
4. 4Install a meeting pulse and scorecard — quarterly Rocks, weekly reviews, a number owned by each person — to maintain accountability (Traction).
5. 5Document and simplify core processes so execution is repeatable and can be delegated as you scale (Traction; Horowitz's scaling techniques).
6. 6Use fair process — involve people in decisions that affect them, explain the reasoning, clarify expectations — to win voluntary cooperation rather than grudging compliance (Blue Ocean).
7. 7Build management systems (Lafley/Martin) that support and measure the specific capabilities the strategy needs — align incentives to the behaviors you actually want (Strategic Pay; What Your CEO Needs to Know About Sales Compensation).

Watch out for

• —Priority proliferation — the surest way to lose traction is to let the list of 'critical' initiatives grow until none gets real resources.
• —Measurement that becomes a control-and-blame system rather than a learning system (Balanced Scorecard's own caution: use it to inform and learn, not just to control).
• —Tying stretch goals to bonuses, which produces sandbagging and gaming (OKRs from Mission to Metrics).
• —Incentives that reward behaviors contradicting the strategy — if the sales comp plan pays for volume when the strategy is premium differentiation, the comp plan wins (What Your CEO Needs to Know About Sales Compensation; Strategic Pay).
• —Assuming alignment because you announced it — coordinated effort has to be built and checked, not assumed (Rumelt).

How to

1. 1Pick the right scorecard metric: ROIC vs cost of capital, and free cash flow, not just revenue or reported earnings (Valuation; The Outsiders).
2. 2Before pursuing growth, test whether the incremental returns exceed the cost of capital; if not, the growth destroys value — reinvest elsewhere or return capital (Valuation; The Outsiders' capital allocation discipline).
3. 3Map your flywheel: the specific sequenced components that reinforce each other, and push them relentlessly turn by turn (Turning the Flywheel).
4. 4Renew the flywheel as conditions change — the Genius of the AND: sustain the architecture AND change the components (Turning the Flywheel).
5. 5Attribute carefully: measure performance relative to real competitors and against the risk taken, not in absolute terms (Halo Effect).
6. 6Allocate capital like an owner: rank all uses (reinvestment, acquisition, buyback, debt reduction, dividends) by risk-adjusted return against a hurdle rate, and act boldly when returns are compelling (The Outsiders).

Watch out for

• —Vanity growth: scaling a business whose unit economics don't clear the cost of capital — the faster you grow, the faster you destroy value.
• —Managing to reported earnings or EPS through financial engineering — Valuation: transactions that don't increase cash flow don't create value.
• —Attributing a good run to your genius; performance is relative and risk-shaped (Halo Effect), and sustained greatness requires continuous adaptation, not resting on results.
• —Confusing momentum with a durable flywheel — momentum without an underlying barrier reverses when a rival matches you.