Strategy That Forces Better Choices

Executive summary

A business strategy is a coherent set of choices about the outcomes the organization will pursue, the arenas in which it will compete, the advantage it will build there, the capabilities that make the advantage real, and the management systems that concentrate resources behind it. The thesis of this lesson is simple: strategy is valuable only when it changes an allocation or refusal. Goals can be admirable without being strategic. Plans can be detailed without resolving a trade-off. A strategy becomes operational when a reasonable opportunity is declined because it weakens a more important choice.

This lesson uses a five-choice cascade associated with A.G. Lafley and Roger Martin, reinforced by Porter’s emphasis on a distinctive position, trade-offs, and fit among activities.^[s1][s2] It also adds an assumption portfolio and evidence loop. Those additions matter because strategy is made under uncertainty: leaders must commit enough resources for a choice to work while buying evidence before irreversible scale. The managerial implication is not “plan less.” It is to separate direction, choice, capability, experiment, and execution; connect them explicitly; and review the assumptions that could make the system fail.

Learning objectives

By the end of this lesson, you should be able to:

• distinguish strategy from aspirations, forecasts, budgets, operational improvement, and project portfolios;
• construct a five-part choice cascade in which each answer constrains and reinforces the next;
• diagnose incoherence by tracing customer value, activities, capabilities, systems, and resource allocation;
• rank strategic assumptions by importance and uncertainty, then design proportionate tests;
• translate strategy into a ninety-day learning agenda, decision rights, metrics, and stop-doing commitments; and
• evaluate failure modes, ethical risks, stakeholder consequences, and conditions that require adaptation or exit.

Foundations: strategy is a theory with consequences

Executives often use “strategy” for four different artifacts. An aspiration states a desired future, such as becoming the most trusted provider. A forecast estimates what may happen. A plan sequences activities. A budget authorizes spending. All are necessary, but none by itself explains why customers will choose the organization, why the organization can deliver that value unusually well, or why rivals will not immediately neutralize it.

The Strategy learning hub connects this cornerstone to competitive analysis, business models, operations, change, project governance, and strategic finance.

Porter separates strategy from operational effectiveness. Improving quality, speed, and productivity matters, but widespread adoption of the same best practices can move competitors toward convergence. Strategy requires a distinctive position, trade-offs, and fit among activities.^[s1] Rumelt similarly argues that good strategy contains a diagnosis, a guiding policy, and coherent actions rather than a list of ambitions.^[s3] These accounts share a discipline: strategy must focus effort through an explanation of the challenge and a coordinated response.

Strategy as a testable theory

A useful strategy is a theory of value creation and capture. It says, in effect: for this customer in this situation, if we configure these activities and capabilities differently, we can create an outcome they value, deliver it at an acceptable cost, and retain enough of the value to sustain the system. Every clause contains assumptions. The customer may not care enough. The channel may be uneconomic. The capability may be slower to build than expected. Competitors may respond. Regulation may change. Partners may capture the margin.

Calling strategy a theory does not mean postponing action until certainty. It means making causal logic visible so evidence can improve it. A three-year revenue target has no falsifiable mechanism. A claim that independent clinics will pay for same-day diagnostic coordination because it reduces lost patient time can be tested through behavior, unit economics, and operational feasibility.

Choice requires opportunity cost

Resources are finite even in well-funded organizations. Senior attention, specialist talent, sales capacity, engineering time, reputation, channel relationships, and organizational tolerance for change all have alternative uses. A “priority” that replaces nothing is an addition. Additions accumulate until every function protects a different interpretation of strategy.

The most revealing strategy question is therefore not “What are our initiatives?” but “What attractive work will we stop, delay, or refuse?” The answer should follow from the chosen arena and advantage, not from arbitrary austerity. A specialist may decline a broad adjacent segment because serving it would require product variety and sales behavior that erode the specialist advantage. A low-cost operator may refuse custom work even when a customer is willing to pay, because exceptions damage the activity system supporting the broader position.

Position, capabilities, and path dependence

Market position alone is incomplete. An attractive arena does not explain why this organization can win. The resource-based view asks whether resources and capabilities are valuable, scarce, and difficult to imitate or substitute.^[s7] Dynamic-capabilities research adds the ability to sense opportunities, seize them, and reconfigure assets as environments change; Teece, Pisano, and Shuen emphasize processes, asset positions, and paths rather than a detachable list of strengths.^[s5] History matters because accumulated knowledge, relationships, routines, and constraints influence which strategies are feasible.

This insight protects against strategy by imitation. A competitor’s visible product feature is not its complete system. Copying the feature without complementary data, processes, economics, channel access, culture, or installed base may reproduce cost without reproducing advantage.

Commitment and learning are complements

Strategy requires commitment, but uncertainty requires learning. March’s distinction between exploration and exploitation shows why organizations must allocate resources between search and refinement; returns have different timing and uncertainty.^[s4] Gavetti and Levinthal distinguish forward-looking cognitive search from backward-looking experiential search, reminding leaders that models can explore possibilities beyond experience while experience disciplines unrealistic models.^[s6]

The false choice is between rigid planning and endless experimentation. The better design is stable intent with staged commitment. Protect the chosen outcome and strategic boundaries while testing assumptions about customers, channels, capabilities, and economics. Scale only when the evidence justifies increasing irreversibility.

Framework: the five-choice cascade with an evidence loop

The cascade below should be read in both directions. Downward, each choice imposes requirements on the next. Upward, capability and system realities may expose an implausible choice. Evidence can revise any assumption, but revisions must preserve coherence across the whole cascade.

1. Winning aspiration: what valuable outcome defines victory?

An aspiration should state a valued outcome and the role the organization intends to play, not merely rank or financial size. “Be number one” describes status. “Enable independent clinics to deliver a reliable diagnosis within one day” begins with customer value. Financial ambition still matters: define the economic conditions under which the outcome is sustainable. But do not mistake the financial result for the reason a customer chooses.

Test the aspiration with four questions. Whose outcome improves? What condition changes? Why is the outcome important now? What ethical or economic boundary must not be crossed? If leadership cannot answer, the statement is inspirational language rather than a strategic anchor.

2. Where to play: which arena will receive disproportionate commitment?

Specify customer, need, use situation, offer scope, channel, geography, and position in the value chain. Not every dimension must be narrow, but the combination must guide investment. “Small businesses” is not useful if product, price, distribution, and support differ radically across small retailers, clinics, manufacturers, and software firms.

Choose using evidence about problem severity, willingness to change, reachable demand, channel economics, competitive intensity, regulatory constraints, capability fit, and learning value. Separate current attractiveness from right-to-win. A large market can be strategically poor for a firm without relevant assets. A small beachhead can be valuable if it permits learning and builds a capability that transfers to a coherent next arena.

3. How to win: why will the chosen customer prefer and sustain this system?

The how-to-win choice names the advantage from the customer’s perspective and the activity difference that makes it credible. Examples include lower total cost through standardization, faster resolution through integrated data, superior reliability through process control, specialist judgment through focused expertise, or lower adoption risk through exceptional implementation.

Avoid benefit stacking. “Faster, better, cheaper, personalized, innovative, and premium” conceals trade-offs. Ask what the organization will do differently, what cost or inconvenience the customer accepts, and which buyers may rationally prefer another offer. If no customer is excluded and no activity changes, the advantage is probably generic.

When the advantage depends on customer meaning and source attribution, translate it into the brand strategy operating system. Brand should compress and evidence the how-to-win choice, not invent a disconnected promise.

4. Capabilities: what must the organization repeatedly do unusually well?

A capability is a repeatable organizational ability, not a department or virtue. “Great people” is not diagnostic. “Diagnose a complex implementation risk within forty-eight hours using integrated product and customer data” is closer. Identify three to seven capabilities that reinforce one another. For each, name the required process, talent, information, technology, relationship, and learning mechanism.

Distinguish threshold capabilities from differentiating capabilities. Secure billing or basic quality may be required to compete but not a source of preference. The differentiating system may depend on how several ordinary components fit. Porter emphasizes that fit among activities can strengthen advantage and make piecemeal imitation less useful.^[s1]

5. Management systems: what keeps choices alive after the workshop?

Management systems include decision rights, resource allocation, metrics, incentives, operating cadence, talent systems, governance, information flow, and exception rules. They are where stated strategy meets power. If a specialist strategy rewards sales only for volume, the incentive invites off-strategy customers. If fast experimentation is essential but approvals take eight weeks, the system contradicts the choice.

For each capability, specify an owner, leading indicator, review cadence, investment, constraint, and escalation path. Link every funded initiative to the choice it strengthens, the assumption it tests, and the work it replaces. This creates traceability without turning strategy into project administration.

The cascade converts ambition into a reinforcing operating system. The return path matters: execution produces evidence, and evidence may require revising an assumption rather than merely pressuring teams to deliver the original forecast.

6. Build the assumption portfolio

List the assumptions underneath the cascade. Typical groups include desirability, demand, access, willingness to pay, competitive response, capability feasibility, unit economics, regulation, partner behavior, and organizational adoption. Score each assumption on importance and uncertainty. Importance asks how badly the strategy fails if the assumption is wrong. Uncertainty asks how weak the present evidence is.

High-importance, high-uncertainty assumptions deserve early tests. High-importance, low-uncertainty assumptions need monitoring and contingency plans. Low-importance, high-uncertainty assumptions may receive cheap validation or be deferred. Do not spend six months perfecting evidence for a low-consequence detail while a fatal customer or cost assumption remains untested.

Tests should match the uncertainty. Interviews can reveal language and mechanisms but rarely prove willingness to pay. A paid pilot tests commitment but may overrepresent early adopters. A concierge service can test value before automation but may hide future delivery costs. A pre-mortem tests coherence, not market behavior. Treat every method as partial evidence.

7. Write the stop-doing architecture

For every positive choice, record the corresponding refusal, transition cost, affected owner, customer impact, and reconsideration trigger. A refusal can be permanent, conditional, or time-bound. “We will not serve enterprise clients” may be too absolute; “we will not build enterprise-specific workflow until the mid-market implementation model reaches target economics and the capability can transfer without custom fragmentation” is testable.

This artifact reduces shadow strategy. Without it, declined work returns through exceptions, executive favors, and budget leftovers. Review exceptions as evidence. One exception may be harmless; a recurring class of exceptions may reveal that the choice or implementation rule is wrong.

Worked example: choosing a growth strategy for a logistics software firm

The case is hypothetical and designed to demonstrate assumptions, choices, numbers, and consequences.

RouteNorth provides transport-management software to Indian manufacturers. Revenue is ₹42 crore, gross margin is 61 percent, and growth has slowed from 32 to 14 percent. Leadership proposes three initiatives: enter large enterprise, expand to Southeast Asia, and launch a low-priced product for small fleet owners. All appear plausible. Together they require different product architecture, sales cycles, channels, support models, and compliance work.

Diagnosis reveals the central challenge: RouteNorth is spread across customers whose implementation needs differ dramatically. Product exceptions consume 37 percent of engineering capacity. Median implementation is seventy-nine days. Sales celebrates contract value, while customer success absorbs configuration complexity. The company has good route-optimization logic but no repeatable implementation capability.

The winning aspiration becomes: “Help regional manufacturers make multi-carrier transport reliably visible and controllable within thirty days, while maintaining software gross margin above 65 percent.” The economic boundary prevents a service-heavy promise from masquerading as scalable software.

Where to play: manufacturers with ₹300–2,000 crore revenue, recurring multi-carrier routes, operations in India, and a finance or operations sponsor willing to standardize. RouteNorth declines very small fleets, highly customized national conglomerates, and immediate geographic expansion. These are not judgments about customer worth; they reflect present capability and activity fit.

How to win: a thirty-day control launch using preconfigured carrier integrations, exception workflows, and a specialist implementation squad. The alternative is neither the broadest platform nor the lowest price. The customer accepts constrained configuration in exchange for faster visibility and lower implementation risk.

Four capabilities must become true: segment diagnosis, configurable integration, implementation orchestration, and exception-learning. Management systems change accordingly. Sales commission includes activation quality, not contract signature alone. Product reserves capacity for the reusable integration library. Implementation reviews occur weekly. Custom requests need an exception case showing transfer value across the target segment.

The assumption portfolio identifies a fatal uncertainty: will target customers accept standardization? RouteNorth runs eight paid pilots. Six launch within thirty-five days; five adopt standard workflows without major customization; two require legacy integrations that destroy target economics. The evidence leads to a sharper qualification rule, not abandonment of the strategy.

Pilot economics are encouraging but incomplete. Average first-year contract value is ₹34 lakh. Direct implementation cost falls from ₹11 lakh to ₹6.5 lakh. Expected software gross margin after launch is 68 percent. However, sales-cycle data cover only eight customers, and the implementation squad includes unusually senior staff. Leadership funds a second cohort with ordinary staffing and postpones Southeast Asia by two quarters.

The stop-doing list has consequences. Two enterprise opportunities worth a combined ₹5.8 crore are declined because they require bespoke architecture. A small-fleet product prototype is discontinued. Twelve roles are reassigned toward the integration library and implementation system. These choices make the strategy credible because resources move.

At six months, median implementation for qualified customers is thirty-three days, engineering exception work falls from 37 to 24 percent, and software gross margin for the cohort is 66 percent. Growth has not yet accelerated enough to prove the full strategy. The evidence supports capability progress and economics within the chosen segment; market-scale and competitive-response assumptions remain open.

The case illustrates an important distinction. Success is not the accurate prediction of every result. It is a coherent commitment that exposes decisive assumptions early, learns without dissolving the choice, and reallocates resources when evidence changes.

Action Plan: make strategy operational in ninety days

Days 1–30: diagnose and choose

Appoint one accountable strategy owner and a small cross-functional team. Start with evidence, not a blank presentation. Review segment economics, customer behavior, competitor activity systems, capabilities, bottlenecks, lost deals, retention, channel performance, regulatory constraints, and resource allocation. Interview frontline employees because they see contradictions that executive summaries hide.

Write a diagnosis in one paragraph: the critical challenge, its mechanism, and why current actions do not resolve it. Generate at least three materially different strategic options. Each must include an aspiration, arena, advantage, capabilities, systems, risks, and stop-doing implications. Avoid options that differ only in growth rate.

By day 30, leadership selects a provisional cascade and records dissent. The decision memo names the rejected alternatives and the evidence that could revive them. This prevents later mythology that only one option existed.

Days 31–60: test coherence and fatal assumptions

Run a cascade audit with functions independently translating the strategy into decisions. Compare their answers. If sales, product, operations, and finance infer different customers or advantages, the strategy is not yet communicable.

Build the assumption portfolio. Design the cheapest credible test for the three highest-importance uncertainties. Use behavior where possible: paid pilots, pre-commitments, switching actions, prototype use, channel economics, or operational trials. Specify success, failure, and ambiguous zones in advance. Kahneman and Tversky’s work on prospect theory is a reminder that people evaluate gains and losses relative to reference points, so framing and loss aversion can distort executive judgment as well as customer decisions.^[s9] Precommitted thresholds reduce motivated interpretation.

Map the required capability system. For each capability, define current evidence, target behavior, owner, investment, lead time, dependency, and make–buy–partner decision. Identify what must exist before scale and what can be learned during scale.

Days 61–90: reallocate and establish cadence

Move resources. Cancel, pause, or reduce at least one activity that conflicts with the strategy. Update decision rights, incentives, qualification rules, portfolio criteria, and metrics. Communicate the logic and trade-offs, not only the slogan. Give frontline teams examples of reasonable requests they should decline.

Launch a bounded operating cycle. Weekly reviews handle experiment execution and obstacles. Monthly reviews inspect capability and resource evidence. Quarterly reviews assess strategic assumptions and external change. Do not reopen the entire strategy because one metric moves; do not protect the strategy when several independent signals invalidate a core assumption.

Use the practical decision-making system for irreversible choices and formal decision records. Use scope control when project additions quietly reconstruct the rejected strategy. Use the after-action review process to turn execution evidence into learning rather than blame.

Measurement: a strategy scorecard that preserves causality

Measure at four levels: choice integrity, capability progress, customer evidence, and economic outcomes.

Choice-integrity measures include resource concentration in chosen arenas, revenue from target versus off-strategy customers, exception volume, stopped-work completion, senior-attention allocation, and percentage of initiatives linked to a cascade choice. A high exception rate is not automatically failure; it is a diagnostic requiring classification.

Capability measures should describe behavior and throughput. Examples include qualified implementation time, reusable integration rate, experiment cycle time, specialist accuracy, first-contact resolution, or forecast reliability. Avoid labels such as “innovation capability” without observable outputs and mechanisms.

Customer evidence includes problem incidence, switching behavior, paid-pilot conversion, time to value, usage of the differentiating behavior, retention in the chosen segment, price realization, referral quality, and reasons for loss. Economic outcomes include contribution margin, customer acquisition payback, return on invested capability, cash exposure, and growth quality.

Useful formulas include:

• strategic resource concentration = resources committed to chosen capabilities and arenas ÷ discretionary strategic resources;
• off-strategy revenue share = revenue from deliberately excluded or exception customers ÷ total revenue;
• capability yield = outputs meeting the chosen standard ÷ eligible attempts;
• assumption burn-down = high-importance uncertain assumptions resolved or materially narrowed ÷ starting high-importance uncertain assumptions;
• strategy-adjusted contribution = revenue from the target system minus direct delivery, acquisition, exception, and capability-sustaining costs; and
• experiment value = expected loss avoided or decision improved minus the cost and delay introduced by the test.

Do not optimize each metric independently. A team can reduce implementation time by rejecting hard but strategically important cases, raise margin by starving capability investment, or improve target-segment share by redefining the segment. Use paired measures and qualitative case review. Strategy measurement is a causal conversation supported by data, not a machine that eliminates judgment.

The portfolio concentrates learning where a wrong assumption would be fatal and present evidence is weak. It prevents teams from over-researching comfortable details while scaling an untested strategic premise.

Set review triggers. Reassess an arena if reachable demand falls below the threshold across two independent sources. Pause scale if unit economics miss the floor after a defined learning period. Redesign a capability if yield does not improve after two intervention cycles. Reopen the how-to-win choice if customers value the outcome but consistently prefer a rival’s activity system.

Failure modes and corrective actions

1. Goals masquerading as strategy

The document promises growth, leadership, innovation, and customer centricity. The warning sign is that competitors could adopt the same words without changing activities. Correct it by specifying arena, advantage, capabilities, systems, and refusal.

2. The initiative pile

Strategy becomes a collection of legacy projects plus new executive ideas. Nothing is stopped. The warning sign is resource allocation unchanged after “strategic transformation.” Correct it by requiring every initiative to name the choice it strengthens, assumption it tests, and work it replaces.

3. Universal customer focus

Teams interpret customer centricity as saying yes to every request. The result is complexity that makes the core experience worse. Correct it by defining the chosen customer and problem, creating a fair exception rule, and referring poor-fit customers honestly rather than overselling fit.

4. Advantage without an activity system

The strategy claims superior service or innovation but does not change how work is performed. Early warning signs are generic capabilities and unchanged incentives. Correct it by tracing the advantage backward into required activities, information, talent, technology, relationships, and trade-offs.

5. Analysis without commitment

Leadership keeps testing because no evidence feels conclusive. The warning sign is repeated research with no decision threshold. Correct it by defining the minimum evidence for a staged commitment and limiting the cost of reversibility.

6. Commitment without learning

Leaders treat disconfirming evidence as execution failure. Forecast gaps trigger pressure rather than diagnosis. Correct it by separating choice review from operating review, protecting honest escalation, and documenting which assumptions the evidence challenges.

7. Strategy by competitor imitation

The organization copies visible features, channels, or language. The warning sign is a roadmap organized around parity. Correct it by analyzing the rival’s complete activity system and returning to the chosen customer’s unresolved problem and the firm’s feasible capability path.

8. The annual ritual

Strategy is discussed once a year and displaced by the budget. Correct it with a regular assumption review, resource-allocation evidence, and clear decision rights. Annual planning can fund strategy; it cannot substitute for ongoing strategic management.

Ethics and limits

Strategy distributes benefits, costs, risks, and voice. Choosing a segment can improve service for some while excluding others. Standardization can lower cost while reducing accessibility. Automation can improve speed while shifting error risk to customers or workers. A serious strategy names these effects instead of treating stakeholder harm as an externality.

Use four ethical tests. The value test asks whether the strategy solves a real problem rather than exploiting confusion or dependency. The distribution test asks who captures value and who bears transition or failure costs. The voice test asks whether affected groups can surface evidence and seek remedy. The reversibility test asks whether the organization can correct harm before it becomes entrenched.

Do not use “strategic fit” to disguise discrimination, retaliation, unsafe work, deceptive selling, collusion, or avoidance of legal duties. Employment, competition, consumer, environmental, privacy, financial, health, and sector-specific rules vary by jurisdiction. This lesson is not legal or investment advice; consequential decisions require qualified local review.

Frameworks also have limits. The five-choice cascade encourages coherence but can underrepresent politics, emergent behavior, ecosystem dependence, institutional constraints, and moral commitments that are not reducible to competitive advantage. Dynamic-capabilities language can become circular if success is used as proof that a capability existed. Resource-based analysis can encourage inward focus. Porter’s activity fit can be misread as a reason to resist necessary change. Use multiple lenses and state what evidence each cannot provide.

In crisis, the organization may need temporary action before a complete cascade is available. In exploration, the initial task may be problem discovery rather than advantage design. Public institutions and social enterprises often have plural goals and non-market accountability. The discipline still applies—choices, assumptions, capabilities, systems, and consequences—but “winning” must be defined in mission-consistent terms.

Checklist and executive practice

Use this checklist before approving a strategy or major strategic initiative.

• The diagnosis explains the critical challenge and its mechanism, not only symptoms.
• The winning aspiration names a customer or stakeholder outcome and sustainable economic boundary.
• Where-to-play choices specify customer, need, situation, offer, channel, geography, or value-chain position as relevant.
• The how-to-win choice names both customer value and a different activity system.
• We can identify customers or opportunities that may rationally prefer another provider.
• Three to seven reinforcing capabilities are described as repeatable abilities with owners.
• Management systems, incentives, decision rights, and metrics support rather than contradict the choices.
• The three most important uncertain assumptions have tests, thresholds, owners, and dates.
• At least one meaningful resource allocation changes and one attractive activity stops.
• Stakeholder harms, accessibility, legal constraints, and remedy paths are documented.
• Frontline managers can use the strategy to decide among competing reasonable requests.
• We know what evidence would trigger adaptation, pause, or exit.

Field exercise: the strategy coherence hearing

Divide the leadership team into five groups, one for each cascade choice. Each group receives the same proposed strategy and independently writes what its choice requires from the other four. Reassemble the cascade. Contradictions reveal hidden ambiguity. For example, a premium specialist advantage may be paired with a broad arena, generalist talent model, volume incentive, and low-touch service system.

Next, assign a “credible dissenter” to argue the strongest alternative strategy and a “stakeholder advocate” to identify costs borne by customers, employees, partners, communities, or the environment. The strategy owner must respond with evidence, not authority. Record unresolved assumptions in the portfolio rather than forcing artificial consensus.

Finally, choose one live opportunity. Ask whether accepting it strengthens the cascade, tests a valuable assumption, creates a transferable capability, or merely adds revenue. Write the decision and the precedent it creates. Strategy becomes learnable when choices leave a trace.

Reflection questions

1. Which current initiative would be impossible to justify if the cascade were taken seriously?
2. Where does money contradict the stated strategy?
3. Which capability is assumed rather than demonstrated?
4. What customer evidence could overturn the how-to-win choice?
5. Which stakeholder bears risk without adequate voice or remedy?
6. What would we refuse if the strategy were real tomorrow?

Key takeaways

• Strategy is a causal theory expressed through connected choices, not a target or project list.
• A choice becomes credible when it reallocates finite resources and excludes an attractive alternative.
• Where to play and how to win must reinforce one another; capabilities and systems make the advantage operational.
• Commitment and learning are complements: protect strategic intent while testing fatal assumptions before irreversible scale.
• Metrics should trace choice integrity, capability behavior, customer evidence, and economics without replacing judgment.
• Strategy reviews must distinguish execution problems from evidence that the underlying theory is wrong.
• Ethical strategy makes stakeholder consequences, legal constraints, voice, and remedy explicit.

References

• [s1] What Is Strategy? Michael E. Porter, 1996, *Harvard Business Review*. https://hbr.org/1996/11/what-is-strategy
• [s2] Playing to Win: How Strategy Really Works. A.G. Lafley and Roger L. Martin, 2013, Harvard Business Review Press, ISBN 9781422187395.
• [s3] Good Strategy/Bad Strategy. Richard P. Rumelt, 2011, Crown Business, ISBN 9780307886231.
• [s4] Exploration and Exploitation in Organizational Learning. James G. March, 1991, *Organization Science*. https://doi.org/10.1287/orsc.2.1.71
• [s5] Dynamic Capabilities and Strategic Management. David J. Teece, Gary Pisano, and Amy Shuen, 1997, *Strategic Management Journal*. https://doi.org/10.1002/(SICI)1097-0266(199708)18:7%3C509::AID-SMJ882%3E3.0.CO;2-Z
• [s6] Looking Forward and Looking Backward: Cognitive and Experiential Search. Giovanni Gavetti and Daniel A. Levinthal, 2000, *Administrative Science Quarterly*. https://doi.org/10.2307/2666981
• [s7] Firm Resources and Sustained Competitive Advantage. Jay B. Barney, 1991, *Journal of Management*. https://doi.org/10.1177/014920639101700108
• [s8] Discovery-Driven Planning. Rita Gunther McGrath and Ian C. MacMillan, 1995, *Harvard Business Review*. https://hbr.org/1995/07/discovery-driven-planning
• [s9] Prospect Theory: An Analysis of Decision under Risk. Daniel Kahneman and Amos Tversky, 1979, *Econometrica*. https://doi.org/10.2307/1914185
• [s10] Chasing a Moving Target: Exploitation and Exploration in Dynamic Environments. Hart E. Posen and Daniel A. Levinthal, 2012, *Management Science*. https://doi.org/10.1287/mnsc.1110.1420

[s1] Porter, “What Is Strategy?” Harvard Business Review, November–December 1996, https://hbr.org/1996/11/what-is-strategy.

[s2] Lafley and Martin, Playing to Win: How Strategy Really Works, Harvard Business Review Press, 2013, ISBN 9781422187395.

[s3] Rumelt, Good Strategy/Bad Strategy, Crown Business, 2011, ISBN 9780307886231.

[s4] March, “Exploration and Exploitation in Organizational Learning,” Organization Science 2(1), 1991, https://doi.org/10.1287/orsc.2.1.71.

[s5] Teece, Pisano, and Shuen, “Dynamic Capabilities and Strategic Management,” Strategic Management Journal 18(7), 1997, https://doi.org/10.1002/(SICI)1097-0266(199708)18:7%3C509::AID-SMJ882%3E3.0.CO;2-Z.

[s6] Gavetti and Levinthal, “Looking Forward and Looking Backward,” Administrative Science Quarterly 45(1), 2000, https://doi.org/10.2307/2666981.

[s7] Barney, “Firm Resources and Sustained Competitive Advantage,” Journal of Management 17(1), 1991, https://doi.org/10.1177/014920639101700108.

[s8] McGrath and MacMillan, “Discovery-Driven Planning,” Harvard Business Review, 1995, https://hbr.org/1995/07/discovery-driven-planning.

[s9] Kahneman and Tversky, “Prospect Theory,” Econometrica 47(2), 1979, https://doi.org/10.2307/1914185.

[s10] Posen and Levinthal, “Chasing a Moving Target,” Management Science 58(3), 2012, https://doi.org/10.1287/mnsc.1110.1420.