How Enterprise Architecture Shapes Strategy in a Volatile World

Enterprise Architecture sits at a unique crossroads between strategy and execution. It is one of the few disciplines that must simultaneously understand the external forces shaping the market and the internal capabilities that determine what an organisation can realistically do next. In an era of global competition, digital disruption, and constant change, architects can no longer afford to align with only one strategic worldview. Instead, we must integrate multiple perspectives into a coherent, actionable lens.

Two classic schools of strategy — the positioning perspective and the resource-based perspective — continue to shape how organisations think about competitiveness. Each has strengths, limitations, and profound implications for how architecture should be designed and governed.

The positioning perspective assumes that the external environment defines strategic freedom. Customers, competitors, suppliers, substitutes, and new entrants collectively constrain what is possible. Michael Porter’s Five Forces framework remains influential because it forces leaders to confront uncomfortable truths about industry attractiveness and competitive pressure. From this view, strategy is about choosing where to play and how to defend that position. Success depends on continuously monitoring market signals, anticipating shifts in customer demand, and responding faster than rivals.

For Enterprise Architects, this perspective reinforces the importance of external awareness. Architecture decisions cannot be made in isolation from pricing pressure, regulatory change, ecosystem dynamics, or platform competition. A technically elegant architecture that ignores these forces risks optimising the wrong outcomes. However, the positioning perspective also has a blind spot: it implicitly assumes that firms within an industry are largely similar, differing mainly in how well they adapt.

This is where the resource-based perspective fundamentally changes the conversation. Rather than starting with the market, it starts with the firm. Organisations are not interchangeable. Each has a unique combination of tangible and intangible resources — skills, knowledge, processes, culture, data, brand, and technology — that competitors cannot easily replicate. From this angle, strategy is not about fitting into an existing industry structure, but about reshaping it.

The early days of Google illustrate this clearly. The founders did not begin by analysing entry barriers or industry profitability. They focused on what they could do uniquely well, and in doing so, they changed the rules of the search market entirely. For architects, this perspective validates a core architectural instinct: sustainable advantage comes from what is hard to copy, not what is easy to benchmark.

Yet resources alone are not enough. Servers, engineers, data, and brands do not create value by themselves. What matters is the organisation’s ability to combine them effectively. This is the realm of capabilities — the complex bundles of skills, learning, and coordination embedded in organisational processes. Capabilities bridge the gap between internal potential and external value.

Thinking in terms of capabilities shifts architectural conversations away from isolated systems and towards end-to-end outcomes. The ability to anticipate customer needs, to sense emerging technologies, to deliver at scale, or to orchestrate partners across borders are not properties of individual applications. They emerge from how systems, people, and processes work together. Enterprise Architecture, at its best, is a capability-design discipline.

Global competition raises the stakes further. Competing internationally introduces new demands: managing currency risk, scanning global technology trends, and transferring tacit knowledge across borders. Firms also face structural disadvantages — liabilities of foreignness, expansion, smallness, and newness — that compound complexity. Architecture must therefore enable learning, adaptability, and local responsiveness without fragmenting the enterprise.

This is where comparative analysis becomes critical. Many organisations believe they are above average, yet few systematically study their competitors. Without credible competitor intelligence, claims of “core capabilities” are often little more than internal myths. Architects should challenge this complacency. Understanding how rivals structure their platforms, manage costs, or scale operations provides essential context for architectural trade-offs.

Benchmarking, when done well, is not imitation for its own sake. Xerox’s response to Canon in the 1980s was not about copying a product, but about learning better processes. For modern enterprises, benchmarking might involve cloud cost structures, DevOps maturity, data platform scalability, or ecosystem integration patterns. The goal is not to be identical, but to close blind spots.

Strategic intent, however, means little without effective implementation. Organisational structure plays a decisive role here. Matrix structures promise synergy across products and markets, but often collapse under their own complexity. Dual reporting lines, overlapping accountability, and decision latency can undermine execution, especially across geographies. Some firms abandon the matrix after painful experience; others, like Disney, sustain hybrid forms through strong leadership and clarity of purpose. For architects, this reinforces a hard truth: structure and governance matter as much as technology.

Finally, strategy and architecture must evolve through change. Not all change is equal. Incremental change refines what already exists; transformational change redefines beliefs, identities, and priorities. The latter demands leadership, not just management. Global organisations must choose change styles deliberately, based on urgency, environmental fit, and internal support. Participative approaches work when time and alignment exist. Dictatorial transformation, while uncomfortable, may be necessary when survival is at stake.

Enterprise Architects are often positioned as neutral facilitators, but in transformational moments, neutrality is not enough. Architects must help leaders translate vision into coherent operating models, align capabilities with ambition, and ensure that change is structurally and technologically possible.

In a volatile world, sustainable advantage does not come from choosing between positioning or resources, structure or culture, incremental or transformational change. It comes from integrating them. Enterprise Architecture provides the connective tissue — linking market insight to internal capability, strategy to execution, and vision to reality. When done well, it does not merely support strategy. It helps shape it.

企業架構如何在動盪世界中塑造策略

企業架構（Enterprise Architecture）位於策略與執行的交會點上，是少數必須同時理解外部市場力量與內部組織能力的學科。在全球競爭、數位顛覆與持續變革的時代，組織已無法只依賴單一策略視角。企業架構師必須整合多種策略觀點，將其轉化為一致且可落地的行動方向。

策略領域中，定位觀點與資源基礎觀點這兩個經典學派，至今仍深刻影響企業對競爭優勢的理解。兩者各有優勢與限制，也對企業架構的設計與治理帶來截然不同的啟示。

定位觀點假設，企業的策略空間主要由外部環境所決定。顧客需求、競爭對手行為、供應商議價能力、替代品威脅以及新進入者，共同形塑了企業的行動邊界。Michael Porter 的五力分析之所以歷久不衰，正是因為它迫使管理者正視產業吸引力與競爭壓力的現實。在此觀點下，策略的核心在於選擇「在哪裡競爭」以及「如何防守或擴大該位置」，而成功仰賴對市場訊號的持續監測與對變化的快速回應。

對企業架構師而言，這提醒我們：架構決策不能脫離市場現實。若忽略價格壓力、監管變化、生態系競爭或平台化趨勢，再精巧的技術架構也可能優化了錯誤的目標。然而，定位觀點也存在盲點——它往往假設同一產業中的企業大致相似，只是適應環境的能力不同。

資源基礎觀點則徹底改變了這個假設。它不從市場出發，而是從企業本身出發。企業並非可互換的個體，每一家組織都擁有獨特的有形與無形資源組合，包括技能、知識、流程、文化、資料、品牌與技術，而這些往往難以被競爭對手複製。從這個角度看，策略並非只是適應產業結構，而是透過自身獨特能力去改變遊戲規則。

Google 的早期發展正是明證。創辦人並未首先分析進入障礙或產業獲利性，而是專注於他們能夠做得特別好的事情，並因此重塑了搜尋引擎市場。對企業架構師而言，這印證了一個核心信念：可持續的競爭優勢來自「難以模仿的能力」，而非「容易複製的做法」。

然而，單一資源本身並不會自動產生價值。工程師、伺服器、資料或品牌若未被有效整合，仍然只是潛力。真正關鍵的是能力（capabilities）——也就是組織中透過流程與學習所形成的複雜技能組合。能力連結了內部資源與外部價值，是將潛力轉化為成果的關鍵橋樑。

以能力為思考單位，能將企業架構的討論從單一系統，提升到端到端價值交付。預測客戶需求、感知新興技術、規模化交付，或跨國協同夥伴，這些都不是某一套系統的屬性，而是人、流程與技術共同運作的結果。從這個角度看，企業架構本質上是一門「能力設計」的學科。

當競爭走向全球化，複雜度進一步提升。企業必須面對匯率風險、全球技術掃描、跨國知識移轉等新挑戰，同時承擔外國性、擴張性、小規模與新進入市場所帶來的結構性劣勢。企業架構因此必須在不犧牲整體一致性的前提下，支持學習能力、彈性與在地回應。

在此背景下，比較分析變得至關重要。許多組織自認表現高於平均，但卻缺乏系統性競爭者分析。沒有可信的競爭情報，「核心能力」往往只是內部自我感覺良好的神話。企業架構師有責任挑戰這種盲目樂觀，透過理解競爭對手的平台設計、成本結構與擴展能力，為架構決策提供現實基準。

標竿學習（Benchmarking）並非盲目模仿。Xerox 面對 Canon 的競爭時，真正學到的不是複製產品，而是改善流程。放在今日，標竿學習可能發生在雲端成本管理、DevOps 成熟度、資料平台可擴展性或生態系整合模式上。目的不是變得一模一樣，而是消除關鍵盲點。

策略若無法有效落地，終究只是空談。組織結構在此扮演關鍵角色。矩陣式組織承諾跨產品與市場的協同效應，但在實務上常因多重匯報、責任模糊與決策遲緩而失控，尤其在跨國環境中更為明顯。有些企業因此放棄矩陣結構；也有如迪士尼般，在強而有力的領導與清晰目標下成功維持混合模式。這提醒企業架構師：結構與治理，與技術同樣重要。

最後，策略與架構都必須隨著變革而演進。但並非所有變革都相同。漸進式變革是在既有信念下持續優化；轉型式變革則重塑組織的價值觀、身份與方向。後者需要的是領導力，而不只是管理能力。全球化企業必須根據時間壓力、環境適配度與內部支持度，審慎選擇變革風格。

企業架構師往往被視為中立的協調者，但在轉型時刻，中立並不足夠。架構師必須協助領導者將願景轉化為可運作的營運模式，使能力與企圖一致，並確保變革在組織與技術層面皆具可行性。

在高度動盪的世界中，真正可持續的競爭優勢，並非來自於在定位與資源、結構與文化、漸進與轉型之間做單選題，而是來自於整合。企業架構正是那條關鍵的連結線，將市場洞察與內部能力、策略與執行、願景與現實緊密串聯。當企業架構發揮其真正價值時，它不只是支援策略，而是在塑造策略本身。

AI Governance for the Enterprise

AI is no longer a peripheral capability in the enterprise. It is rapidly becoming embedded into core business platforms, decision-making processes, and customer interactions. From copilots that assist sales teams to autonomous agents capable of triggering actions across workflows, AI is reshaping how value is created. Yet as AI becomes foundational, governance has not kept pace. Recent analysis of Salesforce’s AI strategy highlights a growing concern across the industry: while vendors race to embed AI into platforms, customers are exposed to new risks around cost predictability, data governance, and operational control.

The current AI landscape is characterised by speed and fragmentation. Large SaaS providers are bundling generative and agentic AI into existing products, often with evolving licensing models that are difficult to forecast over multi-year horizons. At the same time, confidence in AI reliability remains mixed. Even vendors acknowledge that large language models can hallucinate, misinterpret context, or act unpredictably when granted autonomy. For enterprises, this creates a tension between the pressure to adopt AI for competitive advantage and the responsibility to protect customers, data, and financial sustainability.

One of the most immediate risks is data governance. AI systems are only as trustworthy as the data they consume, yet generative AI blurs traditional boundaries around data usage. Sensitive customer, commercial, or operational data can be unintentionally exposed through prompts, model training, or generated outputs if controls are insufficient. For organisations operating in regulated environments, this risk extends beyond reputational damage into regulatory and legal liability. Enterprise architects must therefore treat AI access to data as a privileged operation, governed by the same rigor as access to core transactional systems.

Cost and commercial risk is another emerging challenge. Consumption-based AI pricing, while flexible in theory, introduces significant uncertainty at scale. Analyst warnings about AI licensing structures converting from capped agreements to defined-quantity pricing underscore a broader issue: enterprises may only fully understand their AI cost exposure after adoption is widespread. Without architectural mechanisms to observe, limit, and forecast AI usage, organisations risk budget overruns or unfavourable contract renegotiations at renewal time. This shifts AI governance from a purely technical concern into a strategic financial discipline.

Autonomy introduces a different class of risk. As AI agents are granted the ability to act — not just recommend — the boundary between assistance and decision-making becomes blurred. Automated updates to customer records, workflow escalations, or financial adjustments can amplify errors at machine speed if not governed properly. The absence of human-in-the-loop controls in critical processes can turn isolated model inaccuracies into systemic business failures. For enterprise architects, designing where autonomy is acceptable — and where it is not — is a core governance responsibility.

Compounding these challenges is the rise of shadow AI. Business users increasingly experiment with AI tools outside sanctioned platforms, often with good intentions but little awareness of compliance or security implications. This creates blind spots that traditional IT governance models struggle to detect. AI governance, therefore, cannot rely solely on policy documents; it must be embedded into architecture, tooling, and operational oversight.

In response to this landscape, enterprise-grade AI adoption demands clear architectural principles. First, AI must be mediated through trust and control layers that enforce data classification, anonymisation, encryption, and auditability before any interaction with models occurs. AI should not be treated as a direct consumer of enterprise data, but as a service operating behind controlled gateways that make governance enforceable by design.

Second, automation must remain human-centred. AI should augment human decision-making, not silently replace it in high-impact scenarios. Architectures should explicitly define approval thresholds, escalation paths, and explainability requirements so that responsibility remains clear and defensible. Human oversight is not a limitation of AI maturity; it is a safeguard for organisational resilience.

Third, cost predictability must be engineered, not hoped for. AI usage patterns should be observable in real time, tied to business outcomes, and constrained by access controls that reflect actual value creation. Enterprise architects should collaborate closely with procurement and finance teams to model AI consumption scenarios and ensure contractual terms align with architectural realities.

Finally, AI governance must be treated as a lifecycle capability rather than a one-off initiative. Models evolve, vendors change pricing structures, regulations tighten, and business expectations shift. Governance mechanisms must continuously monitor risk, accuracy, bias, and drift, with clear processes for review, rollback, and remediation. This requires embedding AI governance into existing enterprise disciplines such as architecture review boards, security operations, and compliance assurance.

For Salesforce customers, these principles are particularly critical. As AI becomes more deeply woven into CRM and customer engagement platforms, enterprises must ensure that convenience does not come at the expense of control. AI governance should protect the organisation from unintended data exposure, financial volatility, and operational risk while still enabling innovation and productivity gains.

Ultimately, AI governance is not about slowing adoption. It is about ensuring that AI scales safely, predictably, and sustainably. For enterprise architects, the challenge — and opportunity — is to elevate AI governance to the same level of importance as security, data management, and identity. Done well, it becomes a strategic enabler that allows organisations to embrace AI with confidence, clarity, and trust rather than hesitation and regret.

企業級 AI 治理

AI is no longer a peripheral capability in the enterprise. It is rapidly becoming embedded into core business platforms, decision-making processes, and customer interactions. From copilots that assist sales teams to autonomous agents capable of triggering actions across workflows, AI is reshaping how value is created. Yet as AI becomes foundational, governance has not kept pace. Recent analysis of Salesforce’s AI strategy highlights a growing concern across the industry: while vendors race to embed AI into platforms, customers are exposed to new risks around cost predictability, data governance, and operational control.

The current AI landscape is characterised by speed and fragmentation. Large SaaS providers are bundling generative and agentic AI into existing products, often with evolving licensing models that are difficult to forecast over multi-year horizons. At the same time, confidence in AI reliability remains mixed. Even vendors acknowledge that large language models can hallucinate, misinterpret context, or act unpredictably when granted autonomy. For enterprises, this creates a tension between the pressure to adopt AI for competitive advantage and the responsibility to protect customers, data, and financial sustainability.

One of the most immediate risks is data governance. AI systems are only as trustworthy as the data they consume, yet generative AI blurs traditional boundaries around data usage. Sensitive customer, commercial, or operational data can be unintentionally exposed through prompts, model training, or generated outputs if controls are insufficient. For organisations operating in regulated environments, this risk extends beyond reputational damage into regulatory and legal liability. Enterprise architects must therefore treat AI access to data as a privileged operation, governed by the same rigor as access to core transactional systems.

Cost and commercial risk is another emerging challenge. Consumption-based AI pricing, while flexible in theory, introduces significant uncertainty at scale. Analyst warnings about AI licensing structures converting from capped agreements to defined-quantity pricing underscore a broader issue: enterprises may only fully understand their AI cost exposure after adoption is widespread. Without architectural mechanisms to observe, limit, and forecast AI usage, organisations risk budget overruns or unfavourable contract renegotiations at renewal time. This shifts AI governance from a purely technical concern into a strategic financial discipline.

Autonomy introduces a different class of risk. As AI agents are granted the ability to act — not just recommend — the boundary between assistance and decision-making becomes blurred. Automated updates to customer records, workflow escalations, or financial adjustments can amplify errors at machine speed if not governed properly. The absence of human-in-the-loop controls in critical processes can turn isolated model inaccuracies into systemic business failures. For enterprise architects, designing where autonomy is acceptable — and where it is not — is a core governance responsibility.

Compounding these challenges is the rise of shadow AI. Business users increasingly experiment with AI tools outside sanctioned platforms, often with good intentions but little awareness of compliance or security implications. This creates blind spots that traditional IT governance models struggle to detect. AI governance, therefore, cannot rely solely on policy documents; it must be embedded into architecture, tooling, and operational oversight.

In response to this landscape, enterprise-grade AI adoption demands clear architectural principles. First, AI must be mediated through trust and control layers that enforce data classification, anonymisation, encryption, and auditability before any interaction with models occurs. AI should not be treated as a direct consumer of enterprise data, but as a service operating behind controlled gateways that make governance enforceable by design.

Second, automation must remain human-centred. AI should augment human decision-making, not silently replace it in high-impact scenarios. Architectures should explicitly define approval thresholds, escalation paths, and explainability requirements so that responsibility remains clear and defensible. Human oversight is not a limitation of AI maturity; it is a safeguard for organisational resilience.

Third, cost predictability must be engineered, not hoped for. AI usage patterns should be observable in real time, tied to business outcomes, and constrained by access controls that reflect actual value creation. Enterprise architects should collaborate closely with procurement and finance teams to model AI consumption scenarios and ensure contractual terms align with architectural realities.

Finally, AI governance must be treated as a lifecycle capability rather than a one-off initiative. Models evolve, vendors change pricing structures, regulations tighten, and business expectations shift. Governance mechanisms must continuously monitor risk, accuracy, bias, and drift, with clear processes for review, rollback, and remediation. This requires embedding AI governance into existing enterprise disciplines such as architecture review boards, security operations, and compliance assurance.

For Salesforce customers, these principles are particularly critical. As AI becomes more deeply woven into CRM and customer engagement platforms, enterprises must ensure that convenience does not come at the expense of control. AI governance should protect the organisation from unintended data exposure, financial volatility, and operational risk while still enabling innovation and productivity gains.

Ultimately, AI governance is not about slowing adoption. It is about ensuring that AI scales safely, predictably, and sustainably. For enterprise architects, the challenge — and opportunity — is to elevate AI governance to the same level of importance as security, data management, and identity. Done well, it becomes a strategic enabler that allows organisations to embrace AI with confidence, clarity, and trust rather than hesitation and regret.

What Enterprise Architects Can Learn from IKEA and Global Strategic Management

Strategy is often described as a plan, but in practice it is a long-term commitment to a direction, and a discipline for decision-making. As Enterprise Architects, we sit at the intersection of intent and execution. Our role is not merely to document the strategy, but to translate it into structures, capabilities, platforms, and governance that allow the organisation to succeed over time.

IKEA provides one of the clearest examples of what a strong strategy looks like in practice. Its core idea is deceptively simple: standardized, Swedish-designed, self-assembly furniture at low cost. That simplicity is not accidental, it is the source of its power. By standardising products and processes, IKEA benefits from economies of scale and scope, driving costs down while appealing to a broad customer base. The result is a sustainable competitive advantage that competitors struggle to replicate.

What is striking, however, is not just IKEA’s strategy, but its restraint. As the company expanded internationally, it did not fundamentally alter its core strategy. Instead, it resisted the temptation to over-localise, preserving global coordination wherever possible. This is a lesson many global organisations forget: strategy dilution often happens in the name of flexibility.

Yet IKEA’s early failures in the United States and Japan are equally instructive. Success in one context does not guarantee universality. Cultural norms, living spaces, consumer expectations, and even basic assumptions, such as willingness to self-assemble furniture, varied more than IKEA initially anticipated. The company learned that global strategy is not about rigid uniformity, but about intelligent adaptation. The challenge was not to abandon the core strategy, but to adjust activities at the edges without eroding the centre.

This tension, between global efficiency and local responsiveness, is at the heart of global strategic management. For Enterprise Architects, it translates directly into architectural choices. Which capabilities should be globally standardised? Which should be locally configurable? Where do we draw the line between shared platforms and market-specific extensions? These are not purely technical questions; they are strategic ones.

Strategic management, at its core, is about achieving a sustainable competitive advantage. “Advantage” implies a superior position, “competitive” implies relevance to rivals, and “sustainable” implies durability over time. From an architectural perspective, sustainability is achieved when the organisation’s capabilities, processes, and technologies reinforce each other in ways that are difficult to imitate. Architecture becomes a strategic asset when it encodes these advantages into the operating model.

The strategy-making process is often described as analysis, development, and implementation. In reality, these activities happen simultaneously, especially in a volatile global environment. A perfectly executed analysis can become obsolete overnight, as seen during the 2008 financial crisis or the Greek debt crisis. However, this does not diminish the value of analysis. Without it, decision-making becomes reactive and fragmented.

Environmental analysis remains essential, particularly for global firms. At the macro level, political, economic, social, and technological factors shape the boundaries within which organisations operate. At the industry level, buyers, suppliers, competitors, and intermediaries determine competitive dynamics. Internally, firm resources and capabilities define what is actually possible. Enterprise Architects must be fluent across all three levels, because architecture decisions are constrained, and enabled, by each of them.

Frameworks such as PEST analysis or Porter’s Diamond Model are not ends in themselves. Their real value lies in helping leaders ask better questions. Why do certain countries consistently lead in specific industries? How do demand conditions or supporting industries amplify innovation? And critically, how do these external forces interact with our internal capabilities?

One area where global strategy and architecture intersect most sharply is in sensing and responding to change, especially technological change. Weak signals rarely appear in headquarters reports. They emerge locally, in startups, research labs, customer behaviour, and regulatory shifts. Leading multinational firms recognise this and deliberately distribute their sensing capabilities. Bayer’s use of global research centres and technology scouts is a powerful example. Local intelligence is gathered close to the source, but synthesis and decision-making remain coordinated centrally.

For Enterprise Architects, this highlights an often-overlooked responsibility: designing feedback loops. It is not enough to deploy systems, platforms, or innovation hubs across regions. Information must flow back to the centre in a form that leadership can act upon. Otherwise, decentralised sensing becomes organisational noise rather than strategic insight.

Ultimately, global strategy succeeds when an organisation strikes the right balance between coherence and adaptability. Too much central control leads to rigidity; too much local autonomy leads to fragmentation. Architecture is where this balance becomes tangible. Through shared platforms, clear capability boundaries, and explicit governance, Enterprise Architects help organisations preserve their strategic core while remaining responsive to local realities.

IKEA’s story reminds us that great strategy is not about constant reinvention. It is about clarity of intent, discipline in execution, and humility in adaptation. In a world of increasing complexity, the Enterprise Architect’s role is to ensure that strategy does not remain an abstract ambition, but becomes a living system, scalable, resilient, and unmistakably aligned with the organisation’s long-term advantage.

企業架構師能從 IKEA 與全球策略管理學到什麼

策略常被描述為一份計畫，但在實務上，它更是一種對長期方向的承諾，以及引導管理層做出決策的紀律。身為企業架構師，我們正站在「意圖」與「落地執行」的交會點。我們的角色不只是記錄策略，而是將策略轉化為組織的結構、能力、平台與治理機制，讓企業能夠在長時間內持續成功。

IKEA 提供了一個極具代表性的成功策略案例。它的核心策略看似簡單：以低價格銷售標準化、瑞典設計、自行組裝的家具產品。但這種簡單並非偶然，而是其競爭優勢的來源。透過產品與流程的高度標準化，IKEA 能夠取得規模經濟與範疇經濟，降低成本，同時吸引極廣泛的客群，形成競爭對手難以複製的長期優勢。

然而，真正值得關注的不只是 IKEA 的策略本身，而是它的「克制」。在全球擴張的過程中，IKEA 並未因進入新市場而大幅改變其核心策略。相反地，它刻意避免過度在地化，只在有充分理由時才調整，從而保留全球協同所帶來的效益。這對許多跨國企業而言是一項重要提醒：策略往往不是敗於競爭，而是被內部的妥協與分散所稀釋。

即便如此，IKEA 在美國與日本的早期挫敗同樣具有啟發性。一套在本土市場成功的模式，並不保證在其他市場同樣奏效。文化習慣、居住空間、消費者期待，甚至是對「自行組裝」這件事的接受程度，都存在顯著差異。IKEA 從中學到，全球策略並不等於僵化的一致性，而是有智慧的調整。關鍵不在於放棄核心策略，而是在不動搖核心的前提下，調整周邊活動以回應在地現實。

這種「全球效率」與「在地回應能力」之間的張力，正是全球策略管理的核心課題。對企業架構師而言，這會直接反映在架構決策上：哪些能力應該全球標準化？哪些能力需要保留在地彈性？共享平台與市場客製之間的界線應該畫在哪裡？這些從來都不只是技術問題，而是深層的策略選擇。

從本質來看，策略管理的核心目標是打造可持續的競爭優勢。「優勢」代表更佳的地位或利益，「競爭」意味著相對於對手的比較，而「可持續」則是能長期維持。從架構的角度來看，當企業的能力、流程與科技彼此強化、相互支撐，並且難以被模仿時，競爭優勢才真正具備可持續性。此時，企業架構本身就成為一項策略資產。

策略制定常被描述為三個階段：分析、發展與執行。但在現實世界中，尤其是在快速變動的全球環境裡，這三者往往同時發生。一份再縝密的分析，也可能在一夕之間失去價值——就如同 2008 年金融危機，或 2009 年希臘債務危機所帶來的劇烈轉變。然而，這並不代表分析沒有價值。沒有系統性的分析，決策往往只剩下反應式與碎片化。

環境分析仍然是策略制定不可或缺的一環。從宏觀層面來看，政治、經濟、社會與科技因素形塑了企業運作的邊界；在產業層面，買方、供應商、競爭者與中介者決定了競爭態勢；而在企業內部，資源與能力則界定了企業「能做到什麼」。企業架構師必須能同時理解這三個層次，因為任何架構決策，都是在這些限制與機會之中誕生的。

PEST 分析或波特鑽石模型等框架，本身並不是答案。它們真正的價值，在於幫助管理者提出更好的問題：為什麼某些國家能在特定產業中長期領先？需求條件或相關支援產業如何放大創新能力？而這些外部因素，又如何與企業內部能力相互作用？

在全球策略與企業架構交會之處，有一個特別關鍵的能力：對變化的感知與回應，尤其是科技變化。真正重要的「弱訊號」，往往不會出現在總部的報告中，而是藏在當地市場、新創公司、研究機構、客戶行為或法規變化裡。領先的跨國企業深知這一點，並刻意將感知能力分散到全球各地。拜耳（Bayer）透過全球研發中心與科技偵察員蒐集在地情報，就是一個典型案例。

但分散感知並不代表分散決策。所有資訊最終必須回流至總部，成為可被高層理解與採取行動的洞察。否則，再多的在地觀察，也只會成為組織噪音，而非策略資產。對企業架構師而言，這突顯了一項常被忽略的責任：設計有效的回饋迴路。

最終，全球策略的成功，取決於企業是否能在一致性與彈性之間取得平衡。過度集中會導致僵化，過度分散則會造成碎片化。企業架構正是這種平衡具體化的地方。透過共享平台、清晰的能力邊界，以及明確的治理機制，企業架構師協助組織在回應在地現實的同時，守住策略核心。

IKEA 的故事提醒我們，偉大的策略並不來自不斷推翻重來，而是來自清晰的意圖、紀律化的執行，以及對調整的謙遜態度。在這個高度複雜的世界裡，企業架構師的使命，是確保策略不只是停留在願景層次，而是成為一個能夠擴展、具備韌性，且與長期競爭優勢高度一致的「活系統」。

Why Enterprise Architecture Must Create Urgency, Clarity, and Trust in a Disruptive World

When I joined the organisation as an Enterprise Architect, the expectation was clear and familiar. We would follow a disciplined, strategy-led approach aligned with TOGAF: start from vision and mission, understand business architecture, derive technology architecture, guide solutioning, and govern change through Architecture Development Method cycles. On paper, this approach creates coherence, traceability, and alignment between strategy and execution. In reality, what I observed was a set of deeply embedded organisational challenges that are far more cultural, structural, and behavioural than methodological.

One of the most fundamental issues is the absence of a shared sense of urgency. We are operating in a period of disruptive transformation—tokenisation of assets, real-time settlement, ecosystem-based platforms, and AI-driven operating models are no longer theoretical concepts but active forces reshaping industries. History has shown that organisations do not fail because they lack talent or resources, but because they fail to adapt in time. The stories of Nokia and Kodak are reminders that past success can become a liability when it breeds complacency. Urgency does not mean panic; it means a collective understanding that standing still is itself a strategic decision, and often the most dangerous one.

Instead of initiatives flowing from strategy to architecture and then to execution, many initiatives today originate from project or portfolio management channels and are passed directly to technical teams with a request to “find a solution.” Business objectives are often unclear, implicit, or reduced to a single dimension such as cost savings. Under tight timelines, teams default immediately to convergent thinking, searching for pragmatic, locally workable solutions rather than exploring the problem space. There is resistance to blue-sky thinking, scenario planning, or asking uncomfortable “what if” questions. Energy is spent refining final presentation slides rather than debating options, assumptions, or alternative futures. Design thinking may be referenced, but it is rarely practised meaningfully.

This way of working produces predictable outcomes. Solutions are optimised locally rather than globally. Short-term delivery is prioritised over long-term coherence. Architectural intent becomes reactive instead of intentional. When deadlines are driven by system end-of-life, regulatory pressure, or business launches, tactical fixes crowd out strategic solutions. Even when long-term visions are acknowledged as important, they are perpetually deferred because projects are bounded by fixed timelines and fixed resource allocations. Over time, technical and organisational debt accumulates—not because teams are careless, but because the system incentivises speed over alignment.

Another challenge is that strategy itself is often implicit. It lives in people’s heads rather than in clearly articulated, written, and shared artefacts. Long-tenured staff carry context about why decisions were made, what trade-offs were accepted, and which constraints were temporary. When those people move roles or leave the organisation, that knowledge leaves with them. What remains is a current state shaped by historical decisions and undocumented assumptions. Architects and teams are then forced to reverse-engineer intent from outcomes, a process that is inefficient, fragile, and prone to error.

Innovation is also too often assumed to be an internal activity. Large organisations tend to look inward and backward, reusing existing systems and defending sunk costs. Yet some of the most powerful signals for change come from customers. Their behaviours, frustrations, and workarounds provide direct insight into where operating models and platforms are no longer fit for purpose. When architecture and strategy are disconnected from real customer journeys, innovation becomes abstract rather than actionable.

The difficulty of first-principles thinking compounds these issues. Working-level teams rarely have a holistic view of the enterprise. They operate within functional, system, or domain boundaries and are incentivised to optimise locally. Constraints inherited from past decisions are treated as immutable truths rather than assumptions to be challenged. As a result, organisations repeatedly arrive at local optima while missing better global solutions. Enterprise Architecture exists to elevate thinking beyond these boundaries, but only if it is engaged early and given legitimacy.

Organisational hierarchy further reinforces this problem. Communication is often one-way, flowing downward, while feedback upward is filtered. People at the working level spend a disproportionate amount of time guessing the preferences of senior management rather than searching for the right answer. Risk aversion grows. Decisions are shaped by what feels politically safe rather than what is strategically sound. The organisation gets alignment on appearances, not on outcomes.

Change fatigue is another invisible but powerful force. After repeated rounds of cost cutting, restructures, and transformation programmes, staff become weary. Trust erodes when change feels constant but direction feels unclear. Without co-creating a change vision at the team level, people disengage or comply superficially. Buy-in becomes transactional, and transformation loses momentum before it delivers value.

Even the tools organisations rely on shape behaviour in unhelpful ways. PowerPoint decks and Excel spreadsheets dominate collaboration, fragmenting information across emails and versions. Context is lost, decisions are revisited, and alignment is shallow. Contrast this with the narrative-driven approach popularised by Jeff Bezos, where structured written documents force clear thinking, shared understanding, and meaningful discussion. Good tools do not merely store information; they shape how people think together.

Decision-making itself is often unclear. When choices span multiple teams and stakeholder groups, accountability blurs. Tough decisions stall or are endlessly escalated. Without clear decision rights, governance becomes slow and political. Empowerment and decentralised decision-making, supported by clear architectural principles and guardrails, enable speed without chaos.

Financial opacity further weakens strategic decision-making. Understanding total cost of ownership is surprisingly difficult. Procurement focuses on licence costs, while operational and support costs are fragmented across teams. In matrix organisations, FTE allocation is opaque, and different teams define run costs and change costs differently. Without a shared financial language, application rationalisation and investment decisions become debates about numbers rather than discussions about value.

Many organisations aspire to data-driven decision-making, yet data is often inaccessible, inconsistent, or mistrusted. Before decisions can be data-driven, data must be discoverable, well-defined, and usable by teams. Otherwise, “data-driven” remains a slogan rather than a capability.

Psychological safety is another casualty of prolonged cost pressure. After repeated reductions, people protect their own areas. Silos form as defensive mechanisms. Collaboration declines, and knowledge sharing slows. Architecture, which depends on surfacing risks and challenging assumptions, struggles to function in an environment where people fear being wrong.

Addressing these challenges requires more than refining processes or adopting new frameworks. It requires a re-anchoring of how organisations think about purpose, urgency, and alignment. Strategy must start with “why” and be explicitly articulated, not inferred. Urgency must be created without fear, linking disruptive forces such as tokenisation to real strategic choices and making the cost of inaction visible. Customer journeys should become a central organising construct, providing a shared reference point for prioritisation and investment. A frequently cited example is the digital transformation journey of DBS, highlighted by Harvard Business Review, where success was driven by strong platform foundations, partnership ecosystems, and clear ownership of end-to-end customer journeys.

Organisations must create space for divergent thinking early, before converging on solutions. Scenario planning, “what if” analysis, and option exploration should be treated as risk reduction, not overhead. Roles, decision rights, and engagement models across Enterprise Architects, Solution Architects, Data Architects, and Domain Architects need to be explicit. Financial transparency must be treated as a first-class architectural concern, with standardised definitions of run cost, change cost, and total cost of ownership. Governance should evolve from control to enablement, accelerating good decisions rather than merely preventing bad ones. Collaboration tools should privilege shared narratives and written thinking over slide decks. Data must be made available and trusted before demanding data-driven outcomes. Stakeholder engagement must be intentional, recognising that architecture succeeds through alignment, not mandate.

Enterprise Architecture is not about enforcing frameworks or producing artefacts. It is about creating clarity in complexity, continuity in change, and coherence over time. When urgency is missing, strategy fades. When trust is weak, architecture is sidelined. When purpose is shared and thinking is explicit, architecture becomes indispensable. In a world that will not wait, clarity is no longer optional—and neither is architecture.

為何企業架構必須在顛覆性世界中建立急迫感、清晰度與信任

當我以企業架構師的身分加入組織時，最初的期待是清晰而熟悉的。我們將遵循以策略為導向、符合 TOGAF 的架構方法：從願景與使命出發，理解業務架構，推導技術架構，指導解決方案設計，並透過架構發展方法（ADM）週期來治理與監控變革。理論上，這樣的做法能夠在策略與執行之間建立一致性、可追溯性與長期連貫性。然而在實務中，我所觀察到的挑戰，遠不只是方法論是否被遵循，而是更深層的文化、組織結構與行為模式問題。

其中最根本的問題之一，是組織缺乏一種「共同的急迫感」。我們正身處於高度顛覆性的時代——資產代幣化、即時清算、以生態系為核心的平台模式，以及 AI 驅動的營運方式，早已不是遙遠的未來，而是正在重塑產業的現實力量。歷史一再證明，組織的失敗往往不是因為缺乏人才或資源，而是因為無法及時改變。過去的成功若演變成自滿，反而會成為負擔。急迫感並不等同於恐慌，而是一種集體共識：原地不動並非中立選擇，而往往是最危險的決定。

然而，現實中許多計畫並非從策略出發，再由架構引導執行。相反地，計畫往往由專案或投資組合管理端發起，接著直接交給技術團隊「想辦法找解法」。業務目標往往模糊不清，或被簡化為單一面向，例如成本節省。在緊迫的時程壓力下，團隊從一開始就進入收斂式思考，尋找最務實、最快可行的局部解法，而不是先探索問題空間。對於藍天思考、情境規劃或「如果⋯⋯會怎樣？」這類問題，常會出現抗拒。大量心力被花在完善最終簡報投影片，而非深入討論選項、假設與未來可能性。設計思維或許被提及，但很少真正被實踐。

這種工作方式帶來的結果其實並不意外。解決方案往往只在局部最佳化，而非整體最佳化；短期交付被優先考量，長期一致性則被犧牲；架構從原本應該是「有意識的設計」，退化為「被動回應」。當專案受到系統汰換期限、法規要求或業務上線時程的驅動，戰術性方案便不斷擠壓戰略性解法的空間。即使大家口頭上承認長期目標的重要性，實際上卻總是被延後，因為專案被固定時程與固定資源所綁定。久而久之，技術債與組織債不斷累積，並非因為團隊不夠用心，而是因為制度本身獎勵速度，而非一致性。

另一個常見問題，是策略本身往往是隱性的。它存在於人的腦中，而非清楚書寫、公開共享的形式。資深員工往往掌握關鍵背景：為何當初做出這些決策、接受了哪些取捨、哪些限制只是暫時性的。然而，一旦這些人轉調或離職，知識也隨之流失。留下來的，是由歷史決策與未被記錄的假設所形塑的現況。架構師與團隊只能從結果反推意圖，這不僅低效，也極為脆弱。

創新來源的誤判，也是組織常見的盲點。大型組織容易向內、向後看，重用既有系統，保護沉沒成本。然而，最有力的變革訊號，往往來自客戶。客戶的行為、抱怨與變通方式，清楚揭示了哪些營運模式與平台已不再適用。當架構與策略脫離真實的客戶旅程，創新就會變得抽象，而非可落地。

第一性原則思考在這樣的環境中格外困難。多數第一線團隊缺乏對整個企業的全貌理解，只能在功能、系統或領域邊界內運作，並被激勵去進行局部最佳化。過去的決策與限制被視為不可動搖的前提，而非可以重新檢視的假設。結果是，組織不斷在局部最佳解中打轉，卻錯失更好的整體解法。企業架構的價值，正是在於打破這些邊界，將思考提升至整體層級，但前提是它必須被及早納入，並被賦予正當性。

組織的階層式思維進一步放大了這個問題。溝通往往是單向向下傳遞，向上的回饋則被過濾。第一線人員花費大量精力去猜測高層想聽什麼，而不是專注於尋找正確答案。風險趨避逐漸成為主流，決策依據的是政治安全感，而非策略正確性。表面上看似一致，實際上卻缺乏真正的對齊。

長期的組織變動也帶來明顯的疲勞感。在多輪成本削減、重組與轉型計畫之後，員工容易感到倦怠。當變革看似不斷發生，但方向卻不清晰時，信任便開始流失。如果沒有在團隊層級共同創造變革願景，人們要麼消極配合，要麼表面服從，轉型自然難以產生實質成果。

即使是工具的選擇，也深刻影響了思考方式。大量依賴 PowerPoint 與 Excel，導致資訊分散在無數郵件與版本之中，脈絡容易流失，決策反覆被重談。相較之下，以敘事為核心的書面文件能迫使人們進行清晰思考，建立共同理解，並促進真正的討論。好的工具不只是存放資訊，而是塑造人們如何共同思考。

決策權責不清，也是阻礙之一。當決策橫跨多個團隊與利害關係人時，責任容易模糊，困難的決定被拖延或無限上綱。沒有清楚的決策權限，治理就會變得緩慢且政治化。去中心化的決策，加上清楚的架構原則與護欄，能在不失控的情況下提升速度與彈性。

財務透明度不足，進一步削弱了策略性決策。總持有成本往往難以掌握。採購關注授權費用，但營運與支援成本分散在不同團隊；在矩陣式組織中，人力投入不透明，各團隊對「變更成本」與「營運成本」的定義也不一致。沒有共同的財務語言，應用系統汰換與投資決策就會淪為數字之爭，而非價值討論。

許多組織也高喊資料驅動決策，但資料往往難以取得、定義不一，或缺乏信任。在要求資料驅動之前，資料必須先能被發現、被理解、被使用，否則只會停留在口號層次。

在多次裁員與成本壓力之後，心理安全感也明顯下降。人們傾向保護自己的領域，形成防禦性孤島。協作減少，知識分享放緩。而架構的本質，正是要攤開風險、挑戰假設、探索替代方案，若缺乏心理安全，架構便只能流於形式。

要真正改善這些問題，並非靠流程優化或引入新框架即可，而是需要重新錨定組織對目的、急迫感與對齊方式的理解。策略必須從「為什麼」開始，而不只是「做什麼」，並以明確、可傳承的形式表達。急迫感必須在不製造恐懼的前提下建立，將顛覆性力量與具體策略選擇連結起來，讓不行動的代價清晰可見。客戶旅程應成為重要的組織錨點，提供共同的優先排序與投資依據。廣受引用的 DBS 數位轉型案例顯示，其成功關鍵在於穩固的平台基礎、夥伴生態系，以及對端到端客戶旅程的清楚擁有權。

組織需要在早期就為發散式思考留出空間，將情境規劃與選項探索視為降低風險，而非額外成本。企業架構師、解決方案架構師、資料架構師與領域架構師之間的角色、決策權限與合作模式，必須被明確定義。財務透明度應被視為架構的一部分，建立一致的成本定義。治理應從控制轉向賦能，加速好決策，而不只是避免壞決策。協作工具應支持共同敘事與書面思考，而非只追求簡報美觀。資料必須先被釋放與信任，資料驅動決策才有可能實現。利害關係人的參與必須被刻意設計，因為架構的成功來自對齊，而非命令。

企業架構並不是關於框架或產出文件，而是關於在複雜中創造清晰、在變動中維持連續性、在時間中建立一致性。當缺乏急迫感，策略便會淡化；當信任不足，架構便會被邊緣化；而當目的被共享、思考被顯性化時，架構才能成為真正的放大器。在一個不會等待的世界裡，清晰不再是選項，而是必需，而企業架構亦然。

Corporate Governance Across Emerging Markets

For global investors and owner-managers alike, corporate governance in emerging markets is no longer a peripheral concern. It sits at the heart of capital allocation, risk management, and long-term value creation. China, Brazil, and South Korea offer three distinct but instructive governance stories—each shaped by history, ownership structures, and regulatory choices. Examined together, they reveal how governance frameworks evolve, where tensions persist, and what practical lessons can be drawn for investors and controlling families.

China’s corporate governance system has developed rapidly over the past three decades, largely in tandem with the country’s gradual transition from a planned economy to a market-oriented one. Early reforms in the 1990s focused on corporatizing state-owned enterprises (SOEs) and introducing stock exchanges in Shanghai and Shenzhen. Governance in this phase was heavily state-centric: boards existed, but real authority often rested with government bodies and Party committees. Over time, China introduced a modern company law, independent director requirements, audit committees, and disclosure rules broadly aligned with international practice. Today, governance is shaped by a dual structure. On the one hand, the China Securities Regulatory Commission enforces listing rules, disclosure standards, and corporate governance codes similar in form to those in developed markets. On the other hand, the State-owned Assets Supervision and Administration Commission represents the state as controlling shareholder in major SOEs, influencing board appointments, executive incentives, and strategic direction. For a major investor, the opportunity lies in China’s scale, liquidity, and improving transparency; the risk lies in understanding that control rights, political priorities, and shareholder value do not always align in the same way as in Anglo-American markets. Effective due diligence therefore requires not only financial analysis, but also a clear view of ownership, state influence, and regulatory signaling.

Brazil offers a contrasting governance journey, one driven less by the state and more by capital market innovation. Historically, Brazilian companies were characterized by concentrated ownership, extensive use of non-voting shares, and weak minority protection. In response, Brazil’s stock exchange introduced differentiated listing segments, most notably Novo Mercado, which raised governance standards beyond minimum legal requirements. These reforms demonstrated that better governance can be market-led and value-enhancing. One particularly relevant lesson from Brazil for family-owned enterprises is the role of structured family governance mechanisms, especially the family council. For a large family company now owned by two generations, a family council can serve as a formal forum to separate family matters from business management. The benefits include clearer communication across generations, agreed principles on dividends, succession, and employment of family members, and reduced risk of conflict spilling into the boardroom. It also helps professionalize decision-making without diluting family control. The costs are real but manageable: time commitment, the need for facilitation or external advisors, and the risk that poorly designed councils become symbolic rather than effective. The Brazilian experience shows that when family councils are clearly mandated, linked to but distinct from the board, and focused on long-term stewardship, they can significantly enhance both family harmony and corporate resilience.

South Korea illustrates yet another governance model, dominated by large business groups known as chaebol. Many of the country’s most prominent listed companies—such as Samsung Electronics, Hyundai Motor, and SK Hynix—are globally competitive firms with sophisticated operations, yet their governance has long been shaped by founding-family control. Samsung Electronics provides a useful example. It has strengthened formal governance practices over the past decade by increasing board independence, separating the roles of chair and CEO in practice, enhancing disclosure, and engaging more actively with international investors. At the same time, ultimate control remains closely linked to the founding Lee family through ownership structures and influence within the wider Samsung Group. This creates a hybrid governance model: outwardly aligned with global best practices, but internally anchored in family and group control. For investors, the key insight is not to assume convergence automatically means convergence in substance. Instead, governance must be assessed in terms of how effectively boards can challenge controlling shareholders, manage succession, and balance group interests with those of minority investors.

Taken together, these three cases underline a central theme in emerging-market governance: form is converging faster than substance. Codes, committees, and disclosure frameworks increasingly resemble those of developed markets, yet underlying power structures—state ownership in China, family capitalism in Brazil, and chaebol control in South Korea—continue to shape outcomes. For investors, this means governance analysis must go beyond box-ticking and focus on who really controls strategy and capital. For owner-managers, especially in family firms, the lesson is equally clear: well-designed governance mechanisms such as family councils and independent boards are not constraints, but enablers of continuity, credibility, and long-term value creation.

新興市場的企業管治

對全球投資者與企業擁有人而言，企業管治早已不再只是合規或附屬議題，而是資本配置、風險管理與長期價值創造的核心。中國、巴西與南韓提供了三種截然不同、卻同樣具啟發性的企業管治發展路徑。這些路徑深受歷史、所有權結構與監管選擇所塑造，放在一起觀察，有助我們理解企業管治如何演進、其內在張力何在，以及投資者與控股家族可汲取的實務經驗。

中國的企業管治體系在過去三十年間迅速發展，並與其由計劃經濟逐步轉向市場導向的改革進程緊密相連。九十年代初期的改革重點在於將國有企業公司化，並成立上海與深圳證券交易所。當時的管治模式高度以國家為中心，雖然設有董事會，但實際權力多掌握在政府部門與黨組織手中。隨著制度成熟，中國逐步引入現代公司法、獨立董事制度、審計委員會，以及與國際接軌的資訊披露要求。時至今日，中國的企業管治呈現出一種「雙軌結構」。一方面，中國證券監督管理委員會負責監管資本市場，制定上市規則、披露標準與公司管治守則，在形式上與成熟市場相當接近；另一方面，國務院國有資產監督管理委員會則以國家股東代表的角色，深度介入大型國有企業的董事任命、高管考核與策略方向。對潛在大型投資者而言，中國市場的吸引力在於其規模、流動性與不斷提升的透明度，而主要風險則在於控制權、政治目標與股東價值之間未必完全一致。因此，投資中國不僅需要財務分析，更需要清晰理解企業所有權結構、國家影響力與監管訊號。

相較之下，巴西的企業管治改革更具市場導向色彩。歷史上，巴西企業普遍存在所有權高度集中、非表決權股份盛行、以及少數股東保護薄弱等問題。為回應這些缺陷，巴西證券交易所推出了分級上市制度，其中最具代表性的是 Novo Mercado（新市場），其公司管治要求明顯高於法定最低標準。這一制度證明，企業管治的改善可以由市場力量推動，並對企業估值與投資者信心產生正面影響。對家族企業而言，巴西經驗中特別值得借鑒的是家族管治機制，尤其是家族委員會的角色。對一間由兩代家族成員共同持有的大型家族企業而言，家族委員會可成為一個正式平台，將家族事務與企業經營清楚區分。其好處包括促進跨世代溝通、就股息政策、接班安排及家族成員任用建立共識，並降低家族衝突干擾董事會運作的風險。同時，它有助於在不削弱家族控制權的情況下提升專業化程度。當然，設立家族委員會亦有成本，包括時間投入、可能需要外部顧問，以及設計不當時流於形式的風險。巴西的經驗顯示，只要家族委員會職權清晰、與董事會有所區隔，並專注於長期家族與企業的共同利益，便能同時提升家族和諧與企業韌性。

南韓則展現了另一種企業管治模式，其經濟由大型企業集團（財閥）主導。南韓多家知名上市公司，如三星電子、現代汽車及 SK 海力士，均具備世界級競爭力，但其管治結構長期受到創辦家族控制的影響。以三星電子為例，過去十多年來，公司在形式上大幅強化企業管治，包括提升董事會獨立性、實務上區分董事長與行政總裁角色、改善資訊披露，以及更積極地與國際投資者溝通。然而，在實質層面，最終控制權仍透過股權結構與集團內部影響力，與李氏家族緊密相連。這形成了一種混合式的企業管治模式：對外高度符合國際最佳實務，對內則深植於家族與集團控制之中。對投資者而言，關鍵並非假設形式上的趨同等同於實質上的一致，而是深入評估董事會是否具備挑戰控股股東的能力、是否能妥善處理接班問題，以及如何在集團利益與少數股東權益之間取得平衡。

綜合中國、巴西與南韓的經驗，可以看到新興市場企業管治的一個核心現象：制度形式的趨同速度，往往快於權力結構的實質改變。公司守則、委員會設置與披露框架愈來愈接近成熟市場，但國家所有權、家族資本主義與企業集團控制等深層結構，仍深刻影響實際管治結果。對投資者而言，這意味著企業管治分析必須超越清單式檢查，聚焦於誰真正掌握策略與資本配置權力。對企業擁有人，特別是家族企業而言，清晰而有效的管治機制——如家族委員會與具獨立性的董事會——並非限制，而是確保企業延續性、建立市場信譽及實現長期價值創造的重要工具。