企業走進海外市場，不是只問「要不要去」，而是要問：用誰的通路、誰擁有資產、誰掌握資料、誰承擔風險，以及情勢改變時能不能退場。這一週會把進入模式、聯盟、模式轉換與退出成本放在同一張決策地圖裡。Foreign entry is not a yes-or-no choice. It asks who owns the assets, who controls customer data, who carries political and partner risk, and whether the firm can change or exit when the market changes. This week connects entry modes, alliances, mode shifts, and exit costs in one decision map.
出口、授權、加盟、代理、策略聯盟、合資、併購、綠地投資（greenfield investment），看起來像一串名詞；實務上，它們是在決定公司要把哪些活動留在自己手上，哪些交給夥伴，哪些先暫時外包，哪些必須一開始就自己做。Exporting, licensing, franchising, agency, strategic alliance, joint venture, acquisition, and greenfield investment are not just textbook labels. They decide which activities stay inside the firm, which activities are shared with partners, and which activities can be outsourced for now.
本週的重點不是背模式，而是練習判斷：什麼時候需要擁有（ownership），什麼時候需要合作，什麼時候需要快速試水溫，什麼時候必須預先設計退出條件。好的進入策略會同時說清楚「為什麼現在這樣進」、「什麼訊號出現時要改」、「最壞情況下如何退」。The goal is not to memorize modes. The goal is to judge when ownership is needed, when cooperation is enough, when a firm should test the market quickly, and when it must design exit conditions before entering. A strong answer explains why this mode fits now, what signal would trigger a change, and how the firm can exit if the case turns bad.
流程可以折疊起來。需要確認自己現在應該產出什麼時，再打開檢查。Keep the flow collapsed until you need to check what you should be producing at that moment.
先看 BYD 匈牙利設廠或 McDonald's 退出俄羅斯的短片。把影片轉成一個決策問題：公司是在選進入模式、調整模式，還是承受退出成本？Watch the BYD Hungary or McDonald's Russia video. Turn it into a decision question: is the firm choosing an entry mode, changing a mode, or paying exit costs?
輸出：一個完整決策問題Output: one clear decision question
把模式放在「控制權」與「資源承諾」的光譜上。不要只背定義，請記住每一種模式交換了什麼。Place modes on a spectrum of control and resource commitment. Do not memorize definitions only. Remember what each mode trades away.
輸出：模式比較表Output: mode comparison table
用交易成本、OLI、資源與動態能力、競爭優勢、國際化順序來看同一個進入決策。每個理論都要變成一個可回答的管理問題。Read the same entry decision through TCE, OLI, RBV and dynamic capabilities, competitive advantage, and internationalization sequence. Each theory becomes one management question.
輸出：理論到問題的配對Output: theory-to-question matching
用一個小題目判斷：如果技術容易被模仿、地主國規則複雜、在地通路很重要，哪一種模式比較合理？為什麼？Use a short scenario to judge: if technology can be copied, host-country rules are complex, and local channels matter, which mode fits best, and why?
輸出：一段可辯護的理由Output: one defensible reason
用 BYD 匈牙利設廠、印度 EV 製造政策、企業退出俄羅斯，看到「進入」其實會受到關稅、在地化要求、政治風險與資產沉沒成本牽動。Use BYD in Hungary, India's EV manufacturing scheme, and corporate exits from Russia to see how tariffs, localization requirements, political risk, and sunk assets shape entry and exit.
輸出：案例到理論的對照Output: case-to-theory mapping
用第六、七集的供應商延誤、風險審查與「急著確認」作為前奏，再接到 Jordan 舊車弊案。重點不是看劇情，而是看：把海外營運交給代理人後，總公司要靠哪些資料、現地查核與契約條款保住控制權？Use Episodes 6 and 7 as the setup: supplier delay, risk review, and rushed confirmation. Then connect them to the Jordan used-car fraud. The point is not plot summary. The point is to ask what data, local verification, and contract clauses protect control when overseas operations are delegated to an agent.
輸出：代理人風險與治理工具表Output: agent-risk and governance table
用三組專家級 Prompt 產出初稿、查證、修正與自我測驗。前面的課中練習不送出；下方 Prompt workflow 的輸出、reflection 與基本學習紀錄會上傳到 Google Sheet。Use the three expert prompts to draft, verify, revise, and test yourself. Earlier in-class practice tasks are not submitted. The Prompt workflow outputs, reflection, and basic learning record below are uploaded to Google Sheets.
輸出：Prompt 學習紀錄與本機備份Output: prompt learning record and local backup
寫下：我現在能分辨的模式差異、我最容易誤判的地方、下一個案例我會先查哪個證據。Write: the mode difference I can now explain, the place where I may still misjudge, and the evidence I would check first in the next case.
輸出：課後 takeawaysOutput: after-class takeaways
這張光譜不是排名。左側速度快、資源承諾低；右側控制強、退出成本高。中間的授權、代理、聯盟與合資，最需要設計治理規則。This spectrum is not a ranking. The left side is faster and lighter. The right side gives more control and higher exit cost. The middle modes need governance design.
交易成本經濟學提醒我們，市場交易不一定便宜。當合約很難寫完整、夥伴可能投機、資產高度專屬、品質難以監控時，「自己做」可能比「交給市場」更合理。這就是所有權問題：擁有不是面子，而是為了降低談判、監督、協調與被綁架的成本。Transaction Cost Economics reminds us that market transactions are not always cheap. When contracts are incomplete, partners may behave opportunistically, assets are relationship-specific, and quality is hard to monitor, ownership may be cheaper than market contracting. Ownership is not about pride. It reduces bargaining, monitoring, coordination, and hold-up costs.
OLI 把海外投資拆成三個條件：所有權優勢（Ownership advantage）、區位優勢（Location advantage）、內部化優勢（Internalization advantage）。如果公司有品牌、技術或流程優勢，目標市場有需求、人才或政策條件，而且把活動留在自己手上比授權更好，FDI 就比較說得通。OLI breaks foreign investment into three conditions: Ownership advantage, Location advantage, and Internalization advantage. FDI makes stronger sense when the firm owns a brand, technology, or process advantage, the target location offers demand, talent, or policy benefits, and internal control is better than licensing.
資源基礎觀點問：公司的優勢是否有價值、稀有、難模仿，而且被組織有效使用？動態能力再往前一步問：環境變動時，公司能不能感知、抓住機會、重組資源。海外進入不是複製貼上，重點是能力能不能在新市場重新運作。The Resource-Based View asks whether a firm's resources are valuable, rare, hard to imitate, and organized for use. Dynamic capabilities go further: can the firm sense change, seize opportunities, and reconfigure resources? Foreign entry is not copy and paste. The question is whether capabilities can work again in the new market.
企業進入海外市場，有時是利用既有優勢（exploit）：把成本、品牌、技術、營運流程帶到新市場。有時是為了取得新能力（learn or augment）：靠近人才、供應鏈、科技聚落或高階客戶。兩種邏輯不同，進入模式也不同。A firm may enter a foreign market to exploit an existing advantage: cost, brand, technology, or operating process. It may also enter to learn or augment capabilities: talent, suppliers, technology clusters, or demanding customers. These two logics lead to different entry modes.
Uppsala 學派強調逐步累積市場知識，從心理距離較近、承諾較低的市場開始，再逐步加碼。Born Global 則提醒我們，某些數位、利基、科技或網絡型企業可以很早就跨國經營。差異不是誰比較先進，而是公司如何取得市場知識與降低不確定性。The Uppsala model emphasizes gradual learning: firms start with closer markets and lower commitment, then increase commitment over time. Born Global firms remind us that some digital, niche, technology, or network-based firms can internationalize early. The difference is how firms acquire market knowledge and reduce uncertainty.
退出不是失敗的同義詞。退出可能是政治風險上升、制裁、夥伴衝突、需求消失、匯率重挫、聲譽風險或策略重心改變。重資產企業通常比輕資產企業更難退，因為工廠、設備、人員、供應鏈與政府關係都可能形成高昂的退出成本。Exit is not the same as failure. A firm may exit because political risk rises, sanctions hit, partners conflict, demand disappears, currency collapses, reputation risk grows, or strategy changes. Heavy-asset firms usually face harder exits than asset-light firms because factories, equipment, people, supply chains, and government relationships become costly to unwind.
| 理論鏡頭Theory Lens | 在實務上問什麼？Practical Question | 要找什麼證據？Evidence to Check | 對模式選擇的影響Mode Implication |
|---|---|---|---|
| TCE | 把活動交給夥伴，會不會比自己做更貴、更難管？Will using partners become more expensive or harder to control than doing it internally? | 合約複雜度、資產專屬性、品質監控難度、夥伴可替代性。Contract complexity, asset specificity, quality monitoring difficulty, partner replaceability. | 交易成本高時，合資、併購或自建子公司更可能合理。High transaction costs push toward JV, acquisition, or wholly owned operations. |
| OLI | 公司有什麼優勢？目標地有什麼吸引力？為何要內部化？What advantage does the firm own, what location advantage exists, and why internalize? | 品牌、技術、成本、區位政策、市場需求、授權風險。Brand, technology, cost, location policy, market demand, licensing risk. | O、L、I 同時強，FDI 說服力較強；只有 O 強，授權或出口也可能足夠。When O, L, and I are all strong, FDI is stronger. If only O is strong, licensing or exporting may be enough. |
| RBV / DC | 核心能力能不能移轉、保護、重組？Can the core capability be transferred, protected, and reconfigured? | 流程可複製性、人才與訓練、資料回流、當地調整能力。Process transferability, talent and training, data feedback, local adaptation capability. | 能力敏感且難以外包時，控制權需求上升；需要在地能力時，聯盟或合資有價值。Sensitive capabilities raise control needs. Need for local capability gives alliances or JVs value. |
| 優勢利用Advantage Exploitation | 公司是把既有優勢帶出去，還是要在海外取得新能力？Is the firm taking an existing advantage abroad or seeking new capabilities overseas? | 母國優勢、市場差異、競爭者反應、客戶接受度。Home-market advantage, market differences, competitor response, customer acceptance. | 純粹利用優勢可從低承諾開始；尋找新能力通常需要更深入的在地嵌入。Pure exploitation can start lighter. Capability seeking usually requires deeper local embedding. |
| Uppsala / Born Global | 公司應該逐步加碼，還是一開始就跨多市場？Should the firm commit gradually or internationalize early across markets? | 市場知識來源、網絡、數位化程度、心理距離、資金壓力。Market knowledge sources, networks, digital intensity, psychic distance, funding pressure. | 知識不足時先低承諾；平台或利基企業可更早跨國，但更要設計學習迴路。Low knowledge supports low commitment. Platform or niche firms may go early but need fast learning loops. |
| Exit Mode | 如果要退，最難拔掉的是什麼？If the firm exits, what is hardest to unwind? | 資產沉沒成本、員工、供應商、品牌、制裁、買家、政府關係。Sunk assets, employees, suppliers, brand, sanctions, buyers, government relations. | 退出成本高時，初始承諾要更審慎，或用分階段投資保留轉圜。High exit cost calls for more cautious initial commitment or staged investment. |
每一個理論都要推向可行動的輸出。只寫「這是 OLI」還不夠，必須能回答模式、備選方案、轉換訊號、治理與退出。Each theory must lead to an action-ready output. Saying "this is OLI" is not enough. The answer must cover mode, backup mode, trigger, governance, and exit.
先觀察：BYD 在歐洲不是只賣車，也在建構在地製造、供應鏈與品牌信任。Watch for this: BYD is not only selling cars in Europe. It is building local manufacturing, supply chains, and trust.
先觀察：退出不是按下停止鍵，而是處理資產、員工、品牌、買家與聲譽。Watch for this: exit is not a stop button. It means handling assets, employees, brand, buyers, and reputation.
BYD 2023 年宣布在匈牙利 Szeged 建造歐洲第一座乘用車工廠，官方新聞稿強調在地就業、供應鏈與歐洲生產中心。2024 年 10 月，歐盟對中國製電動車課徵五年期反補貼稅，BYD 稅率為 17.0%。這不是單純的「海外設廠」故事，而是展示進入模式會如何被關稅、品牌信任、供應鏈與在地化壓力推動。BYD announced in 2023 that it would build its first European passenger-car factory in Szeged, Hungary. The official release emphasized local jobs, supply chains, and a European production center. In October 2024, the European Commission imposed five-year countervailing duties on China-made BEVs, including 17.0 percent for BYD. This is not a simple factory story. It shows how tariffs, trust, supply chains, and localization pressure can push entry-mode change.
印度 SPMEPCI 要求核准申請者至少投資 Rs.4,150 crore（約 5 億美元）建立電動乘用車製造設施，並在三年內達成 25% 的本地附加價值，五年內達成 50%。符合條件者可在一定限制下用 15% 關稅進口高價 EV。這提醒我們：政策不是背景，而是進入模式設計的一部分。對想進印度的 EV 企業來說，單純出口、代理銷售與在地製造不是自由排序，政策會改變各方案的成本與可行性。India's SPMEPCI requires approved applicants to invest at least Rs.4,150 crore, about USD 500 million, in e-4W manufacturing facilities, reach 25 percent domestic value addition within three years, and 50 percent within five years. Approved applicants may import qualifying EVs at a 15 percent duty under limits. Policy is not background. It is part of entry-mode design. For an EV entrant, exporting, agency sales, and local manufacturing are not freely ranked options because policy changes their costs and feasibility.
McDonald's 2022 年宣布出售俄羅斯業務，預估記錄約 12 到 14 億美元的主要非現金費用，並進行「de-Arching」，也就是移除 McDonald's 名稱、標誌、品牌與菜單。BP 則在俄烏戰爭後宣布退出 Rosneft 19.75% 持股，後續揭露俄羅斯相關費用達 255.20 億美元。兩者都在退場，但重資產能源投資與餐飲品牌退出面對的資產、政治與會計壓力完全不同。McDonald's announced in 2022 that it would sell its Russian business, expecting a mainly non-cash charge of about USD 1.2 to 1.4 billion and a "de-Arching" process that removed the McDonald's name, logo, branding, and menu. BP announced it would exit its 19.75 percent Rosneft shareholding, and later disclosed Russia-related charges of USD 25.520 billion. Both firms exited, but heavy-asset energy investment and restaurant-brand exit carry very different asset, political, and accounting pressure.
McDonald's official exit release | BP Rosneft exit release | BP 2022 results
策略聯盟的價值通常來自互補：一方有技術或品牌，另一方有通路、制度知識、客戶關係或製造能力。問題是互補也會帶來依賴，學習也可能變成學習競賽。判斷聯盟時，不只看夥伴是不是有名，而要看資料權、決策權、技術邊界、績效指標與退出條款是否清楚。Strategic alliances create value through complementarity: one side may bring technology or brand, while the other brings channels, institutional knowledge, customer relationships, or manufacturing. But complementarity also creates dependence, and learning can become a learning race. Judge alliances by data rights, decision rights, technology boundaries, KPIs, and exit clauses, not by partner fame alone.
| 補充案例Additional Case | 最適合理論鏡頭Best Theory Lens | 學生可以怎麼問？Student Question | 來源Sources |
|---|---|---|---|
| TSMC Arizona / ESMC Germany | OLI, TCE, Ownership | 半導體先進製程為何需要高控制權？德國 ESMC 合資與美國 Arizona 自建廠，分別解決了什麼區位與治理問題？Why does advanced semiconductor manufacturing require high control? What location and governance problems do the Arizona fab and Germany's ESMC JV solve differently? | NIST CHIPS: TSMC Arizona | TSMC ESMC release |
| Disney-Reliance India JV | TCE, RBV, Mode Shift | 如果 Disney 在印度從全資控制轉向合資，它是在降低交易成本、取得在地能力，還是重新分配風險？哪些說法是官方事實，哪些只是我們的理論推論？If Disney shifts from full control to a JV in India, is it reducing transaction costs, gaining local capability, or redistributing risk? Which statements are official facts and which are our theory-based inferences? | Disney official release | India CCI approval |
| LEGO Vietnam factory | OLI, Competitive Advantage Exploitation | LEGO 為何用自有工廠複製精密製造、供應鏈與永續製程，而不是只授權或外包？品質一致性在這裡扮演什麼角色？Why does LEGO use its own factory to replicate precision manufacturing, supply chain, and sustainability processes instead of only licensing or outsourcing? What role does quality consistency play? | LEGO official release | AP News |
| Costco China vs Temu | Uppsala vs Born Global | Costco 的中國展店較像逐步累積市場知識；Temu 的跨國擴張較像早期快速國際化。兩者差異來自商業模式、資產型態、資料回流，還是供應鏈？Costco's China store rollout looks closer to gradual learning, while Temu's international spread looks closer to early rapid internationalization. Does the difference come from business model, assets, data feedback, or supply chain? | Costco annual report | PDD annual reports |
退出成本常常在進入時就被寫進去了。重資產、制度風險高、買家少、品牌能見度高的模式，退出壓力通常更大。Exit cost is often built into the entry decision. Heavy assets, high institutional risk, few buyers, and visible brands usually make exit harder.
先把位置說清楚：第六、七集不是 Jordan 舊車弊案的完整爆點，而是前奏。第六集讓你看見供應商延誤、關係壓力與契約治理；第七集讓你看見風險審查、資料查核與「先當作確認」的危險。後續 Jordan 舊車案把這些問題推到更嚴重的代理人風險：當海外現地資訊掌握在代理人、地方夥伴或分公司手上，總公司如果只看文件、電話與表面承諾，很容易把交易成本低估。First, locate the material correctly: Episodes 6 and 7 are not the full Jordan used-car fraud. They are the setup. Episode 6 shows supplier delay, relationship pressure, and contract governance. Episode 7 shows risk review, information checks, and the danger of treating an assumption as confirmed. The later Jordan used-car case raises the same issues to a more serious agency problem: when local information sits with an agent, local partner, or branch office, headquarters can easily underestimate transaction costs if it relies only on documents, phone calls, and surface promises.
這一段要練的不是「誰做錯了」，而是「哪一種進入模式會把哪些控制權交出去」。代理、經銷、聯盟與合資都可能帶來速度與在地知識，但也會讓公司暴露在資訊不對稱、機會主義、雙重合約、回扣與品質失控的風險裡。The question is not simply who made a mistake. The question is what control rights are given away by each entry mode. Agency, distribution, alliances, and JVs can bring speed and local knowledge, but they also expose the firm to information asymmetry, opportunism, dual contracts, kickbacks, and quality-control failures.
| 影片片段Video Moment | 對話線索Dialogue Cue | 學生要看什麼What to Watch | 對應理論Theory Link |
|---|---|---|---|
| E06 00:31:11-00:32:24 | 「keep calling」式的追單"keep calling" follow-up | 供應商延誤時，只靠電話催促很難知道真實原因。問題不是對方有沒有故事，而是能不能把故事轉成可查核資料。When a supplier delays, phone follow-up does not reveal the real cause. The issue is not whether the story sounds human, but whether it can be converted into verifiable evidence. | TCE, monitoring cost |
| E06 00:48:59-00:50:40 | 「by the book」與長期關係"by the book" vs. relationship | 長期往來可以降低搜尋成本，也可能讓人放鬆程序。治理的重點是把信任變成規則，而不是用信任取代規則。Long relationships can reduce search cost, but they can also weaken process discipline. Governance turns trust into rules. It does not replace rules with trust. | Relational governance, incomplete contract |
| E06 00:53:40-00:54:00 | forwarding agency、shipment dates、credit limitforwarding agency, shipment dates, credit limit | 這些不是行政細節，而是控制點。能否查到貨代、船期與信用額度，決定公司能不能把海外交易從「聽說」拉回「可驗證」。These are not clerical details. They are control points. Forwarder, shipment schedule, and credit limit move a foreign transaction from hearsay to verification. | Internalization, due diligence |
| E07 00:29:44-00:30:43 | 「skimmed through」的確認"skimmed through" confirmation | 風險審查如果只求趕上進度，最需要確認的資料反而被略過。這是進入模式常見的時間壓力：速度越快，查核設計越要清楚。If risk review is rushed, the most important items may be skipped. Entry speed increases the need for clear verification design. | Uppsala learning, risk analysis |
| E10 00:48:25-00:51:24 | executive list 與身分查核executive list and identity check | Jordan 舊車案開始露出身分與職稱不一致。當代理人或地方公司說自己能代表市場時，第一步不是談條件，而是查清楚誰真的有權承諾。The Jordan used-car case begins to reveal inconsistent identity and titles. When an agent or local firm claims market access, the first step is not bargaining. It is checking who has authority to commit. | Agency risk, ownership boundary |
| E13 00:08:30-00:08:49 | dual contract 與 commissiondual contract and commission | 雙重合約與佣金不是小瑕疵，而是治理失靈。它會讓價格、責任、利潤與法律風險都失真。Dual contracts and hidden commission are not minor flaws. They distort price, accountability, profit, and legal risk. | Opportunism, exit cost |
代理人不是「幫我做事的人」而已。代理人會替公司接觸市場，也會替公司遮住市場。進入模式越依賴代理人，越要把身分、信用、船期、現地查核與合約一致性變成固定流程。An agent is not just someone who helps the firm. An agent gives market access and can also block market visibility. The more the entry mode depends on agents, the more identity, credit, shipment, local visit, and contract consistency must become routine controls.
交易成本經濟學會問：如果公司不能完整監督代理人，合約又寫不完，外包是否真的便宜？OLI 會問：這個交易是否該內部化，還是可以用授權、代理或聯盟處理？資源基礎與動態能力會問：公司是否有能力在海外重新建立查核、信用評估與現地學習流程？退出模式則提醒我們：當舞弊發生後，退出不是停止合作這麼簡單，還包括追償、信用保險、客戶關係、品牌修復與內部問責。TCE asks whether outsourcing is really cheap when monitoring is weak and contracts are incomplete. OLI asks whether the transaction should be internalized or handled through licensing, agency, or alliance. RBV and dynamic capabilities ask whether the firm can rebuild verification, credit assessment, and local learning overseas. Exit mode reminds us that fraud does not end when cooperation stops. It can require recovery, trade insurance, customer repair, brand repair, and internal accountability.
請把一個海外代理交易拆成四欄：代理人聲稱能提供什麼、總公司實際能查到什麼、如果查不到會造成什麼風險、要用哪一個條款或流程補上。至少要包含：身分與授權、買方與地址、信用與付款條件、貨代與船期、佣金與利益衝突、終止與追償條款。Split an overseas agency transaction into four columns: what the agent claims to provide, what headquarters can actually verify, what risk appears if it cannot verify, and which clause or process closes the gap. Include identity and authority, buyer and address, credit and payment terms, forwarder and shipping schedule, commission and conflict of interest, termination and recovery clauses.
把「延遲出貨、身分不一致、雙重合約、現地查核不足」分別配到 TCE、OLI、RBV/DC、Exit Mode。每個答案都要寫一句理由。Match delayed shipment, inconsistent identity, dual contract, and weak local verification to TCE, OLI, RBV/DC, and Exit Mode. Add one reason for each match.
若代理商無法提供可查證買方資料，但市場需求看起來很大，公司應該：A 繼續談、B 降低出貨量、C 要求第三方查核與信用保險、D 直接改成綠地投資。請選一個最合理的下一步。If an agent cannot provide verifiable buyer data but demand looks large, should the firm: A continue negotiation, B reduce shipment volume, C require third-party verification and trade insurance, or D switch directly to greenfield investment? Choose the best next step.
什麼訊號出現時，代理或經銷模式應該轉成合資、在地辦公室，或暫停退出？請至少列出三個訊號。What signals should move a firm from agency or distribution to JV, local office, or temporary exit? List at least three signals.
可用來快速回顧第六集的商業關係、銷售壓力與職場脈絡。Use it to review Episode 6's business relationship, sales pressure, and workplace context.
可用來回顧第七集的風險審查、資料確認與提案競爭。Use it to review Episode 7's risk review, information checks, and proposal competition.
可搭配 Jordan 舊車案的身分查核與代理人疑點。Use it with the identity-check and agent-risk part of the Jordan used-car case.
可搭配後續提案、現地支援與舊車案治理教訓。Use it with the later presentation, local support, and governance lessons from the used-car case.
作為劇集背景與集數資訊參考；正式分析仍以課本、字幕片段與可靠商業資料為主。Use it for show background and episode information. Formal analysis should still rely on textbook concepts, subtitle moments, and reliable business sources.
把以下問題配對到 TCE、OLI、RBV/DC、優勢利用、Uppsala/Born Global、Exit Mode。Match each question with TCE, OLI, RBV/DC, Advantage Exploitation, Uppsala/Born Global, or Exit Mode.
一家台灣設備公司想進入法規複雜、政府採購關係重要的市場。產品技術很敏感，售後維修必須穩定，當地客戶要求長期服務。下列哪一個初始模式最能兼顧風險與學習？A Taiwanese equipment firm wants to enter a market with complex regulation and important government procurement relationships. The product technology is sensitive, after-service must be stable, and customers require long-term support. Which initial mode best balances risk and learning?
參考答案：3 通常較合理，但答案必須說明為什麼不是 1、2、4。若證據顯示市場穩定、買家明確且服務資產可取得，4 也可能變得合理。Suggested answer: 3 often fits better, but the answer must explain why not 1, 2, or 4. If evidence shows stable demand, clear buyers, and available service assets, 4 may become reasonable.
畫一個 2x2 矩陣。橫軸是「控制權需求」，縱軸是「市場不確定性」。把 exporting、licensing、agency、alliance、JV、greenfield 放入象限。每個象限寫一個可能失敗原因。Draw a 2x2 matrix. Horizontal axis: need for control. Vertical axis: market uncertainty. Place exporting, licensing, agency, alliance, JV, and greenfield into the quadrants. Add one likely failure reason in each quadrant.
情境 A：一家品牌零售商在某國多數店點為加盟，政治風險升高，但當地需求仍在。情境 B：一家能源公司持有大型合資股權，制裁讓未來現金流高度不確定。請分別選擇退出選項，並說明最大成本。Scenario A: a branded retailer operates mostly through franchisees in a country where political risk rises but demand remains. Scenario B: an energy company holds a large JV stake and sanctions make future cash flow uncertain. Choose an exit option for each and explain the largest cost.
用 BYD 案例寫出 OLI 三欄表。每一欄都必須至少有一個來源連結。Build an OLI table for BYD. Each column must include at least one source link.
為一個策略聯盟設計三個治理條款：資料權、決策權、退出條件。Design three alliance governance clauses: data rights, decision rights, and exit conditions.
找出一個「初始進入模式需要轉換」的訊號，並說明要轉成哪一種模式。Identify one signal that the initial entry mode should change, and state the next mode.
下列 Prompt 的重點，是讓你看到一組專業指令如何同時指定情境、角色、證據規則、理論架構、輸出格式與反問。使用 NotebookLM 時，先上傳課本章節、案例與課堂資料；使用一般生成式 AI 時，必須要求它引用可靠學術資源、開放取用資料庫、公司正式揭露、政府或國際機構網站，不能只依賴模型內建記憶。The prompts below show how a professional prompt specifies context, role, evidence rules, theory lenses, output format, and challenge questions. In NotebookLM, upload textbook chapters, cases, and class materials first. In a general GenAI tool, require reliable academic sources, open access databases, official company disclosures, government sources, or international organizations. Do not let the model rely only on its built-in training data.
選擇一家公司與一個目標市場，寫出一頁備忘錄。內容必須包含：建議模式、第二選項、不建議模式、理論依據、兩個案例證據、三個待查證資料，以及退出條件。Choose one firm and one target market. Write a one-page memo covering: recommended mode, second-best option, not-recommended mode, theory basis, two case evidence points, three items to verify, and exit conditions.
不要寫成「我覺得」。請寫成「在目前證據下，A 比 B 更合理，因為...；但如果...，模式應該調整為...」。Do not write "I think." Write: "Based on current evidence, A is more reasonable than B because...; however, if..., the mode should shift to..."
用來理解交易成本、公司邊界與資產專屬性。Use this for transaction costs, firm boundaries, and asset specificity.
理解 Ownership、Location、Internalization 的原始脈絡。Read for the original context of Ownership, Location, and Internalization.
動態能力觀點的經典論文資訊頁。A reference page for the classic dynamic capabilities article.
理解 Born Global 不是口號，而是需要明確衡量的國際化現象。Shows that Born Global needs careful measurement, not slogans.
補充 Uppsala 模型與心理距離概念。Useful background for the Uppsala model and psychic distance.
查全球 FDI 與跨國投資趨勢。Use it for FDI and cross-border investment trends.
比較不同國家外資限制。Compare restrictions on foreign investment.
看政策如何重塑進入模式。Study how policy reshapes entry-mode choices.
BYD 在歐洲在地製造的官方資訊。Official information on BYD's localized European production.
查歐盟對中國製電動車反補貼稅的正式說明。Official statement on EU countervailing duties on China-made BEVs.
用來比較品牌、員工與店點退出成本。Use for brand, employee, and store-exit costs.
查不同企業退出、暫停或維持俄羅斯業務的分類。Compare how firms withdrew, paused, or remained in Russia.
補充進入模式快速複習。Quick review of entry strategies.
補充 OLI 架構快速複習。Quick review of the OLI framework.
前面的課中練習不送出；本區只上傳 Prompt workflow 的輸出結果、reflection 與基本學習紀錄。你也可以另外下載 JSON / Markdown 留存。Earlier in-class practice tasks are not submitted. This section uploads only Prompt workflow outputs, reflection, and basic learning-record fields. You can also download JSON / Markdown for your own copy.