Hold the two poles honestly before you dissolve them. Levitt's 1983 provocation had real teeth: technology and travel homogenize wants, price rewards scale, and the company selling one thing to everyone out-economizes the one selling customized things to each. The case for standardization isn't laziness, it's arithmetic. The rebuttal is Ghemawat's: distance still matters, most 'global' firms earn the bulk of their profits in their home region, and the history of international marketing is littered with converged-world bets that local rivals quietly ate. Both poles are strategies of evidence, which means the real question is empirical — market by market, element by element.
CAGE is the distance meter. Cultural: language, taste, religion, aesthetics — the distance that sinks imagery and flavors. Administrative: regulation, tariffs, legal systems, political relations — the distance that decides whether you can even operate, and the one ambitious teams most often discover last. Geographic: physical remoteness, climate, logistics — the distance that reprices supply chains. Economic: income levels, price sensitivity, infrastructure — the distance that redesigns pack sizes and payment plans. The advanced move is to run it per element. For a two-wheeler, cultural distance to Lagos might be modest (mobility needs rhyme) while economic distance is decisive (the buyer is a commercial taxi operator, not a commuter). For a food brand the weights flip. Distance isn't one number. It's a profile.
The integration–responsiveness grid converts that profile into a posture. Low scale pressure and low local pressure: export — ship the home formula, learn cheaply. High scale pressure, low local pressure: global standardization — one product, one playbook (aircraft engines, luxury icons, semiconductors). Low scale, high local: multidomestic — effectively a local company per market, coherent only in the capital account (the classic multinational-FMCG posture of the twentieth century). High and high: transnational — the hard, modern default, where you standardize the backbone (platform, brand assets, quality systems, data) and localize the skin (flavors, price ladders, channels, campaigns). The grid's real value is forcing you to admit that posture is chosen per business, sometimes per category. One company can legitimately run different answers in different divisions.
Decomposition is the working method: list the marketing mix and score each element on its own. What standardizes best: distinctive brand assets (a logo's recognition compounds across borders), core technology platforms, quality standards, positioning at the benefit level. What adapts first: price architecture (income ladders demand different entry rungs, and sachet logic exports well), channels (the kirana, the boda-boda mechanic, the LINE app — distribution is the most local thing in marketing), product formulation at the taste-and-regulation layer, and communication execution (the insight often travels, the idiom rarely does). Bajaj's Boxer shows the scoring in action: brand standardized, platform standardized, but payload, suspension, and service network rebuilt for a buyer who is a business — the motorcycle as a commercial asset earning daily revenue, sold through the mechanics who keep it earning. McDonald's India is the same scoring from the other side: golden arches and service system standardized, the menu rebuilt around a beef-free, vegetarian-heavy market until the McAloo Tikki became a local institution.