Originally written in French. Translated by AI — the meaning has been preserved, not the prose.
Angle
A country stands on an infrastructure of everyday practices — how people pay, what they present to identify themselves, whom they trust, which risks they work around, which platforms they already use — and it is that infrastructure, not the language, which decides whether a foreign product becomes usable there. This infrastructure is invisible from outside for a precise reason: it is made of gestures the inhabitants find normal and therefore never mention. An internationalization plan that stops at localization treats the only layer that lets itself be seen from far away, and leaves intact the one that decides.
Synthesis
An executive's sentence serves as the starting point: "We already have the software translated into Portuguese, so we can go after Brazil." It is wrong in an instructive way — it confuses a layer treatable by project with a layer that can only be treated through encounter.
The checkout scene is enough to bring the second one out. Four questions are asked there that nobody asks in France, and each opens onto a different arrangement: credit or debit, and you discover that instalments are an ordinary mode of consumption even for a toy or a pair of shoes; the tax number, and you discover that a national identifier circulates in the most banal purchase; instant payment, and you discover a method so universally adopted that it is no longer asked about, it is presupposed. None of these three answers can be deduced from the language, and none can be guessed from Paris.
What connects these arrangements is that they answer constraints the observer does not have. A payment method that avoids handing over your card answers a cloning risk; one that spares a few percent of commission answers a vital margin; a street economy tolerated and equipped with card terminals answers the absence of a social safety net; a trust founded on "we know him" answers an institutional verification that does not carry. Reading these answers as backwardness or as advance amounts to installing your own situation as the axis of measurement, and depriving yourself of the only useful information: what each of them answers.
The case of Amazon against Mercado Livre gives the thesis its sharpest form. If the most established brand in online commerce, with its power and its translated interface, does not automatically become a country's dominant infrastructure, then the argument "we are leaders elsewhere" carries no weight. The entrant does not meet an empty space: it meets circuits of payment, logistics and trust that are already occupied.
What remains is to explain why this layer so regularly escapes analysis, and the answer lies in its very nature. What most strongly structures an economic culture is what nobody on the ground notices: interviews, studies and documentation describe what gets discussed — the exceptions, the novelties, the problems — and pass over in silence what is taken for granted. The foreigner therefore has an access the inhabitant does not, and that access expires as the gestures become normal for them too. That is what makes ordinary observation, far from the places arranged for the outside gaze, more productive than a survey — and what explains that the only decisive layer is also the only one no plan talks about.
Tensions / contradictions
Two trust mechanisms coexist in the same country without meeting: a strong administrative identification, even for a coffee, and a purely personal trust that opens a gated neighbourhood to a street vendor. Nothing here says which prevails when both apply to the same transaction.
Also unresolved: the tension between the foreign observer's access and their propensity to misread. The very gap in position that lets them see the difference pushes them to rank it.
Questions
- Which categories of product plug into this infrastructure so little that localization really is enough to bring them in?
- How does a team preserve what a newcomer saw during their first weeks, before habit erases it?