The same company in every market

In 1983 Theodore Levitt argued that technology was converging the world's preferences, and that the winners would be companies selling standardised products everywhere at scale.¹ It remains among the most cited and most contested articles in marketing. Kotler was one of its sharper critics within three years, warning that the strategy courted danger by mistaking economies of scale for customer logic.² The empirical work since, most thoroughly Marieke de Mooij's, has been unkind to the strong version of Levitt's thesis: incomes converge, values do not, and consumption keeps reflecting culture rather than income.³

The argument settled into a continuum every marketing course still teaches. At one end, straight extension. At the other, invention. Warren Keegan's ladder between them remains the cleanest statement of the options: extend the product and the message, adapt the message, adapt the product, adapt both, or invent something for the market.⁴

For four decades, where a company sat on that ladder was determined less by strategy than by budget. Adaptation cost money per market, so companies adapted where it was justified and extended into the rest. The cost acted as a forcing function. Somebody senior had to look at a list of countries and decide which ones deserved a real version of the company.

That forcing function has now been removed from part of the mix and left intact in the rest.

The mix no longer adapts at one speed

Take the four Ps, and the three that services marketing adds, and ask what it costs today to make each one properly local.

Promotion has collapsed to nearly nothing. Copy, campaign messaging, sales collateral and social output can be produced in fifteen languages, consistently, against a term list, at roughly the cost of producing one.

Product, for software, has largely followed it. Interface strings, onboarding, empty states, error messages, in-product help and notification email are all text, and they are the text customers actually read most. For physical goods, product adaptation remains as expensive as it ever was: tooling, certification, regulation.

Physical evidence splits the same way. A deck, a template, a contract and a trade-stand graphic are cheap to localise. Packaging is not, because it carries regulated text, supply-chain implications and a higher cost of being wrong.

Price has not moved at all, and it is the P where standardisation does the most visible damage. A price is a positioning statement before it is a number. Purchasing power differs, the competitive set differs, and the anchoring differs. So does presentation: consumer prices in the European Union must be shown as a final price including value added tax and all other taxes,⁵ while the American retail default is the opposite. A company that renders its pricing page fluently into German and leaves the architecture American has announced how much thought went into that market.

Place has not moved either. Which marketplace, which reseller, which distributor, which trade fair, and on what terms, are decisions with contracts attached. Distribution partners also generate their own material about you, which is the largest unmanaged brand surface most companies have.

People have not moved. A salesperson or a support agent in a market is the brand in that interaction, and when they lack usable material they improvise, which is how a voice becomes the category default.

Process has not moved. Onboarding, invoicing, payment terms, how a complaint is handled. These read as trust signals and they differ by market more than most companies realise.

So the mix now divides cleanly. The symbolic layer, what a company says and shows, has become nearly free to adapt. The structural layer, what a company charges, where it sells, how it delivers and who handles the customer, costs what it always did.

The failure that asymmetry produces

Companies respond to relative prices. When one half of the mix becomes free to localise and the other half does not, the predictable outcome is heavy adaptation of the cheap layer and none of the expensive one.

The result is a company that sounds native and behaves foreign.

This is worse than being uniformly foreign, because fluency sets expectations. A customer reading idiomatic copy in their own language, with local references and the right register, reasonably infers a company that has thought about their market. They then meet a price architecture built for somewhere else, a payment method nobody uses locally, a support rota on the wrong time zone, and contract terms that read as aggressive in their jurisdiction.

Service marketing separates promises made, promises enabled and promises kept.⁶ The symbolic layer is where promises are made, and it has become cheap. The other half is where they are kept, and it has not. Widening the first without the second is the definition of a promise gap, and it is now the default outcome of doing nothing.

What is at stake

Brand equity, in the customer-based formulation, is the differential effect that brand knowledge has on customer response to marketing.⁷ The equity is not in the mark. It is in what people know, feel and assume when they meet it.

Which gives the precise statement of the risk. If brand knowledge differs by market, a company does not have one brand with international equity. It has several local brands sharing a logo, and their equity does not aggregate.

Every exposure either pays into one account or opens another, and across markets the filing is done by people with different reference points and category conventions. Hall's distinction between high and low-context cultures is a crude tool and still a useful one: a message that is efficient in a low-context market sounds curt or evasive where more is expected before the point.⁸

A sentence can be translated perfectly and land in a different position in the competitive set. A claim that differentiates in one market is table stakes in another. A price that looks premium in one market looks unserious in another. None of this is an accuracy problem, so none of it is caught by checking accuracy.

The decisions a leader has to make now

The capability is available to every competitor, requires no proprietary asset, and will therefore not differentiate anyone for long. This is operational effectiveness rather than strategy, and the distinction matters because best practice diffuses until everyone occupies the same frontier.⁹ What differentiates is whether the decisions underneath it were taken deliberately. Seven of them matter.

Write down what is invariant and what is adaptable, for every P and not only for copy. The invariants are the promise, the name, the claims that must hold everywhere for legal and strategic reasons, and the proof. The adaptables include idiom, formality, which proof points are chosen, price architecture, channel and service hours. No system can make this distinction. If it is not written down, it is being decided daily by whoever is closest to the work.

Adapt in the right order. What you say should never run ahead of what you have built, in any given market. Where the price, the channel and the support are not local, fluent local copy is a liability rather than an asset, because it raises an expectation the rest of the company cannot meet. The cheapest correct move in an underserved market is often deliberately plainer communication.

Place each market on Keegan's ladder explicitly, and revisit it annually. Not every market earns adaptation. Tier by revenue potential and cultural distance, then record the choice so that it can be wrong and corrected rather than invisible.

Name an owner per market with authority to reject, across the mix. Not a translator. Someone accountable for whether the whole offer makes sense locally, whose objection stops a launch.

Set review cadence by the cost of being wrong. A pricing page, a contract template and a careers page carry different risks and should not share a schedule.

Measure associations, not accuracy. The only test that matters is whether customers in different markets describe the company in the same terms. Ask fifty people per market what the company does and what kind of company it is, then compare the words that come back. Translation quality scores answer a question nobody is asking.

Decide what the company will not say or do anywhere. A refusal list travels better than a claim list, holds across languages, and is the clearest instrument a leader has for keeping a company recognisable as output volume rises.

What stays scarce

The cost of producing symbols has fallen close to zero. The cost of judging whether they are the right symbols for a market has not moved, because it depends on someone who knows both the company and the market well enough to feel the difference. Neither has the cost of actually being present in a market, which is the part that turns a rendered promise into a kept one.

This is what happens whenever a production cost collapses. It does not remove the value from the activity. It concentrates the value in the discrimination step and in whatever did not get cheaper alongside it.

The companies that will look worst over the next few years are not the ones that translate badly. They are the ones that arrive fluently in fifteen markets and turn out, on inspection, to be one company's pricing, one company's contracts and one company's support hours wearing fifteen convincing accents.