Rolling out training across an international boutique network

The four points of degradation

A training programme does not degrade evenly between head office and the boutique. It degrades at four identifiable points, and each is treated differently.

Translation. Content translated late, market by market, arrives late, market by market. The subsidiaries that receive their version three weeks after the others start with a handicap nobody makes up, and the sense of being a secondary market settles in for good.

The calendar. When each country picks its own launch date, comparison between markets becomes impossible. The figures from a country that deployed in January do not compare with those from a country that deployed in April, and the analysis stops before it has begun.

The human relay. This is the costliest point and the least visible. Content designed at head office, explained to regional trainers, passed to country trainers, then relayed by a store manager in a morning briefing reaches the advisor in its fourth generation. Every relay is competent and acting in good faith. Each one rephrases, summarises and arbitrates what seems important to him. The final message is not wrong, it has simply become something else.

Measurement. If every market tracks its own indicators in its own tool, there is no way of knowing where the programme works and where it fails. The training department ends up arbitrating without data, which amounts to arbitrating according to whoever speaks loudest in committee.

What a multilingual rollout actually covers

Three very different things hide behind that word, and a provider who does not distinguish them has probably not solved them.

Translating means rendering the same text in another language. It is necessary and it is not sufficient: translated content remains designed for its market of origin.

Localising means adapting what does not translate. Units, date formats, character names, cultural references, sometimes the situations themselves. A client objection that is common in Paris is not necessarily common in Riyadh, and an example that works in Seoul can be opaque in New York.

Deploying simultaneously means making every version arrive at the same time. It is the part the market most underestimates, because it has less to do with translation than with how production is organised. An experience designed from the outset to be multilingual, with the text separated from the mechanics and the visuals, adapts without being rebuilt. One designed in French and translated afterwards has to be reworked for every language.

That initial design is what allows Emraude to deliver up to 20 languages simultaneously, with no limit on the number of translations. The real constraint is not the number of markets: it is the availability of source content and the pace of internal approvals.

Those three operations then meet the realities of the field. Montserrat Garcia, Global Education Executive at Rabanne, discusses a rollout in seven languages in our episode on designing an immersive onboarding experience in luxury retail, along with the workarounds invented market by market, going as far as WhatsApp where the LMS was blocked.

The Sephora case: two populations, one programme

The workplace safety micro-learning built for Sephora illustrates the difficulty well, because it adds a population problem to the geographical one.

The challenge was not only to cover an international network. It was to hold together two very different working realities, the boutique teams and the head office teams. The subject was shared, but a situation handled on the sales floor and a decision taken in a management meeting have neither the same shape nor the same vocabulary.

The solution adopted was not to produce two separate programmes, which would have made the internal culture diverge, nor a single average programme, which would have spoken to nobody. It is two interactive eight-minute mini-games, differentiated by role, sharing the same frame and the same intention. More than 50,000 employees were trained, in 20 languages. The whole scores 4.85/5 for satisfaction, the highest in the portfolio.

The transferable principle sits there: consistency is not obtained by imposing the identical. It is obtained by holding a common frame and letting what must vary vary, meaning the role, the market and the level of experience.

Simultaneous rollout or waves?

Both approaches have their case, and the choice depends above all on what you want to be able to observe.

Simultaneous rollout suits situations where consistency comes first: a global product launch, a change of standards, a campaign in which every market has to be aligned on the same day. It makes comparison between markets immediately readable, since everyone starts at the same moment. It does require everything to be ready everywhere, which shifts the pressure onto the production phase.

Rollout in waves suits you when you want to learn as you go. A pilot market reveals the real friction, and the corrections benefit those that follow. The risk is well known: the markets served last receive a better version, which is good for them and bad for the comparability of the results.

One simple rule helps settle it. If the training concerns a non-negotiable standard, deploy simultaneously. If it concerns a commercial practice that will gain from being refined, proceed in waves and accept that the first markets also serve as a test.

What to require of a provider

Five questions are enough to tell a provider who has already deployed internationally from one who will be discovering it on your project.

Is the text separated from the mechanics? If adding a language means reworking the screens, the cost of every additional market will stay high indefinitely.

How many languages have you delivered simultaneously, and on what project? A theoretical figure is worth nothing. A verifiable reference is worth a great deal.

Who translates, and at what point in the project? The most common answer, a translation provider brought in at the end, is also the most expensive. A language delivered after the fact discovers the constraints of a screen already fixed: text longer than the original does not fit a button drawn for French, a sales situation written in Paris translates exactly and stops being plausible in Riyadh, a character's name is unpronounceable anywhere else. When instructional design, art direction and localisation belong to the same team, those trade-offs are made during the design instead of being inherited after it. Simultaneity is not a feat of translation, it is a consequence of the order in which the work was done. The real cost of localisation, the variation from one market to another and the case of poorly equipped markets are set out in how luxury brands use gamification to train at global scale.

What will I see in the dashboard, market by market? Completion alone does not tell you what failed, or where. Require the drop-off points, the results question by question, the replay rate and the time spent per sequence, and require above all to be able to read them market by market rather than as a network average. An average conceals precisely what you are looking for, which is the country where a sequence does not land, and that is the only place worth intervening.

What happens when a market requests a change after launch? The answer to that question reveals the production model more reliably than any reference.

A frequent and costly mistake

Many maisons treat the rollout as the last stage of the project. They frame the content, approve the pedagogy, sign off the mock-ups, and then turn to the question of markets and languages.

The reverse order is what works. The number of markets, the list of languages and the mode of deployment are design constraints, not end-of-project parameters. Designing for twenty markets from the outset does not produce the same object as internationalising a French project afterwards, and the difference in cost between the two is considerable.

You can see how these constraints translate across the client experience programmes we design. Two articles complete this one: training luxury retail teams sets out the six skills that decide the boutique experience, and measuring the effectiveness of retail training explains how to choose an indicator before launch rather than after.

Frequently asked questions

How long does it take to roll out across an international network?

Allow under two months between framing and going live for a custom programme, translations included. The duration depends far less on the number of markets than on the availability of source content and the pace of internal approvals.

How do you hold local approvals without delaying the whole rollout?

By framing them before production rather than during it. Decide who approves what in each market, what is negotiable and what is not, and set an identical response window everywhere, together with what happens if it is exceeded. A simultaneous rollout rarely fails on translation: it fails because one market replied three weeks after the others and nobody had settled that case in advance.

How do you hold the same service standard from one market to another?

By removing the relays that rephrase. A programme each advisor receives directly, in her own language, at the same moment, transmits the same message everywhere. That is the structural difference with a cascade of trainers, where every tier reinterprets in good faith.

Can a language be added after launch?

Yes, if the experience was designed for it from the outset, which is to say if the text is separated from the mechanics and the visuals. There is no limit on the number of translations. Otherwise, each additional language costs almost as much as the first.

Preparing a rollout across several markets? Request a demo