When most marketing teams talk about personalisation, they mean something specific and relatively narrow: using what you know about an individual — their name, their behaviour, their position in a funnel — to serve them more relevant content at the right moment. Send this segment this email. Show this audience this ad. Surface this content to this user.
This framework works. It's been validated by enough A/B tests and revenue attribution models that it's now the standard operating assumption in most growth marketing functions.
It also assumes, almost entirely, that your audience decides the same way.
That assumption holds in single-market campaigns. It breaks down quickly when your audience spans 100+ nationalities — and the channel itself carries cultural meaning before a single word is read.
What personalisation actually requires
I've spent more than a decade building campaigns that cross cultural boundaries: recruiting students from India, Southeast Asia, Africa, Latin America, and the Middle East. Marketing education products to families in countries where the decision to study abroad is a collective family commitment, not an individual one. Acquiring users for platforms in markets where trust is built through community endorsement rather than institutional credibility signals.
What I learned — often by getting it wrong first — is that personalisation at the cultural level requires something fundamentally different from personalisation at the behavioural level.
Behavioural personalisation asks: what does this person want, and when do they want it? Cultural personalisation asks: how does this person decide, who do they decide with, and what makes them trust the entity asking them to act?
These are not the same question. And most personalisation frameworks are only built to answer the first one.
The channel carries meaning before the message
One of the most consistent findings across multicultural campaigns is that the medium itself carries cultural signal. WhatsApp isn't just a messaging platform in South Asia and West Africa — it's the communication layer for trusted relationships. Email, in some markets, reads as formal and institutional. Instagram credibility works differently in markets where aspirational content is interpreted through collective rather than individual identity.
This means that channel strategy in multicultural marketing is not just a reach and efficiency question. It's a trust question. You can say exactly the right thing in exactly the wrong medium and produce a response that ranges from indifference to active suspicion.
- How trust is built varies significantly across markets — peer testimonials versus institutional rankings versus family-community endorsement are not interchangeable signals.
- How decisions are made — individual versus collective — changes who the effective audience for a campaign actually is. In collective decision cultures, marketing only to the individual student or buyer misses the decision-makers entirely.
- How urgency and value are communicated has cultural registers that don't translate directly. Scarcity messaging that drives action in one market reads as pressure and creates resistance in another.
- What "good" looks like — in creative, in visual language, in tone — is not universal. Design choices that signal quality and credibility in one market signal expense, ostentation, or inauthenticity in another.
Why this doesn't get fixed by translation
The most common mistake I see in multicultural marketing is treating cultural adaptation as a translation problem. Take the campaign, translate the copy, localise the imagery, and ship it.
Translation is necessary. It's not sufficient.
The structure of the campaign — the sequence, the proof points, the call to action, the assumed decision context — is usually built on assumptions that don't survive cultural context. You can translate the words of a campaign that assumes individual decision-making into Mandarin or Hindi or Arabic, and it will still be wrong. Because the assumption underneath the structure hasn't been touched.
Getting this right requires designing from each audience's cultural context upward, not translating downward from a single-market original. That's a fundamentally different process. It's slower. It requires different research. It requires people on your team or in your agency network who have lived in and marketed into these contexts, not just studied them.
What this means for personalisation strategy in 2026
AI tools have made behavioural personalisation dramatically more accessible. Models can now handle segmentation, content variation, timing optimisation, and channel sequencing at a scale that would have required large teams five years ago. That's genuinely useful.
But AI tools are trained on data. And the data that most personalisation models are trained on over-indexes heavily on Western, English-language, individual-decision-maker behaviour. The models are good at what they've seen. They're less reliable on what they haven't.
This creates a specific risk: multicultural campaigns built on AI-personalised frameworks that are superficially more targeted but structurally still wrong. Better segmentation applied to the wrong model of how people decide.
The solution isn't to distrust personalisation. It's to be precise about what you're personalising. Behavioural signals are a starting point, not a complete picture. Cultural context shapes which signals matter, what they mean, and how they should inform the campaign structure — not just the content served within it.
The marketers who will do this well over the next few years are the ones who understand that personalisation is a tool, not a strategy. The strategy requires knowing your audience at a level that data alone doesn't provide.
That knowledge is built through markets, not models.