Multicultural research

Ask a room of comms professionals which language matters most in Lakemba and you'll hear Arabic. It's a reasonable answer. It's also wrong.

At the 2021 Census the largest language other than English spoken at home in Lakemba was Bengali, with 3,168 speakers, ahead of Urdu on 2,257 and Arabic on 1,774. Across the suburb, 72.7 per cent of residents speak a language other than English at home, across 64 languages. A campaign built on the reasonable answer misses the largest group in the suburb.

That's what multicultural research is for. Not to confirm what you already believe about a community, but to find the gap between the community you're picturing and the one that's there.

The three questions worth paying to answer

Most research briefs we see are really one of these, and naming which one changes the method.

• Who is actually here? Language, birthplace, generation, arrival cohort, religion, age profile, and how those have shifted across the last three censuses. Desk work, mostly, and cheap relative to what it saves.
• What do they think, and why? Focus groups and in-language surveys that go past the polite answer. This is where fieldwork earns its cost, and where a facilitator who shares the language and the background gets a different conversation than one who doesn't.
• Will this work before we spend the money? Concept testing, message testing and comprehension testing on real creative with the real audience, early enough that the answer can still change something.

Data first, because it is nearly free

Census data will not tell you what a community believes. It will tell you, precisely, who you're talking to, and it will overturn assumptions before you've spent a dollar on fieldwork.

Greater Dandenong: 63.8 per cent of residents speak a language other than English at home, across 138 languages, led by Vietnamese and Khmer. Fairfield: 69.4 per cent, across 125 languages, led by Vietnamese, Arabic and Assyrian Neo-Aramaic. Cabramatta: 79.6 per cent, with 9,250 Vietnamese speakers in a single suburb. All ABS Census 2021, all publicly available, and all of it routinely skipped in favour of a national top-ten language list that describes nowhere in particular.

We've published the whole extract. The language map covers every local government area and 8,434 suburbs, with the 2011, 2016 and 2021 series so you can see direction of travel rather than a single snapshot. Use it whether or not you work with us.

Fieldwork that gets a real answer

In-language focus groups are not English focus groups with a translator in the room. The recruitment is different, the venue is different, the incentive structure is different, and the moderator has to be someone participants will disagree in front of.

That last point decides the quality of the data. A participant who suspects the moderator will report back to the department gives you the answer they think is expected. A participant sitting with someone from their own community, in their own language, in a venue they already visit, tells you the thing you actually needed to hear.

We run focus groups, in-depth interviews, in-language surveys, community consultations and co-design workshops, with facilitators drawn from the communities themselves and translated instruments that have been back-checked rather than assumed.

Testing, before the money is committed

The cheapest research you'll ever buy is the round that stops a campaign from launching in the wrong direction. Concept testing across four language groups costs a fraction of the production budget it protects, and it is consistently the first line cut when a timeline tightens.

What testing surfaces is rarely a translation error. It's a metaphor that doesn't travel, an image that reads as a different age group, a spokesperson who carries authority in one community and none in another, a call to action that assumes internet banking. None of that shows up in a linguistic review. All of it shows up in a room.

Where the findings go

Research that ends in a deck is a cost. Research that changes the brief is an investment. Our reporting is built to be actionable: what we heard, what it means for your message, and what specifically to change, by language group, in the order the changes matter.

From there it feeds straight into multicultural communication, which is where the findings become creative, and into translation, where the terminology decisions get locked into a termbase so the next three years of content inherit them.