Because our market research agency is located in Belgium, we often have to deal with qualitative marketing assignments that require to conduct interviews in different languages. As you may know Belgium itself has 3 official languages (Dutch, French and German) which correspond to different cultural orientations. For firms with a national marketing strategy, it’s therefore important to make sure the market research is carried out taking those cultural differences and the resulting dynamics into account.
Two interview guides
it all starts with designing two separate interview guides. The reason why you have to not merely handle translations is that cultural differences also goes along with sensitivities for certain topics. If you were to study the adoption of a new social media tool in Switzerland for instance, it would be very recommended to make sure you tackle privacy perception issues in the German-speaking part of the country. German cultures have been shown to be correlated to high sensitivity with privacy issues, which may eventually impede the adoption of your online service.
A cultural effect can also be perceived at the country scale. In Belgium for instance the online invoice payment service Zoomit has never been successful (despite huge communication efforts from the banks). Belgians have always been very conservative as far as payments and invoicing are concerned. When we conduct market research on financial services we are always surprised to see how all generations (even younger ones) follow old-fashioned patterns when it comes to payment : printing invoices, annotating them, and sorting them in binders.
One interviewer only
This is yet another question that is often asked: who will handle the interviews in different languages / countries ? Our answer is straightforward: one research only.
Most market research agencies will go the easy way, hiring freelancers in the different countries to quickly do the interviewing part. The advantage is obvious: you can let several interviewers work in parallel which will accelerate the timing of your projet. This has however a big disadvantage. By doing so the “learning effect” will be very much reduced. Imagine you have to do a series of 20 interviews in Belgium (2 languages), The Netherlands and France. You’ll logically carry 5 interviews in each “territory” : 5 in France, 5 in The Netherlands, 5 in Flanders and 5 in Wallonia. You’ll most probably have to handle 4 different interviewers who will not talk to each other, will not share their findings after each interview. Not only will you face inconsistent results (because each interviewer will handle the job differently), but your interviewing guide will not improve along the way and there will be minimal learning effect between each interview.
Only one to code
The last step (which very few market research agencies still propose because it’s a costly one), consists in coding the interviews after they have been transcribed. Coding interviews is already difficult in one language ; it become even more when qualitative research is carried out in different languages / countries. To ensure consistencies and avoid bias, coding should be done by one person only, which assumes that either this person can speak fluently several languages or that interviews have been translated and transcribed. We prefer to go for the second option.
Posted in Marketing.