We unearthed Australia's favourite brands – and they’re not all what you'd expect

It seems like a simple question: what is your favourite brand, and why? It’s one that Edentify has been asking people as they sign up to our online research community for several years.

But understanding the answers is not a simple task. It’s an open-ended question we have asked to over 6,000 people, covering hundreds of product categories and tens of thousands of brands. I’ve seen many researchers simply ignore this type of rich data as it’s just too hard to analyse.

Recently at Edentify we applied our advanced text mining tools to this vast array of responses to identify our community’s favourite brands and what they like about them.

Text mining allows us to analyse and classify large amounts of open text data in seconds. Almost immediately we could identify the top brands. And it turned up a couple in the top 5 that we might not have expected:

	Nike This is a brand whose image and quality means it is always a favourite
	Apple Appreciated for the functionality and ease of use of its products.
	Coles Great quality products at great prices. Coles has something for everyone.
	Cadbury Who doesn't like chocolate? It's the favourite brand of a favourite product.
	Nescafe The real surprise here. It might not appeal to the coffee snobs, but there are a lot of fans of the taste and value of this instant coffee.

Why should you consider text mining?
Text mining opens up research possibilities. In this case, by implementing text mining to cut down analysis and coding time from hours to minutes, it allowed us to ask an open-ended question. This meant we heard about people’s favourite brands on their own terms. It highlighted those brands that are genuinely close to people’s hearts and the ones that came to mind first in a way that choosing from a long list of brands would not.

Is it really that easy?
Approaching analysis with text mining is certainly easy, but it is not and should not be a fully black-box approach. Automated analysis doesn’t always get it right, so we have built in the ability to over-ride where needed and to adapt and learn for the future. The idea is to automate where possible, then use our most powerful analysis tool – the human brain – to refine the classifications and interpret the insights.

The power of text mining is that it helps us make sense of the best and most natural responses, those in the respondents’ own words. It speeds up the slow, manual work of coding and classifying, giving you more time to spend thinking and understanding.

For more information on how text mining can help you get the most out of open-ended text, contact Edentify.