Tom H. C. Anderson - More Than Market Research

Today Seth Grimes, Chairman for the Text Analytics Summit and contributing editor and Breakthrough Analysis columnist for Intelligent Enterprise magazine published an article which dealt with a text analytics project Anderson Analytics conducted for Unilever recently. You can read or listen to the article here b-eye (Business Intelligence Network).

I think the Unilever Dove, pro-age project was an interesting application of text analytics on web based data, so I thought it was well worth a posting here. After all, “Customers are always right” but can a company listen to everything their customers have to say?

More than ever, companies today have far greater access to the voices of consumers. Millions of people make their opinions heard daily by participating in online forums and blogs. And for companies that place value on listening to their customers, these voices are likely to be heard by the right people.

In order to understand the consumer reaction to the recent pro•age marketing campaign, we helped Unilever employ text analytics to understand what their customers were saying.

Pro•age is a Unilever product line designed for the mature woman. In March, the refusal by network TV to air its commercial which tastefully portrayed nude women over the age of 50 was covered heavily by the media, particularly on Oprah and Entertainment Tonight. Since then, Unilever opened an online forum so that women would be able to view the commercial in its entirety online and then discuss their feelings about the pro•age campaign and other age and beauty related issues.

Unilever engaged us to help them understand the volume of text posted by thousands of women on their forum. At the time of the study, the pro-age section of forum had already received well over 3,000 posts. Anderson Analytics used our proprietary text analytics methodology that combines qualitative research, text mining and psychological analysis. This methodology uses state-of-the-art tools to identify meaningful word combinations, to visualize text data, and to offer analysis that measures over 40 psychological traits.

The text analysis process looks for common themes throughout the online posts and also identifies unique discussions that typically are likely to be dismissed by human analysts. For instance we found that among those expressing an opinion about the ad, 97% strongly supported pro•age, and less than 7% indicated concern in regard to the nudity. By analyzing the text data, we also came to the conclusion that the Dove and the pro•age brand are being viewed by many as a champion for the cause of women over the age 50.

While we weren’t surprised to see positive discussions about aging, beauty and the appropriateness of the TV commercial, we also found many other topics dealing with inter-generational issues, comments on American society and the world at large, and intense contrast and comparison between the US and European media were also prevalent on the forum.

This is the true power of text analytics and text mining. It does not presume to know and offer predetermined options for respondents to select. Instead, it finds order and insight within vastly unstructured and unfiltered information.

Text mining will allow Unilever to gain a much more complete, 360 degree, view of their customers ensuring that not just a small sample of customers are listened to, but rather each of the thousands of women posting their comments will be ‘heard’. The findings in the study will help Unilever continue to better understand American women and offer them products and messaging they want.

As Catherine Catherine Cardoso, Associate Insights Manager at Unilever, said “Text analytics is a new methodology for us, we were very pleased with the results and the depth of insight. The results were helpful beyond understanding reactions to our campaign. We also gained an understanding of what motivates people on discussion boards, which issues are most important to women in our target group, and how to create better products and messaging for them.” Catherine also sees great potential for text analytics and said, “We’ve been thinking about other ways to utilize this technology which would allow us to not only continue to listen to and understand our consumers, but to create a more real-time two-way communication.”

Is your company utilizing text analytics or thinking about ways to do so? I’d love to hear from you.

- Tom H. C. Anderson