Text mining gives structure to content and meaning to sentiment

Published on .

Reprints Reprints

Rebecca Wettemann is VP of Boston-based Nucleus Research. BtoB recently asked her about the marketing uses of text mining. BtoB: What are the major benefits of text mining to marketers? Wettemann: To rapidly digest all the unstructured text content that's on the Web—information about both your own company and your competitors—so you can promote or counter it. Another use is in educating oneself about prospective customers. BtoB: How would it do that? Wettemann: To better understand your prospects, what other partnerships they may have, how they're being treated in the press, etc. Knowing your prospective customers' challenges with their own customers or image can be a great way to hone a pitch. BtoB: What about its use in a marketing campaign? Wettemann: It could allow you to know and make sense of what's been said about your company over the past 30 days, for example. Or you could analyze all customer feedback about your Web site, say, before overhauling it. BtoB: Aren't old-fashioned surveys already used for that? Wettemann: Yes, but some of the most valuable responses to surveys are where respondents can say anything they want. Previously, you couldn't possibly tabulate open-ended answers manually. Text mining allows you to digest a lot of this qualitative information in a meaningful way, giving you a good idea of people's sentiments—which might in turn be the needs of a new marketplace or segment. We've seen organizations take transactional data in context—from surveys, yes, but also in their interaction with call centers, or online forums or blogs—and then figure out what kinds of offers to make. BtoB: It seems logical that you could turn text mining's analysis on your own partners and customers. Wettemann: One new attempt to do that is Salesforce Ideas, [which allows] a company's employees, channel partners and customers to submit what they think. But again, that can constitute a huge amount of unstructured sentiment. When you layer text mining on top of that, you now could have a significant opportunity to understand these people's needs, desires and demands. M
In this article:
Most Popular