Using Text Mining to Explain Consumer Behavior Regarding Brands

Advertising Age is a site with a good content on Marketing and Media. Walking through the content of the site, I found an interesting article about the use of Text Mining to Explain Consumer Behavior regarding brands, entitled Text Mining Provides Marketers With the 'Why' Behind Demand, written by Michael Bush.

He tells the case of Stephanie Hoppe, 7-Eleven's senior director-marketing,that knew there was an incredibly competitive market for iced coffee. But before entering the battle by expanding its own product across its network of stores this past June, Ms. Hoppe and 7-Eleven needed to figure out the why that was driving consumer demand. He defines Text mining as a process in which an agency can mine the conversations taking place on blogs, Twitter or other social-media sites in order to identify the emotional aspects and reasons behind consumer behavior. Typical database information supplies marketers with the who, what, where, when and even the how, but not the why.

Ms. Hoppe Used a tool called Digital Anthropology from Omnicom Group agency Rapp, and got a deep-dive analysis of what consumers were saying about why they drink iced coffee, what flavors they prefer, what presence its competitors had in the social space and what people were saying about their offerings.

"Text mining can't replace traditional data, but the combination of the quantitative and qualitative data they both offer creates a more holistic view of the consumer. In most cases there is no name connected to the data compiled from text mining or text analytics but that may soon change", said Megan Bannon, cultural anthropologist at Rapp.

According Sid Banerjee, text analytics is starting to replace the focus group because it's cheaper and faster. Clarabridge collects data from three sources: survey responses, customer/representative interactions and social-media sites. He is CEO of the Clarabridge, a customer experience management firm. He said these data are significant in a recession because it helps you make sure you are cutting the right marketing dollars. "Being able to micromarket when you're marketing is important but being able to microcut is also important," he said. "You want to take a scalpel to your business, not an ax."

This article shows how the use of Text Mining is an interesting approach to know the consumer behavior. Using Text Mining, a company can collect data from many sources, knowing about what the consumers are saying about their products.

Voice of the Customer: Text Analytics for the Responsive Enterprise

Seth Grimes wrote a very nice report entitled Voice of the Customer: Text Analytics for the Responsive Enterprise, published recently in B-Eye Research.

He defines the report subject: "Text analytics helps business users discern and capture the Voice of the Customer from online media such as blogs, forum postings and news articles; from email, chat interactions and contact-center dialogues; and from surveys and other mechanisms for collecting customer feedback: from the totality of enterprise information sources.
This report describes how text analytics has become an essential part of Voice of the Customer solutions. It describes VoC techniques and processes and explains the fundamentals of text analytics technology and solutions. It covers implementation options and best practices, addressing the basic question: How do organizations get started with VoC text analytics? And it presents the findings of a best practices survey that asked end users and consultants about their goals, information source, return on investment (ROI), and advice for organizations looking into VoC text analytics."

The report is well divided; it has an introduction, three parts and an appendix. In the introduction, he did a question: What are your organization’s customers – and your prospects and the media – saying about your company and your products and your competition? And gave the answer: "Voice of the Customer (VoC) initiatives can answer that question and help you formulate your response. VoC is not just an extension of customer relationship management (CRM) to the customer experience. It is an approach that can guide enterprises in meeting the spectrum of sales, marketing, customer support, brand and reputation management, product and service design, and quality demands."

In the first part, called The Voice of the Customer, he defines the concepts, needs and challenges, sources and voices, and also the VoC elements and VoC process; finish talking about some users experience.

He talks about the fundamental business problems in the voice of the customer, and generally the organizations ask:
- How can we improve customer satisfaction?
- How can we maintain and increase our competitive edge?
- How can we boost profitability?

He also talks that the need is to hear, understand and act on customer, and the biggest VoC challenges are to identify and access all relevant sources and make sense of their contents. He classifies voice sources in operational systems, solicited feedback and "open source" intelligence. The open source intelligence is the information available to anyone: forum postings, news and review articles, blogs and other social media.

In the VoC elements, he talks about the concept of the net promoter score and customer lifetime value. In the VoC processes, he defines the setting goals, how information gathering and filtering, and explains the facts and opinions concepts: what distinguishes opinions from facts it that opinions convey sentiment, they convey attitudinal information. He also explains what is data enrichment and analysis. He finishes the first part describing user experiences.

In the second part, entitled Technical Insights, he defines the technical context and explains the text technology basics. The text analytics is an answer to the unstructured data challenge. With the availability of text analysis tools, large-scale, automated analysis of textual sources is finally attainable, and this shift has occurred for BI, for predictive analytics and for VoC efforts. He explains that sentiment analysis is one of the most exciting applications of text analytics nowadays, because sentiment is central to human communications, it is also of key importance for VoC study.

In the last part, called Study: VoC text Analytics Best practices, he talks about best practices for the implementation of VoC text analytics initiatives.

There is an appendix with links to some case studies and vendor solutions, provided by IBM, SPSS and Business Objects.

This is a very nice report, written by Seth Grimes, an expert in text analytics and should help everyone that is interested in use the concepts of text analytics in the companies.

The use of text analytics to increase guest satisfaction in big hotels

Last week, Intelligent Enterprise published a good article about the use of text analytics in Gaylord Hotels, this group has big hotels with more than one thousand rooms each.

They are using the concept of text mining to analyze quickly the customer satisfaction surveys to make faster decisions to solve the issues.

It is very interesting how the use of a new technology allow a new approach for an old issue.