Text Analytics: Interpreting the Consumer’s Story

Text content is available all over the internet. All sources of raw text data are emails, forum questions, social media messages, customer support requests, responses to open-ended survey questions and even this blog post. Text data are drawn from something written by a human being and it is possible to examine these raw text data to discover informative patterns.

What is Text Analytics?

Text analytics, also known as text mining or text data mining, is a valuable method to collect continuously processed, high-quality data from raw text. A code script analyses the raw text data for patterns or classifications automatically in text analysis. One easy example is the extraction of such terms from specific frequencies. You may be interested, for example, in knowing the frequency at which people discuss your brand in comparison to other brands.

Another example, and more important, is the frequencies at which words with certain distances from other words occur. Of example, you may want to know which words occur the most frequently near the term 'customer service.' Will customers write terms like 'perfect,' 'warm,' and 'friendly' close customer service mentions or will they say anything else? Words that often occur together are called 'collocations,' and provide a potentially rich source of knowledge about text.

Another method is an examination of emotions, which marks comments as either positive or negative. This information can be useful in deciding how a customer feels about a product or service, or in deciding whether certain text-based features are associated with a positive or negative feeling.

Impact of Text Analytics on Market Research and Insight Generation

Market researchers now regularly collect raw text data in the form of open-ended response questions. When a client wants to share this information, they may also have access to text data 'in the wild' (such as social media posts) or to customer text data (such as forum questions or customer service emails). These raw text data can be analysed by a computer script for interesting patterns. One basic technique is to use frequencies to create the familiar infographic about the word cloud. Yet there are other solutions available beyond a basic word cloud, with more sophisticated techniques.

Businesses should understand general sentiment, feeling correlations, regular collocations, writing style, and more, and use this information to address questions about the values and drivers of customers behind those values.

In the past, human scholars have often read open ends, and human reading does have its advantages. We can make correct and intuitive classifications while humans read slowly. That remains a valuable skill even with new technologies and research methods. However, machines can read and make quick analyses in large data sets, allowing the human researcher to process large amounts of data in a short time. Often, computers apply similar analytical methods at each data level, resulting in good precision. Human programmers and researchers will then be able to intelligently apply their experience and use text analytics to draw informed conclusions. This makes the technology perfect for use in market research.

Using text analytics as a combination of human intelligence and experience coupled with machine speed and precision can open doors to new insights and faster discoveries for businesses interested in understanding how consumers really feel when typing out a message.