Customer experience text mining, like its geological counterpart, is sifting through vast amounts of debris to find the gold. As noted by Ronen Feldman and James Sanger, authors of The Text-Mining Handbook, what we call “unstructured data” isn’t completely unstructured: text follows some basic tenets of natural language. Text mining means analyzing text to 1) determine what the original author was trying to say or 2) learn something completely new.
Like data mining, the idea is simple, but what’s “under the hood” in text-mining applications can be very complex. One common technique is called “categorization”, which has been used since the late 1960s in areas such as medicine and news services. A simple example would be deciding whether customer emails represented “happy”, “unhappy”, or “neutral” customers, based on the types of words used in those emails.
In recent years, text-mining algorithms have been more widely adopted by businesses, thanks to Moore’s Law, which drives continued computer performance improvements. But equally importantly, the setup and usage of text-mining systems have become much easier with the adoption of packaged and on-demand solutions.
At JetBlue, customer experience text mining was introduced as a result of the infamous New York ice storm of 2007. After being overwhelmed with fifteen thousand emails in just two days, a text-mining vendor helped the airline learn that customers were upset about the delays and cancellations and disappointed that JetBlue didn’t have a backup plan. On a more positive note, some customers wrote to compliment airport staff and in-flight crews on their handling of a difficult situation.
Since that crisis, JetBlue has worked to more systematically mine customer sentiment, as well as providing “tangible data around how to augment JetBlue services,” according to Bryan Jeppsen, the airline’s customer feedback analyst. By tying feedback data to a specific aircraft or even a seat number, they can find and fix problems that have a direct impact on the customer experience.
More generally, contact center executives are looking for more effective tools to address customer experience and operational performance issues.
When creating customer feedback, “We took a pretty decent stab at anticipating what our customers might want to tell us,” said Meredith Sime, associate director of customer experience with AT&T’s U-verse. “But we did not initially invest enough resources in mining the wealth of information from an unstructured format.” Those instincts were borne out once they started analyzing the data. “We learned a tremendous amount of information that helped us drive improvement plans, and it taught us to ask better questions.”
Text analytics is not just about automating a process to gain efficiency. Progressive business leaders see text analytics for customer feedback as a necessary part of a customer-centric enterprise. What golden nuggets of information are hidden within the customer feedback your organization collects?
About the author: Bob Thompson is an international authority on customer-centric business management who has researched and shaped leading industry trends since 1998. He is founder and CEO of CustomerThink Corporation, an independent research and publishing firm, and founder and editor-in-chief of CustomerThink.com, the world’s largest online community dedicated to helping business leaders develop and implement customer-centric business strategies. His book Hooked on Customers (April 2014) reveals the five habits of leading customer-centric firms. For more information visit https://hookedoncustomers.com
Images purchased under license from Shutterstock.
Related articles:
