3 text analytics use cases for improving customer experience

Text analytics is being used by a growing number of data-driven organizations to generate powerful customer insights that support customer-centric decision making.  

Text analysis aids businesses by providing an efficient and scalable solution for analyzing text data from survey open-ends, review data, social media conversations and customer service transcripts that help them understand the needs, expectations and experiences of customers and target audiences.

Text analysis tools are designed to make sense of loads of unstructured text that traditional business intelligence tools would struggle to analyze in an insightful way.

In this article we’ll explore three common text analytics use cases for customer experience teams.

Reveal the drivers of positive and negative customer experiences

One of the most common use cases for companies that use text mining involves comparing positive feedback from survey open-ends or reviews against negative feedback. This approach efficiently reveals what people like and dislike about your selection of products or services.

This use case requires text data to be associated with a score or a rating. An example would be an open-end associated with a net promoter score (NPS), customer satisfaction score (CSAT) or a star rating on an online review.

Taking this approach, you might find that many customers leaving five-star reviews are mentioning how great over-the-phone interactions were. This would suggest your call center agents are delivering good solutions for customers. Conversely, if you could then see that a large portion of the one and two-star reviews were complaining about the product they purchased not working as intended then this would indicate that quality control should be a focus area for improvement.

This kind of analysis can help you boost customer satisfaction by zeroing in on the aspects of your customer experience that customers care about most.

Understand differences across markets and customer segments

A second common example of text mining for customer insights revolves around segmenting and comparing survey responses or reviews based on attributes associated with respondents. Segmenting based on geography, for example, can help you understand how customer experience differs across regions or international markets. Other popular ways of segmenting your text data include age, buyer behaviour (e.g. repeat customers vs. one time purchasers), personas, order value or any other demographic datapoint.

Consider an example where we are comparing customer feedback from the US against the UK. This approach might tell us that British customers are more likely to experience delays in deliveries whereas Americans are more likely to receive damaged products. These insights would help the logistics teams in each market set their priorities in a way that is most likely to drive improvement in customer experience.

Track changes in customer feedback over time

Customer wants and needs are dynamic, being constantly influenced by a combination of economic conditions, trends and upstart competitors. For businesses, keeping up can be difficult and this limits their ability to put customers at the forefront of strategic decisions.

Text mining techniques can help businesses identify and manage changes in their customer experience. This is useful for both the proactive identification of emerging trends as well as quantifying the impact of strategic changes in customer experience strategy.

A good text analytics tool will allow you to define the topics you want to track and visualize how the prevalence of these topics is changing over time in your text data. This enables you to focus on the things that are having the biggest impact on customer experience.

Here’s an example of how Relative Insight visualizes changes in topical discussion over time…

Visualizing changes in public conversation about electric vehicles over time

Let’s look at a final example. Analyzing changes in customer survey responses over time, we spot a reduction in survey responses mentioning poor quality. This might suggest that quality control improvements have been successful. Alternatively, an increase in survey responses including the word ‘delayed’ or ‘delay’ might indicate a potential supply chain or logistics problem.

Interested in giving any of these text analytics use cases a try? Book a discovery call with one of our experts to explore what’s possible.