Monday, December 8, 2014

How to Gain Customer Empathy via Text Analytics

A Voice of Customer -VOC solution needs to be integrated into a wider customer experience strategy.

It is the age of social collaboration, organizations today are no longer limited by the physical walls,  the functional borders, or the inside out views; customers and partners are more involving through social engagement and collaboration in the business ecosystem. And being customer centric is the strategic priority for many forward thinking businesses in the digital era, hence, analytics needs to become the mainstream culture theme. Technically, how to gain customer empathy through customer sentiment analytics? More specifically, how would you analyze and structure a lot of free text comments from customers?

A Voice of Customer -VOC solution needs to be integrated into a wider customer experience strategy. At the same time, it's a shame not to take care of all valuable feedback that is available at the fingertips. There are a number of software solutions out there that are specifically devoted to customer sentiment categorization and keyword analysis and so on, and  it makes the most sense to have text analytics integrated entirely within a wider customer experience management solution. Unfortunately, most companies have to do the hard work and read the free text comments, going through every single one. You can always use some kind of filtering such as adding rating questions in the surveys along the free text, so that you only read, analyze and categorize the ones that matter most, such as Detractors, or customers that for some reason value high one aspect and low another and need an in-depth analysis.

Discover the root cause of customer’s feedback. Another method you could use is to ask the customer his/her main reason to give you the feedback. Here, you actually ask the customer for the root cause of his/her feedback.This does have the advantage of objectivity (of interpretation) of employees structuring the feedback plus the fact that you are more efficient in the analysis of feedback. And you know where to look for improvement possibilities. For example a customer said he/she visited the store and had to wait for 30 minutes to be served. The employee was friendly, but not very pro-active... Root cause possibilities within the survey: waiting times, quality of service, friendliness. If the customer answers "quality of service" you have your first root cause. You may want to use a second root cause: Proactivity, Know-How, ... In this case, a customer would say proactivity.

The right tools or the combination of approaches can make customer sentimental analytics more effective. There're a lot of different software and enterprise solutions out there for this purpose and analyzing customer feedback is more than just comments from a survey. You can use a combination of manual read through and text mining to categorize the comments (single words, clauses, phrases, etc.) vs. the targeted attributes. You can then use different analytics tools to analyze the key comments, how they are clustered, and how they are linked to the customer experience. It takes in your survey data and presents charts and graphs that show sentiment across various categories.

Ownership of customer surveys within an organization should be handled centrally; to ensure a strategy and good quality, together with the development needed based on customer feedback. Many companies struggle with finding a structure of how to successfully work with customer experienced-based decision making. It will be a challenge to find an enterprise solution without a common strategy and many different ongoing ad-hoc solutions. Ownership of customer surveys should be centralized or coordinated centrally, otherwise there is danger of the customer getting multiple surveys from different departments ending up in survey fatigue. similarly, the output and actions coming out of the surveys should be coordinated centrally, otherwise, you end-up with functions working at cross purposes in the interest of improving the functional level scores. So the ownership of the customer survey control definitely belongs to a central spot, most importantly for customer contact control, but also to enforce consistency and provide expertise which will drive greater adoption of the initiative across the enterprise. Once the study is defined, ownership of the effectiveness (Closed Loop resolutions, major client action plans) live with the owning division or sponsor. The central team monitors the committed action plans by division and suggests coordination when like requirements exist across divisions or geographies.

It takes analytics thinking to do customer sentimental analytics step by step: 

1) Generate high-frequency phrases in text response. This step will provide you trends and high-frequency text in your customer surveys. It will also provide you trends and some insight into customer experience.
2)Define dictionary based on the high-frequency phrases. You need to create lexicon / dictionary of words to summarize the information.
3) Create Customer Journey Chart and associate this dictionary with customer journey to identify the area of pain points and take corrective action.
4) The above steps will give you information, but to build actionable insights, you need to do some manual efforts in analyzing the key comments and point out the pain areas and actions required. 
5) Availability of open source software has helped in generating this information at very minimal investments. You can create all these steps using the software and with low manual efforts, you can bring in insights about the surveys.

Just like any type of analytic, the key point here is not technology but to generate the insights, which can help you in improving the customer experience and overall provide an edge over your competitors. Doing Sentimental Analytics is the means to explore what’s in customers’ mind, gain customer empathy, and take action for building a more customer-centric business.


This is being done by

This comment has been removed by the author.
This comment has been removed by the author.

Customer sentiment analysis is the process of using natural language processing (NLP) algorithms to analyze customer feedback and comments, and measure the overall sentiment toward a company or its products and services. This can be done through surveys, social media posts, and other channels where customers provide feedback.

Post a Comment