How Classifying Data Drives Value in Your Voice of the Customer Program

Co-author: Gavin James

In simplest terms, a Voice of the Customer (VOC) program captures your customers’ experiences and feeds them back into the organization to drive improvements that help grow your business. Whether customers feel good or bad about their experience with your products and services, your VOC effort enables you to hear what’s being said so you can take action to reduce risk, leverage opportunities, and build stronger relationships with your customers.

To focus your efforts across millions of customer comments, classifying and analyzing the data helps you accurately measure customer issues and track experiences over time so you can take appropriate action.

Leveraging customer feedback across multiple channels

Although you may rarely record feedback from direct customer contact like the sales floor or service desk, it’s likely your business has access to a wealth of feedback data through electronic channels such as email, call center recordings, surveys, and increasingly through comments on Facebook and Twitter. By capturing this data for analysis and tracking, you can measure the current state of the customer experience, and track your progress as you make improvements over time.

Driving improvements with targeted actions

To effectively take action to improve your customers’ experience, you need an accurate read on where and when to focus your efforts. Tailoring your VOC data with context specific to your business can help you get there faster. At Beyond the Arc, we use sophisticated analytics software to help businesses track issues, and uncover emerging issues to stay on top of the customer experience as it changes throughout the customer lifecycle.

Here’s a look “under the hood” at how we create classification engines so you can act on Voice of the Customer data:

  • Capturing main ideas – Our analytics tools use Natural Language Processing (NLP) to automatically capture the main ideas from customer comments. The NLP is augmented by business knowledge and terminology specific to each line of business in the company.
  • Classifying key issues - To identify key issues, a classification engine sifts through millions of documents to track the main ideas. Again, business knowledge is integrated to craft actionable categories. Oftentimes, this part of the work is an extension of a reporting system already in place, but one geared towards hundreds of documents rather than millions.
  • Aligning data with measurable outcomes - Our classification engine uses a combination of statistical and linguistic techniques that include NLP, C5, and Apriori algorithms to discern the most meaningful way to classify each customer comment.
  • Engaging the human touch – Customer feedback is first categorized by business experts to ensure accurate meanings are assigned to comments about various experiences. We confirm accuracy over time as the business and the customers are always evolving. Whenever we update the classification engine, we can test the effects by comparing old and new findings to ensure the most accurate outcomes.

With customer data categorized to the specifics of your business, you increase the relevance of customer feedback to more effectively focus your actions across lines of business. You can also measure the impact of your improvements:  Are the big problems changing? Are customers celebrating your brand? VOC analytics can put the answers, and key issues, in your hands so you can address them at the right level at the right time.

Like this post? Download and take it with you!