Based on the numerous interactions that a service organization has with its customers, a large amount of data is captured and stored in company systems and databases. This data can be categorized as either Structured or Unstructured. In order to understand the difference between the two, lets imagine an iceberg floating in the freezing Arctic ocean. The ice you can see above the surface is generally much smaller in size than the ice which is submerged under the water. Structured data is the portion of the iceberg that you can see above the surface, approximately 20%. That data is easy to understand and interpret. Most companies use this data in their standard management reports because it is easy to comprehend and usually resides in an organized repository such as the CRM or Service Management System.

While this structured data is a critical piece of the information required to run your service business, there is other data in your company that is also equally important but usually underutilized. This information is the data that is difficult to report on, understand and interpret. This is the portion of the iceberg that is under the surface which represents about 80% of all your company data. This unstructured data includes customer feedback, knowledge base search strings, case notes, and service resolution details.

Structured data can help you quickly determine critical business measurements to understand WHAT is going on in your business. However, structured data rarely contributes to explain WHY you are experiencing the results. Historically, the only way to discover the WHY in both these scenarios was through a manual process of reading survey comments and case notes, which takes a significant amount of time and often produces inaccurate results. Unstructured data analytics provides a solution that can help you more quickly get to the “WHY.” Companies adopting unstructured data analytics can gain a competitive advantage by understanding why their organization’s performance is changing and more quickly adjust to address the issues.

There are different ways to leverage unstructured text mining and analytic techniques to evaluate and understand unstructured data in your customer service environment. Here are the most common techniques:

• Term Frequency – word frequency and relative importance to the set.
• Similarity Matching – matching based on similarity algorithms
• Topic Modeling – dominant theme clusters
• Sentiment Analysis – analyze the opinion or tone
• Named Entity Recognition – proper noun determination
• Event Extraction – relationship between words to determine events
• There are a number of Open Source tools available on the internet that use these techniques.

If you consider all the unstructured data in your organization, some of the most critical yet undiscovered information is the unstructured data about your customers. This data can be found in several locations:

• Customer comments from surveys
• Customer feedback on social media
• Incidents/Cases/Service request report to customer support
• Customer community threads
• Knowledge base content

After performing an Unstructured Data analysis, we can gain a better understanding of WHY customer call for service.  The customer comment data will provide a deeper insight on both what the customers are saying about your company, products and services and WHAT the customer is doing with your product and services.

Traditionally, service organizations have only used structured data to manage and measure their service performance. Leveraging the unstructured data and applying text mining and the open source analytic tools and techniques provides greater depth and insight into the overall customer experience. This also allows companies to redesign more proactive service, meet customer expectations and be more competitive in the marketplace.