Bridging Structured & Unstructured Data in Federal Agency’s Global Survey Results

A large government agency with international operations conducts an annual customer satisfaction survey in multiple languages to gather feedback on services provided at different locations worldwide. The survey’s mix of closed- and open-ended questions cover a wide range of service offerings across numerous branches of the organization. The agency wanted to create a report capturing the “voice of the customer” to better understand the drivers behind satisfaction ratings for various customer segments in order to tailor services, operations, and programming to these distinct groups.

By utilizing advanced text analytics software, ORI was able to deliver a report capable of driving decisions, marking a turning point for open-ended feedback analysis for the agency.

Challenge

Previous reports focused on closed-ended data fields were well received within the agency, but reporting on open-ended customer feedback failed to provide sufficient detail to guide decision making. Analysis of these text-based responses lacked the nuance needed to create actionable insights or allow for easy comparison across different customer segments. Additionally, insights were only loosely connected with closed-ended data, presenting a missed opportunity for detailed segmentation to highlight specific problem areas.

Solution

Leveraging our advanced text analytics platform, ORI quickly and efficiently analyzed text from more than 50,000 survey responses to translate feedback into themes for improvement, transforming a once-defunct report into a high-value analysis yielding actionable insights. In a game-changing approach, ORI paired analysis of text-based unstructured data with an examination of other closed-ended structured data on the related service, location, and branch to detect patterns and identify improvements to increase satisfaction.

To analyze text at the clause level, ORI parsed text responses to open-ended questions and applied natural language processing (NLP). By combining clause-level sentiment and effort analysis with automated theme detection, we created detailed thematic categories and subcategories, isolated key areas of customer dissatisfaction, and identified the reasons behind that dissatisfaction.

Outcome

By utilizing advanced text analytics software, ORI was able to deliver a report capable of driving decisions, marking a turning point for open-ended feedback analysis for the agency. The analysis identified the top concerns of agency customers, narrowing in on specific services, processes, and programming that can be improved to yield far-reaching benefits throughout the agency.

As a value-add, the agency was presented with interactive dashboards that allow it to dig deeper into survey results to connect satisfaction ratings with various themes, sentiment, and effort and quickly compare results across different segments and agency branches. The dashboards and new themes derived from the analysis provide a richer understanding of what drives satisfaction ratings for services across different locations and segments.