Analyzing Sentiment to Resolve Customer Issues

Customers interact online: posting reviews, commenting, sharing stories, and joining communities and forums. What they say matters to your reputation, and how you respond matters to their customer experience. But how do you track and manage—and act on—what goes on out there?

One of our clients—a Pacific Northwest software giant—had a lot of customer interaction and activity on global social media sites across multiple locales and languages. But they had no process for systematically capturing, tracking, assessing, and applying that wealth of user sentiment data to help resolve customer issues.

Here’s how we helped

  • Provided a way to track and analyze content, and form strategies
  • Created a proprietary tool to compile and evaluate user posts
  • Aggregated and translated posts to English using machine translation—including slang and informal language
  • Analyzed posts for opinion-related terms and phrases
  • Assigned a custom-developed sentiment score based on quantities of positive or negative words
  • Reported sentiment scores by language, market, platform, product, date range, etc., via a dashboard—enabling our client to see patterns and develop appropriate responses

The results?

  • Huge quantity of social data analyzed and aggregated
  • Cost-effective, systematic access to community sentiment
  • Rich BI that helped identify hot customer issues by market
  • Shortened response time from weeks to hours