Although “sentiment” has been tracked since the 1960’s to measure consumer confidence, automated sentiment analysis is only recently emerging as a compelling business tool for managing customer experience efforts. Sentiment analysis allows you to track both facts and opinions –such as customer transaction history as well as their comments on Twitter and Facebook– to more fully understand perceptions about your brand, products and services, and identify problems with customer experience so you can act quickly and effectively. (See our Bank of America case study.)
How to benefit from sentiment analysis
At Beyond the Arc, our data team uses text mining tools to monitor negative and positive comments and tracks acceptable ranges for each over a period of time. If the percentage of total comments falls outside of set ranges, we look for themes and categorizations by analyzing comments captured over this timeframe. Using this method, you can take advantage of sentiment analysis to:
- Gain insights from social media conversations
- Build a sentiment index to measure company and product reputation
- Develop a strategy for responding to negative sentiment to positively impact perception
- Identify and engage top influencers to reinforce desired messages
Overcoming the challenges of sentiment analysis
While sentiment analysis offers valuable insights, it’s not an exact science:
- Sentiment tools are not precise in discerning sarcasm and wit.
- Due to a lack of standards, different tools can deliver different sentiment scores.
- Sentiment analysis results are not completely accurate. In fact, Seth Grimes, one of the leaders in the space, argues that sentiment is an “80% solution.”
Automated sentiment analysis can yield significant insights: from understanding challenging customer experiences, to identifying competitive threats, and even tracking emerging market opportunities. Although sentiment analysis can’t give you a perfect read, skilled data analysts can work the angles to get the most out of what this technology can do.