How to harness data to understand customer sentiment

Customers are constantly talking about the products and services that they use through different social media channels. They are not shy in letting friends know through Facebook if service was slow, or tweeting about a new feature found on a website. While organizations have access to all of these reviews, blogs, and tweets, they often do not have the ability to discern patterns and trends to make insightful decisions. Not only is all this disparate content hard to analyze, but the immense volume of data can be overwhelming, making it impossible to decipher what is relevant and what the motivation behind each post actually is. As such, organizations are missing out on critical data around exactly what customers think of their brand, product or service.

Now, organizations have the ability to understand the sentiment behind this unstructured social data. Using the Watson Sentiment Analysis service, part of the AlchemyLanguage API, companies can understand what constitutes a negative versus a positive tweet, review, or post, and can look for words that carry a specific connotation. Using this knowledge, organizations can then figure out which person, place or thing that the message is discussing and what the overall sentiment is. The aggregate of all individual posts provides insight into trends and signals that can be used to enhance business decisions. Watson provides the ability for developers and enterprises alike to take scattered, unorganized data, and derive value and meaning.

Hands Holding Colorful Speech Bubbles Social Media Concept

So, you can understand the sentiment around certain products, services, companies, etc., but then what? How can this data be effectively leveraged? Here are a few ideas:

• Understand customers’ likes and dislikes by analyzing what prompts positive and negative feedback, and use that insight to shape new services

• Optimize product features based on consumer sentiment

• Find out who your brand’s most vocal detractors are and reach out to them to improve relationships and change perception

• Measure effectiveness of public relations campaigns by gauging percentage of positive response from key influencers like journalists and analysts

With Watson, unstructured information that was once nearly impossible to parse manually turns to knowledge, and knowledge becomes power.

To learn more about Watson Sentiment Analysis and how it works, watch this brief deep dive video provided by one of Watson’s engineers: