Using Sentiment Analysis to Measure Social Media Engagement

As products become more commoditized and companies try to outperform their competitors, a better customer experience is becoming an increasingly important way to differentiate. Customer engagement plays a large role in the overall customer experience. There has yet to be a standard definition of customer engagement, but Ron Shevlin, an analyst at Aite Group, LLC defines engagement as: "Repeated interactions that strengthen the emotional, psychological or physical investment a customer has in a brand." Essentially customer engagement is a measure of relationship strength.

Most organizations’ already accept the fact that active customer engagement correlates with growing income and profit because of the way it makes customers more loyal, generates positive word-of-mouth, and reduces the probability that a customer will switch suppliers. It is now commonly believed an organization that can positively affect engagement is more likely to see these engaged prospects consider the company’s products making prospects, convert to customers, and purchase more regularly. Organizations developing customer engagement strategies focus on creating a positive and consistent online and offline customer experience.

As engagement becomes a key behavioral indicator, organizations will need to be able to link points of customer engagement with bottom-line improvements. Therefore it is becoming increasingly important to monitor and measure customer engagement.

There is yet no universal formula for measuring customer experience. One approach is to take the quantity, quality, breadth and depth of interactions between the company and the customer into consideration when measuring customer engagement and then measuring the correlation between the pattern of engagement and known stages of the purchasing process.

The advent of social media has raised the question of whether we can apply the same approach to measuring engagement to online interactions. As marketers look for way to measure of engagement of social media, research suggests many marketers are using time spent on a site as the most important performance metric, followed by unique page views. Since most social sites are not transactional, increased interaction does not necessarily indicate a more engaged or more loyal relationship. While the number of postings, completed profiles, participation in polls or forums indicate interaction they may not indicate loyalty.

The question then is, “Do the behaviors such as purchase frequency, loyalty program participation, etc. used to determine engagement in the traditional brick-and-mortar, email and e-commerce channels apply when trying to measure engagement of customer who are participating in blogs, social networks, private communities, and forums? The answer is probably not. So what should we use? Some of the latest thinking is to measure sentiment as an indication of engagement.

Sentiment analysis similar to text mining involves parsing and analyzing the comments and suggestions customers post on social media. It concentrates on looking at the context, tone, emotion, polarity and objectivity of the comments rather than solely the words. Twitter and Facebook have built-in applications that perform keyword-based sentiment analysis and Google is experimenting with its Google Trends service.

Sentiment analysis is still in its infancy but proponents believe it will enable organizations to recognize which social media participants are loyal and engaged and which are not. It’s only a matter of time before we’ll have better ways to measure the engagement of the social media customer. In the meantime, this is one approach gaining traction.