Social Enablement: The Importance of Sentiment Analysis

This is the final post on a subject that isn’t a primary focus in my role at Adobe but is integral to our search marketing activities: social media. As I’ve mentioned before, SEO is changing into more of a customer optimization strategy. Best practices (which sprint to keep up with constantly changing algorithms) demand that we integrate social channels into paid, organic, and site strategies at a deeper level.

Social marketing, like search, cannot be effective without measuring the type of activity, responses, and feedback through Facebook, YouTube, Twitter, and other platforms. In this case, we’re talking about sentiment analysis, the contextual and noncontextual cues embedded in social engagements. Sentiment analysis should drive campaign decisions ranging from which channels to deploy to what reactions we want to entice.

Fewer than 50 percent of digital marketers understand whether social is working for them. How can that be? With 86 percent of marketers indicating they value social highly and 88 percent wanting to know the most effective social tactics to deploy, it befuddles me that far fewer value the importance of knowing how their social campaigns are working. Often, the reason is a lack of focus on understanding consumer sentiment.

Adobe’s VP of Social and Analytics Bill Ingram commented last year that “social marketers have largely had to rely on instinct to uncover not only what resonates but what will maximize future engagement on social platforms.” Instinct is a risky, immeasurable basis for committing thousands of dollars to a social campaign. The reason marketers continue to struggle with social KPIs can no longer be linked to a lack of actionable data gathered and displayed through an easy-to-understand interface. The Adobe Social platform, for example, enables sentiment analysis with predictive analytics: the who, what, where, and when of audience measurement. Our product not only provides suggested timing for the highest potential engagement levels but also allows users to predict how many comments, likes, and shares any given post is likely to receive.

Most social sites provide general audience metrics, but success is more dependent on actionable data that are revealed through audience participation, so it’s important to comprehend positive, negative, and even neutral opinion trends. Third-party APIs are often more robust when it comes to sentiment analysis. KPIs such as conversation rate, for example, are a good place to start. The conversation rate measures the average comments per post, which tells you how engaging your posts are. Like the long-dead page view metric, with social it’s not enough to get them to see your asset—they’ve got to respond.

Beware: some social metrics reveal less about customer sentiment than others (for example, Followers-to-Following) and they matter little when it comes to analyzing ROI. Social influence is one thing, but as with SEO and search marketing, the bottom line is how well your activities are driving sales. You’ve got to be looking at attributional metrics that reflect sentiment.

Here are a few of the more effective breakout social KPIs to analyze for uncovering consumer sentiment:

To enable social success, which impacts search success, be clear about the metrics you’ll focus on for sentiment analysis and the tools you deploy to capture, organize, and report those metrics. Customer optimization relies on a solid understanding about your social network’s composition and, more importantly, its activity.