It's easy to get caught up in social media, but there are reasons to question the veracity of its data and ponder whether it's delivering valid, useful results.
Among the more challenging aspects of serving as a CIO is sorting through the constant barrage of buzz and hype and arriving at a business and IT strategy that actually makes sense. The social media sphere is a perfect example. Recently, Gartner research director Martin Kihn noted that too many social measurement strategies are downright antisocial and are leading to less-than-accurate results.
Never mind that marketing execs and others are all atwitter over them. "Social marketing measurement … means information pulled from common social media listening tools and not firewalled data from your owned channels, e.g., your brand's own Facebook page," Kihn notes. "Although some executives are aware of the inherent biases and flaws, a majority of digital marketing analytics consumers may not be," he wrote in a recent blog post.
Indeed, it's easy to get caught up in the vendor- and media-generated excitement. But, as he points out, there are valid reasons to question the veracity of social media data and ponder whether it's actually delivering valid and useful results.
Kihn argues that the problems revolve around five primary areas:
1. Virtually all public discussions about companies and brands take place in private, and the comments aren't visible to businesses.
2. Twitter is very overrepresented in data: Kihn pointed out that it comprises about 84 percent of the data stream. It "has a vastly overinflated sense of its own place in the world," he noted.
3. Twitter tends to "represent the high and low end of the conversation spectrum," and, as a result, it should not be viewed as an accurate representation or a focus group.
4. Social objects are not created equal. A Pinterest pin or Facebook post may take more thinking and effort to generate than a short tweet. Dashboards don't necessarily take this fact into account.
5. System weightings are skewed. "Social measurement should actually try to measure something," Kihn wrote. "It should strive to be a representation, an analogy, of actual human behavior and conversation related to your project." He suggests weighting tweets and other posts by length and other factors.
Over time, it's becoming apparent that putting social media posts in the right context and extracting diamonds from all the lumps of coal isn't easy. However, as Kihn points out, an organization's ability to understand, analyze and weight posts is "an improvement over blindly plugging in social listening data and conducting business as usual."