This article can also be found in the Premium Editorial Download "BI Trends + Strategies: BI projects play key role in improving business and IT alignment."
Download it now to read this article plus other related content.
Social media has become one of the largest data banks available to companies looking to learn more about their customers. It's also developed into a key customer touch point, both for spreading marketing messages and gathering feedback. With the emergence of such a powerful resource, many organizations have begun to take social media data seriously and are working to analyze it to better understand customer preferences and improve their ability to detect potential business problems.
Such efforts typically involve the use of social media analytics tools, which are designed to recognize designated keywords in posts and comments on social networking sites, such as Facebook, Twitter, LinkedIn and Blogger. The software then can aid users in assessing what people on those sites think about new products, customer service experiences, ads and corporate developments.
But analyzing social media data remains an inexact science, according to industry analysts and consultants. They said that although social media analytics software can provide valuable information to companies, the results of queries can be skewed because the technology has trouble picking up on tone, slang and nuances such as sarcasm when trying to interpret text data.
More info on social media analytics tools and programs
Get advice on how to decide whether your company is ready for a social media analytics initiative
Read our special report on social media analytics technology, featuring news analysis and expert tips
See how retailer GNC is using social media monitoring software to track customer feedback
Katie Paine, CEO of consultancy KDPaine & Partners in Berlin, N.H., said social media analytics findings often are inaccurate unless companies use detailed search strings and build comprehensive dictionaries into the software they're using to help it compensate for the complexity of online conversation. If an organization doesn't plan to analyze large quantities of data, Paine added, bringing in an analytics software vendor or services provider to do the work on a contract basis can speed up the analysis process and be more cost effective than buying and fine-tuning a social media analytics tool.
At this point, Paine and other analysts said, the most common use for social media analytics tools is crisis aversion. The technology can serve as an early warning system for negative customer feedback, such as complaints about problems with products or customer service. It can also help detect other issues that could be detrimental to a company's image. For example, in 2009 two employees of the Domino's pizza chain filmed themselves violating health-code rules and uploaded the video of their antics to YouTube, where it attracted considerable attention. Done effectively, social media analytics enables organizations to defuse potentially damaging problems before they go viral.
Zach Hofer-Shall, an analyst at Forrester Research Inc. in Cambridge, Mass., said social media analytics also enables companies to get unfiltered information from customers about potential product enhancements. That can be a more effective way of gathering opinions on desired new functionality than surveying customers, he said, noting that asking people about specific improvements limits the possible answers. Such barriers aren't erected in social networks. And there's a vast universe of customers to tap into: According to an updated report on "social intelligence" that Forrester released in May, 86% of Internet-using adults in the U.S. now use social networks.
Know what you want from social media analytics
One of the tricks to successfully implementing a social media analytics program is to know up front what you hope to achieve from it, said Seth Grimes, founder of Alta Plana Corp., a consultancy in Takoma Park, Md. Before making any leaps and purchasing social media analytics software and associated computer systems -- which can have a total price tag in the seven digits -- it's important to make sure that the expenditures will accommodate your aims, Grimes said.
Executive-level involvement is also vital to efforts to incorporate social media analytics into internal business processes, he added. And project teams need to have an understanding of social networking and how analyzing social media data can benefit their organizations, something that best comes from direct experience, according to Grimes. "The key," he said, "is to get your hands dirty."
Susan Etlinger, an analyst at Altimeter Group in San Mateo, Calif., said the largest hurdle in implementing a social media analytics program is getting comfortable with, and proficient at, sentiment analysis. That requires taking a proactive approach in training the employees who will be involved in the program on how to use social media analytics tools to interpret data, Etlinger said. Organizations that do so, she added, position themselves to gain real business value from the process.
For now, social media analytics software isn't advanced enough to be relied on to predict market trends, Paine cautioned -- although that hasn't stopped some companies from trying. "It's scary that people are already using it to make predictions and projections without perfecting its accuracy," she said. Paine thinks it could be another three to seven years before the technology is able to fully understand the complexity of human language and generate reliable forecasts of emerging trends.
Nicholas Evelson is an English major at the University of Massachusetts, Amherst.
This was first published in August 2012