|
Analyzing sentiment can be tricky. Although a hot topic among various social CRM and social media monitoring companies, the art of sentiment analysis has yet to be perfected. In most cases, a fair amount of human intervention is required in order to set algorithms, tag content, and remove anomalies. However, the advantages gained from automating sentiment analysis are big - especially for large corporations trying to weed through the myriad mentions of a brand or products on the social Web. With this goal top of mind, social media monitoring company Radian6 has incorporated sentiment analysis technology into its platform. The tools are immediately available to users, as announced in a mid-December blogpost by Radian6. Ryan Strynatka, the senior product manager at Radian6, says that the analytical tools have been in trial for the past few months. Feedback has been positive, he says. "It's something people know they need and have been asking for," Strynatka says. Forrester Principal Analyst Suresh Vital says unveiling sentiment analysis was a logical next step for Radian6. "It's what they need to do; it's what marketers are looking for; and it's what competitors are doing," Vital says. However, the analyst has some caution involving today's maturity of sentiment analysis. "The reality is, automated sentiment is a bit of a misnomer," he says. The technology isn't at a place today when automated sentiment is always 100 percent accurate. In fact, Vital says in talking to clients who have deployed some form of sentiment analysis, accuracy rests at about 50 percent. There are ways in which users can increase the accuracy and results of their automated sentiment analysis. For instance, playing around with taxonomies and auto discovery can yield better results. "In the near term," Vital says. "Human intervention will still be necessary." Strynatka says that customers will realize that sentiment analysis is not an "on or off" tool. It's a flexible capability that will yield different results for different users. Using sentiment analysis for examining trends is an area he says he thinks will be of interest. Additionally the capability could prove valuable in customer support crises. When users see a spike in mentions, they can drill into the various Radian6 categories and get a feel for the conversation.
"It's a great move to make [sentiment analysis] available to the masses," Vittal says. Only time will tell whether users will be able to make the most of the solution. "Fundamentally, the reality of the space is, you have a market where automated sentiment is a good first step but you need to go beyond the step in focusing on the accuracy," Vittal says. News relevant to the customer relationship management industry is posted several times a day on destinationCRM.com, in addition to the news section Insight that appears every month in the pages of CRM magazine.
|