According to the Pew Research Center, 65% of American adults use a form of social media platform, which is a sizable jump compared to the 7% of adults who were using social media in 2005. The growth of social media users gives businesses a chance to employ analytics and figure out who regularly engages on their social media page. Developing technology is helping organizations find out what their customers are saying about their business, and devise strategies to better influence those customers. This is called social listening.
Social media serves many purposes. For one, it is a fast, convenient source of information; now, Americans get the majority of their news on Facebook and Twitter. In addition, We Are Social Media reported that one of the top reasons people use social media is to share opinions. These opinions cover topics from politics and controversial topics, to the quality of a burger. These opinions do not go unnoticed, and if they “go viral” they can easily attract the attention of millions.
Some of these opinions on social media platforms are about particular businesses. Often, users will tweet at a business to complain about poor customer service, and sometimes that business might reply to offer an apology and a discount. However, social listening technology goes one step further and listens to posts that do not tag businesses. The aim of social listening technology is to find posts that are about a specific business, look for trends, gain insights from customer sentiment, and create strategies from those conversations. According to Forbes, it can be integral for the customer journey cycle.
Social Listening and the Arts
How can social listening help the arts? Consider this scenario: a symphony presents a concert with a world-renowned soloist playing a Rachmaninoff piano concerto. An avid classical musician attends this performance, but leaves unsatisfied. Later, they take to Facebook in order to complain about their experience. The user does not tag the organization, but their friends see it, which influences these prospective customers to not attend an upcoming concert. The symphony, who probably never saw that social media post, suffers a loss.
However, if a symphony’s marketing department uses a social listening tool, they can scan social media posts around the web and realize that the overall customer sentiment regarding the performance was negative. With this insight, they can identify a reason why sales were down that particular weekend and, perhaps, advise the Artistic department on how to better program their concerts for their audiences.
Despite the fact that social media platforms are used by almost every business, the majority of them do not employ social listening tools. A research group, Alimeter, reported that 42% of businesses use social listening, while another research group, Amiando, reports that only 20% of marketers use social listening to benefit their business. Clearly, businesses do not utilize social listening tools, even though it can provide a wealth of information for businesses.
How Does Social Listening Work?
Social media is messy; there are grammar errors, spelling mistakes, and lots of incomplete sentences. Because of this, it is difficult for computers to analyze social media posts correctly. In order for social listening to work, computers must use Natural Language Processing (NLP). Natural Language Processing is a set of techniques that are used to clean text before analyzing it. This includes cleaning the documents of stop words (words that do not have much meaning, but appear a lot in a text document, such as articles and pronouns) as well as numbers, special characters, and converting uppercase letters to lowercase. This can be done by employees, or by different software systems that are made for social listening.
Let's revisit the example of the symphony’s marketing department. They utilize an online social listening tool, and search for all social media posts that contain a specific set of words, for example “Pennsylvania Symphony Orchestra Rachmaninoff,” in order to see what the general consensus was on their last performance. The computer searches for all social media posts that contain those words and then sorts through every post. Next, the computer goes through the text and performs the NLP steps. It corrects any misspelled words by guessing what word it should be, gets rid of "noise" words such as pronouns and articles, and gets rid of any special characters.
After the NLP process, the computer takes the words and processes them by clustering them into groups, determining sentiment based on a dictionary of pre-determined sentiment, or joining groups of words that were mentioned together frequently, such as “poor sound.”
The final result is what the marketing team analyzes. If the results show that the general sentiment of “Pennsylvania Symphony Orchestra Rachmaninoff” was negative, the marketing team can use this information to inform future strategy. Conversely, if the results showed that social media posts used “Breathtaking” or “Musical” to describe the concert, the marketing team will say that this soloist was a success and tells the Artistic Department to invite them back in the next few years.
If you are interested more in the process behind text analytics, this whitepaper by leading software company, Clarabridge, concisely describes the steps of taking text and processing it before analysis.