Researchers from the University of Bristol measure public mood in the UK by analyzing tweets. In Brazil a team led by a computer scientist use Twitter to identify dengue fever outbreaks. HSE24, Europe’s largest home shopping network, analyzes social networks in real-time to identify customer complaints and answer them.
Whether in the interest of the public or the corporation, social data can be combined with custom Big Data to gain even smarter insights. No wonder that last month D&B, a major source of commercial information and insight on businesses, added data to their database from the web, Twitter, and other social media outlets. Let’s review what it can be used for and how.
What Is It Good For
Sentiment analysis measures the public’s opinion on a brand or a product. By gathering, for example, all tweets that contain the keyword "Nike", text analysis of the content could determine people’s attitude towards the brand.
Just like HSE24, you could also analyze social networks in real-time to lookout for complaints and handle them immediately. This helps to prevent public brand damage while keeping customers happy.
Finding opinion leaders
One of your customers could be a Tumblr superstar with over 5,000 followers. Wouldn’t you like to know that and engage in a conversation with her? By combining your CRM with Tumbler user info you could do exactly that.
Profiling target audiences
Social media could help you to find extra information about existing and potential clients, details such as gender, location, age, or hobbies. You could also learn about the people they are connected to and use it to improve your marketing efforts and generate new leads.
Connecting with customers
What about following all your customers on Twitter or connecting with them on LinkedIn? You could learn something new about your customers by reading their tweets and engaging with them directly.
Graph analysis is used to analyze social connections between people - who is connected to whom, which people have a lot of followers, etc. You could search retweets to see who is retweeted the most and is thus a major influence. If you measure the number of followers for each user who retweeted, you could also calculate how much exposure a tweet has gained.
Social content is mostly unstructured and in natural language, so text mining and natural language processing (NLP) can help to process it. Note that parts of the data, like Twitter hashtags and mentions, aren’t in plain language so standard NLP tools might not be able to handle them correctly. Custom processing should be used to gain the extra information that they can provide.
Social Data Sources
Gnip, DataSift, LivePerson Insight, and FirstRain are some of the major social data providers. They supply processed data from social networks like Twitter, Tumblr, and WordPress. Some also include data from Facebook although a lot of it is private. FirstRain does not - according to company CEO Penny Herscher: "We did an analysis of Facebook and decided it was not business relevant for B-to-B companies".
Integrating social data with your company data will help you gain deeper knowledge about your brand and customers. The data is readily available for the right price and the analysis techniques are at your disposal.