5 Tips for Mashing up Big Data, Social Media

James Maguire

Updated · Mar 14, 2013

Can Big Data help enterprises make sense of their social media efforts in order to create more business value? Does the mashup of Big Data and social media enable organizations to understand their customers better and establish a higher level of customer loyalty?

These were among the questions addressed by Dr. Cheemin Bo-Linn, president and interim chief marketing officer of Peritus Partners, and Pam Didner, global integrated marketing manager for Intel Corp. during a session at the recent Online Marketing Summit.

Bo-Linn said enterprises are facing an enormous surge in data — aka Big Data — due to globalization, new technologies, and especially from increasing activity in social media networks. She cited the volume of social media activity as one example. Every day, seven terabytes (TB) of data flow through the Twitterverse while Facebook handles 10 TB of data, Bo-Linn said, citing research from IBM and Peritus Partners.

Social media interactions can give enterprises business intelligence that can drive what-if” scenarios and provide insights into customer behaviors and the impact of social media campaigns. “It can give you actionable insights,” Bo-Linn said.

Mashing Up Big Data, Social Media

Bo-Linn mentioned Netflix, Intuit, Amazon and Nike as among the enterprises successfully mashing up Big Data and social media to create greater business value. For example, Netflix has a wealth of Big Data available for analysis as a result of the viewing, rating and other activities of its 27 million U.S. subscribers, she said.

“Big Data helps Netflix make more informed business decisions,” Bo-Linn explained. “It helps them understand the content to bid for that they know their audience will respond to. Instead of just trying to offer something for everyone, Netflix uses Big Data to more precisely identify the content that will get them the most views. That is helping them in their goal to be the ‘HBO of streaming video services.'”

4 Types of Big Data

Didner acknowledged that the mashup of Big Data and social media can help enterprises make better business decisions. But at the moment, there is no one tool that pools together all types of Big Data in order to provide actionable insights, she said.

“Each tool presents siloed information,” Didner explained. “At Intel, we’re trying to bring all those silos together to see if we can find weekly, actionable insights so we can better optimize our social media communications.”

Didner outlined the four types of Big Data, as defined by Booz & Co.:

  • External structured data such as U.S. census data, credit histories and real estate records.
  • External unstructured data, which typically comes from Google+, Facebook, Instagram, Twitter, Pinterest and other social networks. This type of data represents the largest area of opportunity for the enterprise to gather valuable consumer insights, Didner said.
  • Internal structured data, such as HR records, CRM data, inventories and sales records. This category is best understood by the enterprise, Didner said. But the enterprise needs to shift its focus to external unstructured data to better understand its customers and create maximum business value.
  • Internal unstructured data including text documents, sensor data, SharePoint files and online forums.

5 Tips for Big Data Success

Despite the lack of a tool that brings together all four types of data, Big Data and social media can be useful for enterprises if they follow these steps, said Didner: 

Clearly define the problem to be solved using Big Data, so you’ll understand how you need to use the information. For example, you might think you want to use Big Data to generate more customer sales leads on Twitter. But the problem you may really need to solve is getting more leads across all social media channels.

Known your “knowns” and “unknowns.” What facts do you already know about your customers and competitors? How does your company fit into the competitive landscape? What are the things you will never know, such as the subjective decisions people made to visit your website?

Find the appropriate tools. Clearly defining the problem to be solved will help you choose the most appropriate social media analysis tools, whether it’s Sysomos, Radian6 (acquired by Salesforce.com), or another social media monitoring tool.

Test your hypotheses. Once you have gathered data, filter it and look at it from multiple perspectives (such as over different time frames) to test your hypotheses, Didner recommended.

Draw assumptive insights. Ultimately, the data you get and test should help an enterprise arrive at well-informed assumptions and insights, which can then guide their actions in social media or in other customer-facing channels.

“What’s most important is to constantly test the insights you get from Big Data and social media,” Didner said. “You have to separate the signal from the noise, and there are no short cuts. It takes intentional effort to get those actionable insights out of the big chunk of noise.”

James A. Martin writes about social media, SEO, and online reputation management. Follow him on Twitter, @james_a_martin.

More Posts By James Maguire