IBM and Twitter Deliver First Set of Codeveloped Data Analytics Tools


Bookmark and Share

Since announcing a partnership in October 2014, IBM and Twitter have been relatively quiet about their shared ventures, but earlier this week, the companies delivered a number of cloud data services that will enable users to harness business insights from Twitter data. The companies also unveiled several developer tools for building social data–enabled applications.

With more than half a billion posts a day, Twitter has become the real-time pulse of the customer; its data provides a unique glimpse into consumer sentiment and behavior, but transforming that social stream into action has long been a challenge across the enterprise, especially for those in the customer service realm.

"We are at the forefront of enterprises making sense of social data. Twitter data is of interest to enterprises.... We see customers are more likely to tweet on positive or negative experiences than use any other social media network. As such, Twitter is highly interesting as a source for customer sentiment," Holger Mueller, principal analyst at Constellation Research, says.

Disgruntled customers often take to Twitter to complain about a bad product or service. Passengers waiting for delayed flights to take off, for example, are notorious for live-tweeting the situation and stirring up an angry conversation aimed at the airline. But by the time the complaints reach Twitter, there's little the offending companies can do to fix the current situation, so they engage in full damage control and offer customers discounts or other incentives to keep them from abandoning the brand. Once the problem passes—i.e., the plane takes off—companies can look back and learn from their mistakes in an effort to prevent a similar issue from occurring again. To date, Twitter hasn't been a tool for customer service crisis prevention, but the social network's partnership with IBM promises to change that.

IBM's new social data solutions leverage the capabilities of IBM Bluemix and Watson Analytics, as well as Hadoop. Using Watson Analytics, IBM customers can automate the process of pulling in data from Twitter and running a predictive analysis on it in order to build a model with the solution's visual storytelling features. Because the data merge happens automatically, users can pinpoint hidden patterns and relationships they might not have otherwise noticed to plan more informed actions.

Additionally, application developers that use IBM Bluemix now have access to Twitter data and analysis through IBM Insight for Twitter, a new service launched on IBM Bluemix. And, thanks to certain cluster configurations of IBM BigInsights on Cloud now having access to Twitter content, Twitter data can now be a data source for IBM Hadoop-as-a-Service offering. "Especially around Watson, expectations are high, [and IBM customers expect] that Watson will be able to help business users to make sense of social data and garner insights," Mueller says.

While much of the technology will be functioning behind the scenes, it holds potential for employees on the front lines as well. IBM has been working with more than 100 clients in varying industries that are using the technology to different ends, and helping customer service representatives has been one focus area. Because telecommunications companies tend to experience high customer churn rates, many have protocols in place to prevent customers from leaving, but few of these take real-time factors, such as weather, into account. To minimize churn, one telecom company worked with IBM to collect Twitter data on weather conditions in hyperlocal geographic areas and identify a correlation between conditions that interrupt service and unhappy customers on Twitter that eventually call to cancel their service. The data collected helped the company build predictive models and reduce churn by up to 5 percent.

IBM hasn't revealed specific details about the companies that have been beta testing these solutions, but the use cases for customer service providers will be vast, according to Mueller. Plus, there's more to come out of the partnership, and so far, analysts like what they see.

"The IBM and Twitter partnership has covered some good ground in the months since the announcement, [and there's] more coming...with IBM Social Media Analytics and IBM Watson Analytics.... We will have to see how well that offering will be accepted," Mueller says, "but we know that enterprises [always] look for easy-to-use tools and next-generation applications for better insights."