IBM and Twitter announced the first concrete results from a five-month-old partnership.
The new offerings include a number of cloud-based services that mine Twitter data, as well as new developer tools.
IBM is running the new data analytics services on top of its artificial intelligence technology and IBM BigInsights for Apache Hadoop, one of its big data platforms.
The goal is to help enterprises not just identify relevant Twitter data, but also exploit it.
What's the Deal?
When the partnership was announced, a lot of the conversation focused on who stood to gain the most. After all, IBM was already working behind the scenes with 100 early client deployments.
It was also building up its workforce. It has trained 4,000 engineers and consultants to use Twitter data in business projects. And it hopes to boost the number of these skilled workers to 10,000.
Still, in some ways, the partnership looked like a match made in IT heaven. Twitter is supplying untold amounts of real time data for IBM’s analytics, which are built on its Watson supercomputer.
IBM expects the Twitter data to be specific enough to enable enterprises to use it to develop strategies and make decisions.
In an interview with the New York Times, Chris Moody, vice president of Data Strategy at Twitter, called Twitter "the world's best focus group."
3 New Services
IBM will be able to combine the Twitter data with other sources of public information, including weather forecasts, sales information and product inventory statistics "to uncover powerful correlations that drive more actionable insights."
In a press release, IBM claimed the three new analytics services on the cloud will help businesses and developers:
- Social Data-Enabled Apps: Developers will be able to build apps that can search and mine enriched Twitter content and aggregated insights through IBM's Insights for Twitter service on IBM Bluemix
- Predictive Analytics: By automating the steps of data curation, predictive analysis and visual storytelling, Watson can give businesses the ability to pull relevant Twitter data into projects
- Twitter Data: Clients can combine Twitter data with IBM's full-featured Enterprise Hadoop-as-a-Service offering also available through IBM Bluemix.
IBM offered several generic examples of clients who could potentially benefit.
- Geography: Analyzing churn rates across global subscription-based telecommunications and media companies it was able to correlate weather events, angry Tweets and customer defections. Generally speaking, IBM claims that using this is can improve churn models by as much as 5 percent.
- Customer loyalty: IBM looked at Twitter data, loyalty information and the financial performance of stores and restaurants. It showed that staff turnover impacted negatively on company performance.
- Fashion Trends: IBM was able to identify worldwide fashion trends as they were evolving.
IBM isn't the only company Twitter has partnered with for data mining. And it's not the only company that can mine social data — think Oracle, Salesforce, SAS. But it's the only one that has Watson and business unit dedicated to such a supercomputer.