One single tweet may not matters much. But how about a million tweets? When IBM announced its partnership with Twitter in October 2014, the two unlikely partners seemed to have a promising project. On March 17th, 2015, IBM introduces new tweet analysis tool for business.
By partnering with Twitter, IBM is betting part of its future to expand data by harvesting data in the form of tweets. Twitter, with about 6,000 tweets tweeted per second, or over than half a billion each day, is an endless source of data.
Both companies are bringing their products for the first try. The developer tools and cloud-based data analysis services harvest tweets for sentiment and behavior data, which will then be fed to IBM's Watson artificial intelligence. The developer tools allow people to write application that pulls in Twitter's data.
Insights works like most Watson-powered services using Bluemix. Users can submit a request via an API, and the tools will then search a stored of harvested tweets - updated in real time and going back as far as November 2013.
The requests can be as simple as a search term, or can be a sentiment analysis. The tool that supports reading tweets in 20 languages by the time of its release, will deliver the insight. A company can seek out all the tweets about a given brand, and see whether it is positive, negative, neutral, or ambivalent.
The project that moves around filtering information from gathered tweets isn't at all small. IBM has trained more than 4,000 engineers and consultants to use Twitter's data in business projects. The company's goal in to have 10,000 employees with Twitter data-handling skills. Before officially opening its new tweet analysis tool, IBM was already testing it with 100 corporate clients so far, in a range of industries and projects.
"We thought the early interest would be in consumer products and marketing, but it has turned out to be much broader than that," said Alistair Rennie, General Manager of IBM's analytics business.
By using what IBM calls an "insight as a service," IBM has learned some surprising discoveries: bad weather can make people want to ditch their telecommunication service. The more a customer shops at a particular store or eats at a particular restaurant, the more likely they are to stop shopping there when employees leave. And psychoanalyzing the tweets from popular fashion blogs can help clothing manufacturers predict which items will be big commercial hits and which won't.
Using the data-fueled predictive models, customer churn was reduced by 5 percent, said IBM.
IBM's partnership with Twitter is not at all exclusive since it isn't the only company that is mining social media data for insights. Rival companies like SAS Institute, Adobe, Oracle and Salesforce have all gone 'social'.
But what makes IBM's partnership different from its competitors is that IBM is focusing on using the data to a wide range of uses, beyond just marketing tasks. Other initiatives include analysis, talent management and product development.
The blend of corporate data and sources like Twitter is what would ultimately deliver the payoffs.
With the ability to give massive insights about any imaginable topics, IBM is confident that businesses would pay $2000 to use the Twitter analytics tool on it's Bluemix Platform-as-a-Service to analyze tweets and correlate them with external data sources (weather forecasts, sales information and product inventory stats).
At the moment of its release, the service allows users to search up to 1 million tweets for free while IBM works out a long-term monetization model.
"The first access to Twitter data is free, then it's pay as you go," said Rennie.
On Twitter's side, the partnership plays an even larger role. In addition to give more powerful insight to its advertisement methods, Twitter is now reporting its "data licensing and other revenue." The company grew by more than 100 percent in the fourth quarter of 2014.
"Twitter is the world's best focus group," said Chris Moody, Twitter's Vice President for Data Strategy. "We're a data gold mine. We know a lot about the world and a lot about customers."