IBM (IBM 0.82%) and Twitter (TWTR) recently launched their first co-developed data services and developer tools, which will use IBM's cloud-based data analysis services to mine Twitter's data stream. The two companies announced their partnership last October, with the goal of growing revenue at IBM's cloud business and Twitter's data licensing segment.
Twitter's big data "fire hose" delivers over half a billion tweets daily to a wide variety of companies. IBM will help process that feed with its AI platform Watson and IBM BigInsights, an open-source software platform for processing big data across computer clusters. Their developer tools will let programmers write custom applications that rely on mined Twitter data.
The business of "insights"
Big Blue didn't disclose which companies are testing out the Twitter tech, but it highlighted a few applications in a recent New York Times article. For example, merging tweets with real-time local weather data helped a telecom company get ahead of figuring out where disgruntled customers might be tempted to switch providers because of a weather-related service outage. According to IBM, that model reduced customer churn by 5%.
This convergence of social media and data analytics is referred to as "insight as a service." Plenty of other big tech companies -- including Adobe (ADBE 2.95%), Oracle (ORCL 0.85%), and salesforce.com (CRM -0.29%) -- are also mining social media data to offer similar "insights" to clients.
What the deal means for IBM
For IBM, the Twitter deal represents a way to accelerate the growth of its cloud business. Last year, IBM's cloud revenue climbed 60% year over year to $7 billion, but that represented a mere 7.5% of the company's total 2014 revenue.
IBM is depending on the cloud business to offset falling revenue in its core services, software, and hardware businesses. IBM's top line has declined for 11 consecutive quarters, while aggressive divestment of low-margin businesses have stalled companywide sales growth.
To turn its cloud business into a new pillar of growth, IBM acquired cloud computing infrastructure company SoftLayer Technologies in 2013. Last year, it spent over $1 billion to launch a dedicated business unit for Watson, signed several major hybrid cloud deals, and inked a cloud-based enterprise app partnership with Apple (AAPL -1.00%).
CEO Ginni Rometty believes Watson -- which had lifetime revenue of less than $100 million as of October 2013 -- will evolve into a $10 billion business within a decade. The Twitter deal, which taps into Watson's ability to "learn" by processing large amounts of data, represents a promising step in the right direction.
What the deal means for Twitter
Twitter's fire hose subscriptions are reported under "data licensing and other" revenue, which rose 105% year over year to $47 million last quarter. This means nearly 10% of the company's top line is dependent on pumping billions of tweets to customers.
However, data licensing is a double-edged sword for Twitter. The company licenses its fire hose feed to companies, news agencies, and even high-frequency trading firms. This can be a toxic mix if left unsupervised. In April 2013, a fake tweet from a hacked Associated Press account claimed that President Barack Obama had been injured in an explosion. The Dow Jones Industrial Average subsequently plunged 100 points, but bounced back within three minutes. Today, plenty of penny stock pumpers are still trying to hijack this system with fake and bot accounts.
Fake and bot accounts are a serious problem for Twitter, since they obfuscate its real number of monthly active users, or MAUs. Twitter admitted in its latest earnings report that 8.5% of its reported MAUs, or 24.5 million "people," were either fake users, bots, or third-party apps. However, a recent study from Music Business Worldwide found that two-thirds of top pop music stars' Twitter followers were fake. Those fake accounts undermine Twitter's attractiveness as an advertising platform, but retaining them inflates MAU counts and keeps data flowing into the fire hose feeds.
That's where IBM comes in. With Big Blue's help, that river of chaos can be organized into useful information, and developers can create apps that filter and extract the data they need.
The key takeaway
I believe the deal between IBM and Twitter is a win-win situation that will generate fresh revenue for their respective cloud and data licensing divisions. Twitter is a gold mine of information, but its data licensing customers are sometimes tricked by fool's gold planted by malicious tweeters. IBM is providing companies with better mining tools to help them find the data they need.