An often-overlooked, but critical and growing part of Twitter's business is its huge repository of historical and real-time data.
You might not think that celebrity tweets and people moaning that their flights are delayed could serve any revenue-generating purpose. But, using Twitter's data, a local burger joint can find out that it desperately needs to change its fries recipe, big brands like Dove can create and operate far-reaching social media marketing campaigns like #SpeakBeautiful that aimed to crush negative tweets about people's body image on Oscars night, and officials in cities like Jakarta can not only warn residents about flooding but check in real-time which are the safest evacuation routes. Those are all real examples.
Twitter's VP of data strategy Chris Moody spoke to Business Insider at Mobile World Congress in Barcelona and gave us TWO indications that Twitter's data business is about to become a massive deal this year.
How Twitter makes money from its data
Twitter really began ramping up its data strategy last year, with its $134 million acquisition of long-time data partner Gnip.The Gnip team, including then-CEO Moody, joined the Twitter flock within its revenue division. The data team now comprises 110 people, and Moody says this will expand — both in terms of number and internationally — in 2015
With the help of Gnip, Twitter is able to license its data. It primarily sells subscriptions to business software providers like Oracle, Saleforce, and IBM (more on that later) who use their big data tools to analyze the conversations about their customers. Twitter also sells data licenses to its more direct customers — advertisers — and it uses that data to provide consultancy to clients and, inevitably, sell more advertising.
Twitter's data licensing and "other revenue" totaled $45 million in the fourth quarter of 2014, up 105% year on year. It's just a fraction (9.5%) of the total revenue Twitter made in the period. But there's some major indications that percentage is about to get a whole lot bigger.
Yes, Twitter can make money from its logged-out users
In recent months Twitter's CEO Dick Costolo has been talking up the worth of its logged-out users: The millions of people who see tweets on other websites, TV, in newspapers, but aren't actually registered Twitter users. Earlier this month Costolo said a logged-out user is worth around $2.50 in revenue to the company per year (compared to $4 for a logged in user.) But ad industry executives told Business Insider that figure was "fanciful," mostly down to the belief that they contribute little valuable data.
Moody gave us the impression that it might not be far off.
In the simplest of terms, a marketer with more data on the entire reach of their tweets (beyond Twitter alone) can make better business decisions. An airline, whose tweet about a new route opening up, could use that data to work out which publications are best to direct advertising money towards once tickets become available, for example.
Also, Twitter is working on getting far more data about its "syndicated users" as it calls them, who view tweets outside of the official Twitter platforms. Earlier this month Twitter announced plans to syndicate its tweets and ads into Flipboard and Yahoo Japan. Moody reminded us that data share is part of the deal. This isn't just about creating a larger advertising ad network, it's about data mining too.
Twitter's partnership with IBM is going to help thousands of executives understand the value of its data
In October Twitter and IBM formed a partnership that the companies said would "transform decision making."The deal is huge — and not just because it means IBM's thousands of business customers now have the option to turn on Twitter data within IBM's dashboards straight away.
As part of the deal, IBM has also committed to training 10,000 consultants to train people how to use it. If each consultant only trained 100 people over the next year, that would mark a massive education program — one a small data division like Twitter's could never achieve at the same scale.
Moody told us that in the early days of Gnip, 50-75% of his time was taken up convincing people about the power of Twitter's data. "It's virtually no time now," he said, in part due to partners like IBM. A couple of weeks ago Moody met with the global chief marketing officer of a big brand and set up his pitch deck ready to go. He was stopped in his tracks by the marketer who said "you don't have to convince me, I want you to help me get started."
That's where IBM comes in. Marketers understand how useful Twitter's data can be in informing their business decisions, they just need help wading through the thousands of tweets about their companies and how to prioritize which insights are important.
"Other platforms are not always clear" about the public nature of their data collection
Twitter isn't alone in homing a hugely valuable real-time data platform. Facebook, Google, LinkedIn, and many more online platforms also offer similar services.
But Moody thinks it is transparency that sets Twitter apart and that's what makes it attractive to advertisers and partners.
He told us: "From our perspective, the vast majority of our data public and we are very clear about whats public and whats not. Tweets are public. Direct messages between your friends are not. Other platforms are not always clear on that."
And that's a good thing from a user perspective. If you're complaining about your plane being late (even without @ing the airline) and the brand gets in touch with live updates, that's a good user experience. And it's why customer service is a key focus for Moody this year.
Moody said: "My personal reason for that is we did the world a giant disservice in the 1950s when we introduced call centers. Big companies thought about customers service as an operational expense, reducing the cost of interacting with customers. Twitter is about brand building, and brand value often gets measured in billions for some brands, it's not about making a call cost 50 cents less."
Twitter clearly has big plans for its big data business that shouldn't be underestimated.
