We didn’t get a chance to attend the Web Summit in Dublin in early November. But we did manage to catch some of the livestreams from the event. And the talk by Twitter’s chief revenue officer Adam Bain contained a small nugget with a big message.
Bain is Twitter’s president of global revenue. And during his conversation on the main stage, he explained the social media company’s three businesses. The big one you know: advertising. The third one,commerce, is just emerging.
But the one that interests us is number two: data. Bain explained that the fastest growing part of Twitter’s business is now licensing the data it collects. And with 500 million tweets every day, that is a lot of unstructured data being hoovered up.
Still, how valuable is that data really to business? Tweets, after all, might be the poster child for unstructured data. To address that very issue, Bain told a story. Bain’s team got a call from a company that sells industrial fryers. Those are just what you imagine: Giant vats of oil that restaurants cook French fries in. Much as he’s eager to make sales, Bain couldn’t imagine why a fryer company would want to pay to license Twitter’s data.
In response, the company (Bain didn’t disclose the name) said it wanted to track anyone who tweeted the words “soggy fries.” Why? Because any time a fryer wears down or has a problem, the fries it cooks often turn out soggy. Thus, if customers are complaining about “soggy fries” on Twitter, there’s a good bet that the restaurant needs a new fryer, or some kind of part replaced. So the company wanted to use those tweets to generate sales leads.
“We think there’s a massive set of insights for businesses in the data,” Bain said. “What’s unique about this corpus of data is that we think it’s the largest set of public conversations out there. And there’s lots of businesses using this in interesting ways.”
Getting Access to Tweets
So what’s possible today? Traditional text mining is still pretty much the realm of academics and computer science PhDs. It gives you full access to everything on Twitter, past and present, but requires programming skills in Python or R to extract data through the Twitter API — and good grasp of analytics.
For marketers and social media folks, there are a few options for viewing Twitter data, depending on what you’re trying to do. You can:
- Search for hashtags using a tool like WikiScraper
- Access the Twitter Analytics Dashboard to see the impact of their tweets
- Purchase historical Twitter data using the Gnip tool, recently acquired by Twitter, for prices starting at $500
Consulting help is on the way. Bain talked about a massive build-out of professional services as part of Twitter’s new partnership with IBM. And that means businesses will soon have a variety of options for getting their social media analytics efforts underway.
Image credit: slgckgc