While most Twitter users post about news they've seen or what they're doing during the day, a number of hedge funds and financial firms are doing something else with the site: They're looking for information that can help them make money.
If they can find out what the masses are thinking about a particular company, or if they can be first to react to a news event, then maybe they can get an edge over their competition.
Of course, a hedge fund can't follow the nearly 500 million tweets per day that gets posted on Twitter, but Dataminr (CNBC Disruptor No. 43), a New York-based technology company, can.
The seven-year-old business, founded by Yale roommates Ted Bailey, Jeff Kinsey and Sam Hendel, has developed a complicated program that looks at millions of Twitter posts in real time for information that might be relevant to their financial institution clients — and other customers in government, news and corporate security.
Companies punch in key words, like sectors, names of other businesses, geographies, economic terms and more, and when something pops up, that investment firm gets an alert with tweets in it.
It also gets context: It can tell where the tweet is coming from, if there are other people tweeting about that event and who's doing the posting. Is it a trusted source, like an economist, a reporter or a company CEO? Or is it locals in the area who are witnessing an event? The program makes the information relevant; it's not just providing a stream of tweets that may or may not be useful.
"Once they put in their terms, they wait for a notification that our algorithm surfaces for them," said Steven Schwartz, Dataminr's president of news. "It gives them the earliest contextual awareness of events, and that helps companies make all sorts of decisions. They can keep an eye out for what's happening in a stock or a sector and incorporate that."
For traders, having an information edge is crucial. If a company can find out that a terrorist attack has taken place before everyone else, or if people are talking critically or positively about a product, it can use that information to buy or sell securities.
Nick Beim, a Dataminr board member and partner at venture capital firm Venrock, which has invested in the firm, said that users often receive information an hour before other news sites write it up.
"It's not 10 seconds faster, it's way faster," said Beim.
But can social media really help people make money in the market? A widely cited 2010 study by Johan Bollen, a professor at Indiana University, found that by measuring sentiment on Twitter, one can, in fact, predict whether the market will rise or fall three days later 86.7 percent of the time.
The results are less clear when it comes to individual strategies — it can be hard to pinpoint one single reason as to why a fund rises or falls — but Gene Ekster, a consultant who works with hedge funds to implement alternative data strategies, said that more companies are seeing the value in the data derived from social media.
"Investors are looking for ways to be savvy in terms of social media," said Ekster. "It is a requirement of doing business. It's not necessarily going to make your fund perform above the average, but if you don't have it, then it's almost certainty guaranteed to not perform."
Dataminr doesn't know what kind of returns its clients have made off the information it receives, but Schwartz said that its renewal rate is extremely high. Beim also points to one example of where Dataminr's technology helped clients.
On July 12, 2013, Boeing's Dreamliner, plagued by rumors of faulty batteries, caught fire on the tarmac in Heathrow. Someone tweeted about the incident, an alert was sent out to its clients well before it was reported in the news, and some funds got out of the stock. And it's a good thing they did — its stock price fell by nearly 4.6 percent that day.
"A lot of customers make a ton of money based on an early understanding of major events," said Beim.
He was initially attracted to the company in 2013 because he saw that the company was building more than just an application.
"It was pioneering a new science of real-time information discovery," he said. "That's difficult to do. You have access to the best data sets, machine learning, and you have to create a model that can understand things as it happens."
Twitter, which has a 5 percent stake in Dataminr, has provided Dataminr access to its "fire hose," which is the constant stream of tweets that gets created every day. Although anyone can look at public posts, the fire hose allows Dataminr to instantly analyze every single tweet as it comes in. It's the only company of its kind that has that kind of access.
When Dataminr first pitched their service, there was a lot of skepticism. Firms couldn't wrap their heads around how Twitter could be used in a financial context. Now, though, as the social media site's user base has grown, and after big events like the Arab spring have shown people just how fast news can travel, the sell is much easier. According to PitchBook, it's raised $180.4 million over five funding rounds, giving it a valuation of $700 million.
However, its financial-sector clients are mostly hedge funds and traders — long-only fund managers and retail investors, who don't trade on real-time news and likely aren't using Dataminr to generate returns or mitigate risks.
Peter Bailey, the company's chief strategy officer, still thinks it's useful for long-term investors.
"A long only investor may not change their conviction on their five-year investment horizon, but if you care about that stock, you won't want to be the last to know about a damaging event, and you won't want an investor to call you and say, 'What the heck is going on with this company?'" he said. "Whether or not you're actively trading, information is currency."
Going forward, the company will continue to refine its algorithm, its alerts and what it offerings so it can give people even more insight into the markets, businesses and events. Beim also said that it's going to enter new sectors — anyone who has a need for real-time information is a potential client.
For the investing public, Dataminr wants to become another tool to help people make better buying and selling decisions.
"It offers valuable intel," said Beim. "If you're going to change your position because of something happening in the world you want to understand it clearly. This will help inform that broader strategy."
— By Bryan Borzykowski, special to CNBC.com