Dataminr, a Twitter-Mining Service Used by Banks, Raises $30M

Dataminr, a startup that scours Twitter for market-moving information, has raised another $30 million in venture capital.

The New York based company said Wednesday it had secured the funding from a group of investors led by Venrock and Institutional Venture Partners.

Dataminr, which was founded in 2009, uses proprietary software to mine tweets for information before it appears in news outlets.

On Wednesday, users of the company's service learned eight minutes before news reports that Royal Bank of Scotland (RBS) is expected to cut up to 2,000 jobs — a report that caused the bank's stock price to fall 75 basis points after media outlets reported the story, Dataminr said. In May, six minutes before news reports, Dataminr alerted clients of the derailment of a CSX freight train near Baltimore — an incident that caused a nearly 2% swoon in the freight hauler's stock price.

A month earlier, Dataminr users learned of a false tweet that the White House had been attacked by bombers two minutes before the media reported the hack. Another boon came in April when the Securities and Exchange Commission said that companies can use Twitter, Facebook and other social media outlets to relay tradable information.

"In this day and age the number of signals we can find in Twitter that are before news are very high," says Ted Bailey, Dataminr's founder and chief executive. Twitter "creates a new frontline of information that is really the eyes and ears of the world in terms of ability to record real-world events and breaking information."

The latest investment brings to $46.5 million the capital raised by Dataminr, which has seen demand for its service swell in conjunction with the growth of Twitter as a global information outlet. Dataminr says it has signed up three of the world's largest banks and a number of big hedge funds, although Bailey declined to identify customers, citing confidentiality provisions in the company's customer agreements.

A pact with Twitter allows Dataminr to tap the service directly. The relationship gives Dataminr the ability to extract information from more than 500 million tweets that pour from Twitter daily. The unfiltered access matters, Bailey says. "If you shrink down…you don't find what's best, which comes from the un-followed," he notes. "People who happen at that moment to be the front-edge sensor for breaking information in the world, the first source."

Finding "early information" presents a big technical and mathematical challenge, notes Bailey, who adds that Dataminr also analyzes the information further to determine whether it has sufficient relevance for the traders, portfolio managers and others who use the company's service. "You have to algorithmically determine, is this important enough at this very moment that we want to divert the attention of a Wall Street professional? If you divert incorrectly, you may never get again," Bailey adds.

Users access Dataminr's service via a widget. The service also integrates with email and instant messaging.

Bailey says Dataminr expects to use the latest funds to hire data scientists and engineers and to expand beyond the financial companies and government customers the company serves currently. "While this provides an incredibly palpable advantage for Wall Street, there really are many industries that rely on real-time information," he says.