Skip to Content

Watch out workers, it just got easier to sift your Instant Messages

Employees who want to speak on the sly have often turned to email: it’s a quick, discreet way to gossip about a colleague, and also to engage in more serious conspiracies. These days, though, it’s instant messaging – in the form of Gchat or Skype or Slack – where many workers go to swap ideas and opinions with their colleagues.

But while employees may like the fast, breezy format of “IMing,” the rise of instant messages have proved a headache for company bosses and lawyers. Unlike worker emails, which are easy for higher ups to locate and peer in on (yes, they can do that), instant messages are a motley jumble of data that is hard to parse.

This is a problem not only for nosey managers (who we hope have something better to do) but for company lawyers and compliance officers. If there’s trouble for the company, perhaps in the form of a lawsuit or a criminal investigation, the general counsel is often charged with identifying and preserving key communications – something that is much harder when any smoking gun is tucked in an endless string of “Hi.. LOL.. leavin’ .. brb.. ttys,” and assorted metadata like timestamps.

Now, however, the lawyers’ task has become easier with new analytics technology that can group and process big batches of IMs. The technology, which was launched last week by a San Francisco-based software outfit called Recommind, used predictive coding to herd the unstructured IM data into coherent packages.

“It’s a data analytics engine that analyzes words, context and concept groupings,” said Recommind product manager, Hal Marcus, by phone. To get an idea of what that looks like, here’s a screenshot that shows instant message communication patterns and activity levels:

IM Screenshot

Marcus went on to explain the new tool is not relevant only in litigation situations. He said it is also used by compliance officers who must watch for potential violations of foreign bribery statutes, and by officials at the Security and Exchange Commission. The tool can be applied to Google’s Gchat, Bloomberg messages or Skype. For now, though, Recommind is not designed to search Slack, the popular messaging tool that is displacing email at any companies (in a bid to add enterprise customers, Slack last year added a feature that made it possible for bosses to read messages in limited circumstances.)

In theory, Recommind’s tool could also help law enforcement scour reams of instant messages for criminal plots, but Marcus declined to say if any are using it; citing confidentiality, he would only say “other government organizations” are customers.


In the bigger picture, the new Recommind tool for IM is part of a larger story in which data analytics can be used to slash through the unfathomable amount of digital data generated in any modern office. The company also offers another tool that analyzes derivatives contracts for investment banks, who must collect and store such information for regulators.

As for employees, who may get a cold chill at the thought of higher-ups peering at their old IM chats, they better hope company bosses and lawyers keep these tools on the shelf unless they are truly needed.