A few years ago, I was stranded at an airport. Thunderstorms had disrupted travel everywhere. My fellow travelers and I had to wait until morning for the next available flight

As we stood in a miserably long reticketing line, a passenger started screaming at the agent. He didn’t want a $100 voucher toward future travel. He didn’t want a hotel room. Everything the agent offered made him angrier.

Without ever talking to him, I knew why. While we waited, I fired up Sonar, a now-defunct app that, as its name suggests, pinged, listened, and analyzed. On my phone’s screen were profiles of everyone within 100 yards who’d posted recent updates to Foursquare, Twitter, LinkedIn, or Facebook. So I knew that angry man’s name, where he lived and worked, that he loved barbacoa burritos with extra cilantro--and that he had a big client presentation in Denver at 8 a.m. (Intrusive? Maybe. But his data was public, he was making a big scene, and I was bored.) That $100 voucher was as useful as roller skates: Neither would get him to Denver in time.

But what if that airline could have defused his anger before it escalated? Enter the latest predictive tech tools, which wrangle personal data and preferences, crunch it all with complicated algorithms, and enable ultratargeted service that can better handle your customers’ individual needs.

Facebook is building a new feature for its Messenger app that, combined with what companies can glean from customers’ social streams, offers a potential solution. Businesses will be able to use the app along with public data to replace the online chat tools they often rely on now, which autopopulate conversations with ridiculous phrases that only make customers more irate (“Thank you for contacting our real-time chat…”). An agent could know, when a customer complains that a dress hasn’t arrived, that she’s leaving tomorrow for a wedding--and then send it overnight to her hotel. Such service pays dividends that far outweigh shipping costs.

IBM and Twitter recently built tools that mine Twitter feeds and weather data to identify consumers who are likely to fire off angry tweets if their cable service is disrupted. Those complaints aren’t empty threats: IBM’s data shows a correlation between disgruntled tweets and customer loss. IBM’s technology can scan individuals’ social media data and analyze their personalities to predict responses to an email or an ad--so, one day, businesses will be able to run complaints through IBM’s cognitive computing platform, Watson, to understand how best to address individual customer needs.

One new player to watch is Nashville-based startup Crystal, which makes an app that analyzes thousands of public data sources and shows how to communicate better with your social media connections. Enter someone’s name, and it pulls up a comprehensive profile detailing the most effective way to compose a message to that person. (For me, Crystal recommends getting to the point quickly and making me laugh.) Crystal also offers a Gmail plug-in, which acts as a sort of spell checker for interactions with customers. Before sending your email, it makes such recommendations as “Keep your email under 200 words. Otherwise this recipient might ignore it” and “Instead of super, use a less casual word like extremely.” Crystal will be available to the public this year, but hundreds of AOL, Forrester Research, and HubSpot employees have tested an earlier version. “Customer service reps” using Crystal, says founder Drew D’Agostino, can “serve people specifically, rather than responding just automatically from a script.”

Which would be a huge leap forward. The more you know about each of your customers, the better you can serve them. Even if they’re that angry guy at the airport.