An equestrian suing for worker's compensation in Baltimore thought she had everyone fooled. The woman said in her claim she was injured and couldn't ride horses anymore, and in the courtroom, she had a convincing act.
But online, it was a different story. This was in 2002, when community boards served as social networks. There, private investigator Scott Catron found she had been spilling critical details about her life -- including when and where she would be riding next.
The woman's case quickly fell apart. She got nothing, and the insurance company saved more than $200,000.
Fast-forward to today, and even more of our lives are plastered over social networks like Facebook, Instagram and Twitter. That online explosion in personal information spurred Catron and fellow private investigator Michael Petrie to create Social Detection, a search engine that runs deeper searches on social networks and sites like Craigslist for their cases.
"We didn't have that before the internet came about, but once that it was here, we ran with it," Petrie said. "We knew that this was an emerging trend."
Our increasingly public lives have left a trail of evidence for investigators. Cops and private eyes don't have to follow people around in real life as much as they used to. They just need to open up Facebook. In October, police departments in Oakland, California, and in Baltimore were reported using data from Facebook, Twitter and Instagram to track protesters via third-party monitoring program.
Social Detection is one of those tools. Catron and Petrie built the program after spending more than 20 years investigating insurance fraud with long stakeouts and complicated disguises.
They had done web sweeps before, but that had required meticulous organization, scanning endless pages of posts to find digital needles in haystacks.
The pair needed social media searches that could go deeper than just typing a name in on Facebook.
The average investigator conducts 10 to 15 searches on a person, across several social networks and Google to get any worthwhile results, Catron and Petrie said. They're hoping to narrow it down to just one.
Too much information
People are still busting their own million-dollar fabrications by sharing too much on social media. The investigators estimated that Social Detection has saved insurance companies nearly $7 million in fraud cases since its beta launch in April 2015.
"Everybody wants to be seen or heard online," Petrie said. "They don't think they're gonna get caught."
Social Detection has been able to dig up people trying to hide behind disguise accounts that sport aliases or pseudonyms like their names spelled backwards.
It works by connecting the "digital dots," Catron said, using information available from the claimant's lawsuit. The data set typically includes a name, phone number and address. From there, investigators have access to names of relatives and possible friends, thanks to public records.
Even if there's a fake name, if the profile is linked to the claimant's family members and if photos match, investigators using Social Detection can put two and two together, Catron said.
It's often family and friends who help, inadvertently, to expose the lies, Catron said. The tool helps investigators comb through posts from friends and family members as well.
In one worker's compensation case, a man claimed he could barely walk 20 feet because of an injury -- and yet he was running half-marathons in Chicago.
He kept a low profile on social media, using his middle name to sign up, and didn't include photos. But a stranger's video of a marathon his spouse commented on showed him running in the background.
"They don't know it's out there. I mean, you can't control it," Petrie said. "If you comment or like or do anything on a public page, it will essentially be discoverable."
The team hopes to make Social Detection available on a subscription model, priced from $200 to $1,000 a month, but exclusively for insurance fraud cases. Catron and Petrie hope it could soon be used to help recent college graduates and job applicants clean up their own social media messes before someone else finds them.
Some soul-cial searching
Ghosts of social media's past can often come back to haunt people when they least expect it. Just ask Miss Teen USA, who faced a nightmare just minutes after winning the crown in July thanks to the Twitterverse finding her racist tweets from 2013 and 2014.
Using Social Detection, I looked into whether there were any skeletons showing in my closet. I've kept my social media presence pretty clean, but I was also a part of the unfortunate generation of people who created a Facebook profile as a teen. So yes, there's a lot of cringe-worthy material if you look back back enough.
I went back and deleted most of the high school hijinks on my Facebook when I was headed for college and again when I graduated, but Social Detection brought up some funny stuff I had forgotten about.
Like this Instagram account I made in 2012 to test out the app on Android -- the day after it had become available on Google's mobile operating system. The only photo was a test shot of my freshman year textbooks. It received zero likes. The discovery won't exactly cost me my job, but it did revive a ghost I'd forgotten about.
The account has since been edited to reflect that I use a different Instagram account. Hopefully the two followers on that account will transfer over.
The tool isn't perfect -- I had a hard time digging up dirt on the social media accounts I had vetted, like my Facebook profile, and it would take experienced web sleuths like Catron and Petrie to find obscure links to me, like a random YouTube video I might not even know I'm in.
"There's so much data out there that is missed," Petrie said. "So, do you want to find all of it, or do you want to find just some of it?"