Do you reveal a lot about yourself online? Maybe it’s time to stop Tweeting your every move and uploading personal photos to your Facebook page: Insurance companies may soon be able to automatically comb social networking sites looking for data on consumers that would tell them whether to sell you insurance. Insurers could — in theory, anyway — use what they learn to determine your car insurance rates, or set your life insurance quote.
Who’s looking at your Facebook page? Within a year — or possibly two — insurance companies will be able to automatically sift through the torrent of personal data online to help decide who is most likely to file a claim, predicts Craig Beattie, co-author of a recent report, “Leveraging Social Networks: An In-depth view for insurers.” The report was published by Celent, a research and consulting company that services the financial and insurance industries.
So far, there doesn’t appear to be widespread use of the Internet by insurers to fish for personal information about applicants or current customers, says Beattie. Data-gathering efforts presently are an expensive, labor-intensive process done by individual analysts. As such, it’s currently best suited for digging up dirt on high net-worth people who hold expensive insurance policies
Although some insurers are nibbling at the edges by creating their own social networks, they cannot yet conduct widespread screening of customers to use in, say, setting auto, home or life insurance quotes for the average Joe.
That’s because they can’t yet automatically or reliably link specific individuals with particular Facebook, MySpace or Twitter accounts. An insurer will have to know for certain that the John Q. Smith who has auto insurance with them is the same John Q. Smith who’s tweeting about his love for midnight street racing.
“At the point where the insurer can reliably say, ‘This person is the same as this ID on Twitter and this ID on MySpace,’ and [is on] this group on Facebook, maybe a group called ‘Driving Fast on Small Roads,’ that’s where insurers say, ‘If he’s in that group, we don’t want to insure him,’” says Beattie
Solving that problem should take another year or two, he predicts.
Whoa, say regulators
Technology isn’t the only obstacle. Insurers are regulated by state laws, and some dictate the criteria that can be used in their decisions. State regulations govern what insurers can use when setting prices, deciding whom to accept or reject and even how individuals’ privacy is safeguarded. As you’d expect, some states’ laws are stricter than others.
Even within a state, rules differ by insurance “line” (type of coverage). For example, in California, a state that prides itself on consumer protection, insurers must take three factors into account when setting the price of an auto policy:
- Your driving safety record
- The number of miles you drive annually
- How many years you’ve been driving.
California also lets insurance companies look at 16 other optional factors, things like vehicle type, driver training, academic standing, marital status, gender and claims history by ZIP code.
But the rules are different for, say, home insurance. Insurers in California can’t use other criteria — like credit scores — that aren’t on the state’s list.
Although insurers can’t use factors such as race, religion, creed or national origin for deciding whether or not to sell you a policy and pricing, some states give insurers wide latitude in selecting the criteria they use, says the National Association of Insurance Commissioners. Some list criteria they permit; some list criteria that are banned. Not all states require, as California does, that insurers demonstrate that the factors they use are based on sound actuarial principles.
There are three ways an insurer might consider using your personal information found online:
- To decide eligibility (whether to insure you)
- To “rate” you (set the price for your policy)
- To look more deeply into the behavior of someone who has a policy or has made a claim.
The last use — as a means of verifying the legitimacy of a claim — isn’t a radical departure from existing practices, says Joel Laucher, deputy commissioner for rate regulation at the California Department of Insurance. It’s much the same, he says, as using a private investigator to videotape someone who’s lifting heavy boxes at home while collecting disability benefits for a supposed back injury.
Last fall, a Quebec woman lost her disability insurance benefits after her insurer saw Facebook photos of her enjoying a vacation at the beach.
Nathalie Blanchard, then 29, was on leave from her job and collecting disability benefits, having been diagnosed with major depression, CBC News reported that her insurer, Manulife, stopped paying, pointing to photos she’d posted of herself on Facebook, “including ones showing her having a good time at a Chippendales bar show, at her birthday party and on a sun holiday — evidence that she is no longer depressed, Manulife said.”
The wake-up call would come when insurers start to screen policyholders based on evidence they find online about customers’ behavior or even the behavior of their friends and associates.
That’s what Celent is expecting will happen, given the irresistibly rich trove of information available on social networks. “Insurers will increasingly use public shared data to inform pricing decisions and aid in fraud detection,” says the Celent report.
For example, says Beattie, “Insurers in the UK have already shared between them a list of individuals they think are regularly involved in fraud.” A next step for insurers would be to search for those who are friends or associates of the suspected fraudsters.
Not that you hang out with criminals. But even your innocent behavior could suddenly have new consequences. “If you share your network of people [online], if you happen to ‘friend’ everybody you meet in a bar, for instance, it might well be that you’ve got some unscrupulous characters and that, in the future, you might be penalized by an insurer,” says Beattie.
All of this means that your friends’ and acquaintances’ behavior, including what they say about you and whether they name you as a friend — could begin to matter in ways it never has before.
Already, some states let insurance companies use applicants’ credit scores in their formulas for setting rates and identifying risk. Those insurers see a connection between paying bills on time and the likelihood you’ll file a claim.
Beattie imagines the day when insurers may take such connections even further: “Facebook already keeps track of who your spouse, your siblings, and your children are. If you share that information, why wouldn’t they get a credit score rating on them as well?”
Even so, some are dubious about the reliability of such data. Given all the bravado on the Web, it’s hard to see how an insurer could distinguish real evidence from gossip and lies. An online photo of you smoking a cigarette, for example, doesn’t prove you’re a smoker. Besides, “a lot of [what’s available online] is innuendo about a person’s lifestyle that a company would not be able to screen for in an underwriting guideline,” says Laucher.
Social media is likely to be useful not as a broad tool for vetting customers and setting rates but as a means of further investigating certain individuals whose claims or inconsistencies require a closer look, Laucher says.
But insurers will have hurdles to pass if they want to base rate and eligibility decisions on data they find on social networks. They’ll have to prove — at least in California — that the behavior in question really does cost them more.
They’ll also have to prove they’re screening people in a fair, objective way — that some consumers don’t have an advantage or disadvantage because of subjective factors, like the personal taste of decision-makers.
If society isn’t yet ready for the challenges to privacy posed by social networking, states aren’t either.
“I do think that the technology has taken us to the point where we are getting into areas of privacy that have not been contemplated,” says Rick Holbrook, a rate expert with the California Department of Insurance. It’s possible, he says, that regulators in the future will find some of these contemplated uses of Internet data too intrusive and ban them. Of course, those decisions will be made state by state.
Creepy and intrusive as all this sounds, Beattie is betting we’ll become accustomed to having our online lives screened for all kinds of commercial purposes. Many of us will become willing participants if, as he predicts, insurers also offer rewards for access to your personal data.
social networkCelent predicts that social networking sites eventually will embrace companies that want to access their communities. “Customers with large networks [of friends or followers] and a significant amount of influence may benefit from preferential rates and better service since their comments will benefit search engine results for the brand,” Celent says.
The price of opting out might be simply that you pay more. For example, insurers might charge more to insure a customer who’s less transparent than others, much as lenders do when they offer “no document” loans for which borrowers need not show tax records. They get their privacy, but it costs more as lenders factor in what they perceive as an added risk.
Finally, social media users themselves are helping to break down barriers to privacy as, for example, they share their location with friends, family and anyone else listening.
The privacy advocates at PleaseRobMe.com tried demonstrating the downside of such “oversharing.” The site listed “all those empty homes out there” made vulnerable to thieves as owners broadcast their absence through tools like Foursquare. PleaseRobMe has stopped, its point made, but a thief who is monitoring prolific Twitterers can know when they leave home for the gym for a couple hours.
It’s probably fair to say that, for enthusiastic social media users, these downsides seem distant and inconsequential compared with the fun right now.