Originally posted on FastCoExist: “Using Advertising Technology To Data-Mine Your Social Network To Help You Out”
Big companies know a lot about you. They know if you’re pregnant. They know if you’re moving out of town. And they know if you’re starting a new job. Don’t believe it? A New York Times article from 2012 describes the eerie science of analyzing and extrapolating information from consumer habits.
Shlomi Aflalo, a veteran of that morally grey world of advertising technology, wants to use all of the data analytics methods he learned in the industry to do some good in the world. So he teamed up with Yishai Knobel, the former director of mobile health at Diabetes tech startup AgaMatrix, to createHelparound, a company that data-mines everyone in your Twitter and Facebook social network to figure out who can help you with whatever you need done–whether it’s a handyman to fix your toilet or someone nearby who has a spare painkiller.
“The engine is similar to what most smart advertising companies have. They have ways to profile users and match ads against the user. We’re building something similar except we are going to use it for good,” says Aflalo.
Think of the way Home Depot puts together an online ad campaign targeting a specific consumer group–in this case, handymen. Their first targets will be people who say they’re handymen or are interested in fixing things. Then the company can take that data and figure out other less obvious things that these people have in common–browser style, screen resolution, bandwidth provider, etc. These are small data points that seem meaningless, but actually contribute to building full profiles of who people are and what they do.
HelpAround uses the same tactics. “If a user is looking for a handyman at home, if she needs to install an Ikea table and is looking for help nearby from a neighbor, she comes to us and we have this set of users in her social graph, friends of Facebook friends, people who indicated that they are handymen,” explains Knobel. “We build their profile and complete it based on things that are not obvious.”
Knobel declines to divulge what exactly those things are, but he says Helparound only uses public data. “We don’t bother people. What we do is say ‘Here’s a person who is a friend of your Facebook friend. If you’d like, you can approach him.'” The platform is also good at categorizing items. If someone is hunting for a person to put together that Ikea table, it will know to look for handymen. If a user says they’re looking for Advil, it will know to search for nearby painkillers.
The platform will spook some people. But Knobel doesn’t think of it as a way to spy on your neighbors; instead, he likens it to a search engine. “If you ask Google a question, it infers what you’re trying to get at,” he reasons.
So far during Beta testing, people don’t seem to be too thrown off by strangers approaching them for help. “People are being blown away not by the fact that they can find these people [who can help out], but that people respond,” says Knobel. “It’s something magical.”
Knobel wont reveal exactly how Helparound plans to make money. But having a backbone based on advertising technology certainly won’t hurt.