Wednesday, September 29, 2010

AdSense Forefather Makes 14 Million Business Listings Available for Free

Gilad Elbaz's last company was acquired by Google and became AdSense, the source of 30% of Google's revenue. His new company,Factual, is a marketplace for live, collaboratively and algorithmically maintained bulk data. This month Factual has announced that it now offers read and write access to 14 million U.S. business listings and locations - for free.Gigaom's Liz Gannes called the offering "a great place database in the sky." Adam Duvander atProgrammableWeb said it was "a first step toward creating a place database to which anyone can contribute...extremely valuable to developers." What makes this data offering so exciting? The ability to cross-reference it with any other location data that you already have and create something new.
The most obvious use-case is this: You want to build an app that checks where a user is physically located, then tells them what businesses are nearby. That data isn't readily and freely available otherwise, and even if you buy it from a major provider it's hard to know that what you get is up to date. Factual now offers 14 million listings in the U.S. and 25 million worldwide as just one of many data sets on the site.
The write access that Factual provides offers a solution to the problem of accuracy, the company says. Gannes explains that Factual "combines data from partners, vendors and users, and applies machine learning to extract and validate facts -- for instance, to determine what the actual current phone number is for a business." (See also our reviewof Factual's basic offering from October.)
A mobile "what's near me" app is just the most obvious use case, though. Such data could really be put to use in enriching any existing bit of location data. The Factual data set includes fields beyond name, latitude, longitude and adress - it's also got phone numbers, whether a business delivers food, whether it serves alcohol and much more.
The Real World, Mashed Into Apps
Imagine any of those fields being cross referenced with any other location data you've got.
  • Full bars with the highest Yelp rating within a quarter mile of any movie theatre.
  • Family-friendly restaurants within a mile of every historic landmark in the country, for those weekend educational trips with the kids.
  • 24-hour restaurants nearest any musical venue.
Every one of those mashups of location data is a Web or mobile app that people would use, aren't they? Those are just a few quick ideas.
The data set needs filling out beyond the basics; many of the fields beyond location and phone number hold sparse data right now. But the potential, the price (free), the pedigree and the programmability of this offering are all very promising.
Location as a platform, cross-referenced with place, time, people and content, could be a big part of the future of our computing experience. The first step is to get the data about what is where into as many hands as possible.
"Hopefully," Elbaz says on his very narrative LinkedIn profile, "high-quality, lower or no cost data will unleash a wave of creativity and productivity by application developers and content publishers.
"

No comments: