Even as one of the world’s largest location databases, Factual added more than 600,000 new records last month in the US alone, with each representing a new or updated name, address, phone number, or other pertinent information of local business. The company is in a never ending battle to have both the most comprehensive and most accurate places database possible – with each of these measures of success eternally at odds with the other.
One of Factual’s strategies for constantly gathering the newest bits of information is to enter data-sharing partnerships with its Web and mobile partners. Today, Factual announced a reciprocal partnership with TripAdvisor, one of the most popular travel sites on the Web, operating in more than 30 countries.
The Los Angeles big data company will provide restaurant data in 50 countries, including both places and reviews, and in turn will receive raw user-generated data from TripAdvisor travelers. In some cases the user data may reflect changing contact information, while in others, it represents an entirely new business – in others still, it’s simply bad data that must be flagged and eliminated by Factual’s systems.
“Our model has always been based on our belief in incentivizing partners to give us back their valuable, but still raw data,” Factual founder and CEO and former Applied Semantics co-founder Gil Elbaz says. “These crowdsourced edits come in, then get written to our API, and we are able to turn them around very quickly as structured data that is available to our users.”
It might seem trivial to document and deliver accurate places information, but it’s anything but. “We see 20 percent churn on average for businesses in our database,” Elbaz says. “Even the basics are more difficult to keep track of than you’d think, especially given that we’re talking global.”
Factual already provides access to structured data around restaurants, hotels, doctors, lawyers, and other small businesses to many of the largest names in the consumer Internet sector. Listed on its website are relationships with Foursquare, Yelp, LivingSocial, SimpleGeo and Trulia, while it is widely reported, though not confirmed publicly, that it also powers elements of Facebook’s and Google’s platforms.
Access to Factual’s places database is priced on a monthly subscription basis according to the number of users within a service. Partners that choose to share data with the company receive discounts, depending on the level of access granted.
“Sometimes we think we should charge them, because we’re doing so much work to clean up the data for them,” Elbaz says. “But in the end, we think it’s more valuable to us to have access to the data.”
Not all partners are willing to share data, the founder concedes. Others don’t even want to be publicly associated with utilizing Factual’s database. The reason for both goes back to a trend of the moment: Most every consumer internet company wants to be known for its content. According to Elbaz, many of these companies are worried that giving up their reviews, comments, and photos will weaken their unique value proposition. Others, he says, fear that consumers or analysts aren’t sophisticated enough to look past the fact that they use Factual’s data in ascribing value to their product or service.
“Most times they realize that the edits that they offer back aren’t really that differentiated from their competition in isolation,” Elbaz says. “But in aggregate, they raise all ships.”
At the end of the day, not all data that comes in from partners makes it out to Factual’s end customers. This comes back to the battle between comprehensiveness and accuracy. In fact, the CEO reports that Factual has more than two times the data in its database than it can “verify with a high degree of accuracy,” and thus make available to the public.
Factual is backed by $27 million in venture capital from the likes of Andreessen Horowitz, Index Ventures, Miramar Venture Partners, SV Angel, Felicis Ventures, GRP, Founder Collective, Idealab, and numerous angels. The company, which provides data on more than 64 million places competes with the Google, InfoChimps, Locu, and others this highly valuable category.
“Since launching a little over six years ago, we’ve been striving to get toward to a critical mass of canonical, ubiquitous, clean data used across enough sources that it becomes respected and accepted as the standard of truth,” Elbaz says. “We now have a number of the big names using us and we hope that when a few more pins fall it will solidify us as a true ‘factual’ database.”