When Google released their new Local search results format in October, 2010, SEOs quickly determined that the number of reviews and overall rating associated with a business were ranking factors. They also predicted that these would be quickly be gamed.
It was an easy ranking factor to launch the new integrated SERPs with, as they already had piles and piles of review data from their own properties as well as sites they’d partnered with. But Google’s algorithms are not static, and it couldn’t be too long before the Local algorithm included other ranking factors.
Now, it looks like they might be starting to think about other options. A paper submitted to the upcoming Very Large Data Bases conference, on Hyper-Local, Directions-Based ranking and written by two Googlers along with two other researchers, proposes a method for using direction queries…say, of the sort a user might enter into Google Maps…to determine interest in particular places that could possibly be used for ranking such places.
Greg Linden has a post summarizing the key points of the paper, and wrote, “the core idea is that, when people ask for directions from A to B, it shows that people are interested in B, especially if they happen to be at or near A.”
And of course, who happens to have piles and piles of data regarding direction queries?
I’m sure there are some people out there already thinking about how they could game this. But this got me thinking: what other sources of hyper-local, location-based, user-generated data is out here that could supplement direction queries?
GPS Trails on Android Devices
I’ve always believed that Google developed Android simply as a massive data mining engine. After the recent debacle where it was discovered Apple devices store GPS data, it was also uncovered that Android does too, which is not entirely unsurprising.
*removes tinfoil hat*
However, the paper addresses the main issue with GPS trails, in that “GPS datasets typically contain many positions for a few users,” and Greg points out that “just being in a location doesn’t make it clear that it is your endpoint or that you want to be there.”
All good points.
Location-Based Check-In Services
So what applications do people use when they want to be somewhere and want to tell people about it? Well, with increasing frequency, people are checking into location-based games like Foursquare, Gowalla, and Facebook Places.
Sure, this data would be heavily skewed to early adopters, owners of mobile devices, and geeks, but it could be another indicator of interest in places.
Google already partners with Twitter to use their data. The Twitter API includes a portion for GPS coordinates, and many mobile Twitter apps and third-party supporting sites like Twitpic have a setting that allow users to embed their coordinates in their tweets. Rather convenient, considering the number of people who tweet frequently from doctor’s waiting rooms, restaurants, grocery store line-ups, etc.
Of course, we’re not there yet. This paper is to be presented at VLDB in August, so it’s not even being implemented by Google yet. And by the time something like this is anywhere nearing integration into the Local algorithm, Foursquare and the like may not even exist. But with more users adopting mobile devices faster than desktop computers, and everyone under the sun trying to develop a location-based app, the opportunity is certainly there.