GigWalk, Centzy and Locu: Meeting Demand for Better HyperLocal Data
This week there is a LBS Apps developer meetup happening here in NYC that will be focusing on working with POI and venue data… an area that has always been of particular interest for me dating back five or six years ago driving along Central Park South with my Garmin Nuvi and noticing all the garage and auto repair places that Garmin said were there but that didn’t reality exist. I am still not sure if it was just a GIS nerds’ idea of a joke, or was the state of data really just that bad. Well fast forward to today and some respects we’ve made a lot of progress, and in other ways we’re still at square one.
Seemingly for centuries, mapmakers focused solely on getting from point a to point b, but largely ignored describing the stuff in between in any detail. And even today it seems if you want to create a mobile app that covers all the places around you, best practices seem to indicate that it’s necessary to use data from many different providers doing a lot of cleaning and matching along the way in order to paint a clear picture. The definitive database of places is still elusive it seems.
That said there are some quite interesting start ups, who aren’t letting that get in the way of trying to gather together even more and greater detailed information related to places. A few that come to mind include companies like Gigwalk, Centzy and Locu.
Gigwalk, seems to take the approach that attempting to gather much of this information via anything other than direct 1st person interaction is fruitless, so what do they do? They pay folks… anyone that downloads the app… to pick up micro, $3 or $4 gigs by walking into retailers and taking pictures, noting the address and generally gathering information about the place. And in this economy with the emergence of services like TaskRabbit, I am sure the supply of GigWalkers is plentiful.
Centzy, seems to also be taking on the direct approach of paying folks to gather things like store hours and pricing, currently focusing on service business in NYC and SF including “facials, hair salons, manicures, massages, pedicures, yoga.” A great article in VentureBeat quotes Centzy as saying that ” less than 25 percent of local service businesses put their prices online” which must make the manual data gather for those categories a necessity.
But evidently restaurants are putting their menus online in mass, which is why a company called Locu is taking a different approach by crawling, scraping and otherwise electronically gathering menus from across the web, and throwing good old patent pending machine learning against it. Another great article in the Huffington Post covers the all the details, including the launch of MenuPlatform and Locu’s CEO Rene Reinsberg’s quote on the company’s goal, which is beyond just menus to “create the world’s largest repository of semantically annotated real-time small-business offerings.”
To me what Locu is doing seems the most exciting and interesting and maybe challenging because I’ve felt for a long time that the necessary rich place based data simply doesn’t exist yet on the free Internet. If it did, Google should have already indexed, parsed and done whatever the hell Google does with data to give us better local search results that bridges the divide between digital and physical. But they haven’t… if you really think about it local mobile search result still generally suck. Locu seems focused on trying to fix that and is doing it seemingly in a very Google-ly way.
I noticed a similar dynamic and two different approaches related to gathering local product inventory between companies like Retailigence (manual data gathering) and Goodzer (web data gathering) and am not sure there is one right approach, and like with the current state of POI data perhaps for a while best practices will require using a little of everything.
On a related note Greg Sterling brought up an interesting point related to the Locu business , which is even if you are successful in gathering good valuable data, with so many others also focused on this area, will it simply become a commodity and how do you form a profitable businesses around being a data supplier?
Perhaps, if you do it well enough you can simply build your own local search application on top of the data?
Comments
-
Sponsored Links
