SimpleGeo

Returning home after our first class meeting, I found a weird email waiting for me.  It was a user key for simplegeo.com.  At first I thought it was another aspect of class to go along with lists, blogs, deliciousness, etc.  As it turned out, the timing was largely a coincidence.  I had asked for a trial user account for the SimpleGeo beta back in October after receiving a message on the siac listserve sent around by Puneet Kishor.  Now that that’s settled….

SimpleGeo is a newish take on the geo-web.  They say they want to do for the geo-web what PayPal did for the online shopping sector.  Apparently, that mission statement is just vague enough to get them $1.5 million in seed money:

http://www.techcrunch.com/2009/11/30/simplegeo-funding/

The most interesting thing I have been able to find about this company is that their business model was developed because they failed to get an augmented reality game off the ground.  They figured, “We’ll program that geo-web stuff next month, and then role out the game in May…”  Turns that geography stuff is really difficult, so their game never got finished.  But then they had an epipheny, “Hey, I bet a lot of companies struggle with this stuff.  Can we provide a service (and make some money) by doing the geo-processing needed for online applications?”

My answer: no.  Best of luck to them, though.  As we are about to find out with the twitter project, the current state of geodata on the web is abysmally complex.  My experience is that it takes a month of failed attempts to fully discover all the limitations and possibilities of geodata.  What are the limitations and possibilities you are referring to Matt?  Well, that depends on your application.  In our case we’ll have a lot to figure out on the precision of geotagged and geolocated tweets.  In my flickr project, I have a lot to figure out in order to classify users as ‘insiders’ and ‘outsiders’ for the flickr neighborhoods.  These are just two little projects, but the time investment for each one will on the order of 100 hours.  And most importantly, it will be the lead developers putting in the time.  The only way to figure out how to use web-based geodata is to … well … use it.

SimpleGeo thinks companies can outsource this part of development.  I doubt it.  PayPal was quickly and hugely successful because it plugs into the web application to fulfill a relatively simple, repetitive task.  PayPal gathers payment info, verifies accounts, and transfers funds.  While the banking and regulatory aspects might require lots of lawyers and bankers (hence the success of PayPal’s approach to taking this off the hands of smaller websites), but the interaction between the users and the websites are very, very narrowly defined.  Geodata?  It’s a wonderful mess right now, and it’s hard to see how prepackaged database and api services are going to meet the needs of application developers.

On the other hand, if they’re hiring, I’ll take a job, please.