Developer Workshop: Geodelic
Developer Workshop is a series of profiles exploring the current state of the mobile marketplace from the point of view of the software developers mapping out its future. Each profile focuses on a developer with a compelling story to tell, and offers their perspective on what the industry's doing right, what it's doing wrong and how to make it better. Check out our previous workshops on Shazam, InfoMedia, Viigo, Meet Now Live, Shortcovers and Pint Sized Mobile.
This week FierceDeveloper profiles Geodelic, which offers an application called Sherpa that provides localized information based on users' locations and preferences.
With the recent debut of its second Android device, the myTouch 3G, T-Mobile USA also introduced a new Android Market feature called AppPack to promote new and noteworthy smartphone applications. Among the free applications featured at AppPack's launch: Geodelic's Sherpa, a local discovery application that learns a subscriber's favorite types of locations and preferences over time, promising an increasingly customized user experience that recommends relevant retailers, restaurants and attractions according to one's likes, dislikes and behavioral patterns.
Powered by a learning engine dubbed GENIE--Geodelic ENgine for Interest Evaluation--Sherpa combines user location and interests with related contextual information like time of day, aggregating and presenting personalized, site-specific recommendations and insight about the places and things in the subscriber's immediate vicinity. "Wherever you are, we want to give you the information you need," says Geodelic founder and CEO Rahul Sonnad, who prior to starting the Santa Monica-based firm founded digital video Web services company thePlatform, acquired by Comcast in 2006. FierceDeveloper spoke with Sonnad about developing Sherpa, writing for Android and the media attention surrounding the app's inclusion in AppPack.
Rahul Sonnad on Sherpa's origins: It was all very organic--I got a broadband modem, stuck it in my laptop and started running around with a 3G connection and GPS. It quickly became apparent to me that if you can connect location with the kind of information you can get on the Internet, you've really got something.
As the possibilities of a connected, location-aware mobile device became apparent to me, the challenge was getting my head around the concept and figuring out how it would work. The core idea is so simple and intuitive, but learning how you make it work and scale it out took a long time to understand. You need to know where everything is, and you need some heavy taxonomy--getting content on various businesses and places of interest took a while. You have to figure out what's nearby and collect information on those places that are interesting. You have to pick which businesses users are likely to care about. From there, you need to decide what is the UI and the user experience metaphor--we did a ton of iteration on the user experience. A lot of it was trial and error.
On working with T-Mobile USA: We were originally writing this app for an unknown platform--we were writing it on tablet PCs with broadband cards. This was 15 months ago, when there was no iPhone 3G and no Android. Once the iPhone 3G came out, it became apparent that the iPhone was probably our target platform. But in our talks with T-Mobile, it began to make sense to focus on Android and on the myTouch. And frankly, the prospect of being app No. 1 on myTouch instead of number 60,000 on iPhone was very appealing.
We were talking with a lot of people in the mobile space, including carriers and OEMs, and it quickly became apparent there was synergy with [T-Mobile USA]. We think there's something profound and transformational happening here, and T-Mobile understood that. The touchpoints were relatively light--they weren't giving us detailed stacks or telling us how different features should act. I've worked with other carriers in the past, and this was by far the most painless and pleasant experience.
On writing for Android: We had heard good things about Android, and had a good experience developing on it--there were some challenges with rendering, but no major problems. Android is a much more open environment than the iPhone, which is great on several levels and challenging on other levels. As we move forward, it's possible Android will get more challenging--as it become less homogeneous, it will be interesting to see what that means from a developer standpoint. There are going to be lots of different manifestations. But in terms of launching applications and integrating with other applications, everything on Android is open, and you can do whatever you want. It gives you a lot of rope.
On Sherpa's turn in the media spotlight: At the end of the day, you need a product that's interesting--we have that. People get what this app does, so that sets the stage to receive media attention. But without the T-Mobile launch and being featured on myTouch, we wouldn't have gotten this type of coverage. That gets you into a momentum situation--as you get media coverage, you get promotion, which drives more downloads. Those downloads drive more coverage, which drives even more downloads. The challenge for developers is how you stand out. If you launched when the iPhone first launched, you got coverage--now, it is much harder to get that type of interest given the crowdedness of market.
On advice for aspiring mobile developers: They should stick with what they're doing--it's not even the bottom of the first inning yet. The amount of opportunity that will open up is incredible. Smartphones are going to take over as the dominant Internet device. So whatever you do, don't get off the mobile platform.
Beyond that, developers should try to find an area that is not terribly crowded--there are so many places where it's still a wide-open field. With so many devices coming online, this is the time to stake your claim and become a leader in that space. It will be much harder to be a category leader in a year or two. Do something simple you can do well, and keep pushing it forward.