Posts Tagged‘api’

The Internet Has Spoken: Platforms Are King.

If there’s one trendthat I’ve noticed about any of the successful Internet businesses of the past decade or so is that they tend to be platforms on which others can build their business. Sure there are many highly successful companies that operate in a closed fashion but the trend towards a more open web is undeniable. Nearly every successful Internet based company allows some form of interoperability with the wider world allowing anyone to leverage the platform for their own purposes. Thus today for any fledgling start up the choice on whether or not to open up your service for others to use has already been made for you, but that doesn’t necessarily mean its a bad thing.

There are many great examples of companies as a platform dating back to the early days after the dot com bust. One of the examples that sticks in my mind is eBay which started out as a simple way for anyone to sell their unwanted goods online. Quickly though people realised that eBay was in essence a cheap online shop front, much cheaper than many of the alternatives available at the time. This quickly snowballed and many niche businesses found their home on eBay using the brand to get exposure and the platform to grow a business that wouldn’t have been possible before.

The examples flow thick and fast for nearly all of the current Internet giants. Facebook has shown that whilst its core of replicating your friendship online remains it’s now a gaming platform and promotion network. Twitter owes quite a lot of success to its wide open API which has generated hundreds of quality applications, drove adoption of the service and makes it the de facto target for any Internet mash-up (even Geon!). So why does being an open platform do so much for driving adoption of a service?

Primarily it appears to be due to the amount of free development that one can receive by making their services available to developers. Twitter for the longest time didn’t have an official mobile application, arguably the killer app for something that’s based around short frequent updates. Still that didn’t mean there were a lack of clients available for it like Echofon, Tweetdeck and Brizzly. Opening up their API meant that they could focus more on improving the service and developing new ideas rather than having to spend additional resources bringing their platform to where it was needed. This forms a positive feedback loop that enables the underlying platform to improve whilst ensuring that it still remains relevant to its users.

Of course this all relies on the idea that your service provides something of value to your users. For a lot of companies the services that they provide start out closed off in order to ensure that it functions as expected. Early on development time is at a premium and the additional resources required to ensure the platform is stable can outweigh the potential benefits of doing so. However once a critical mass of users is crossed it makes sense to open it up in order to drive adoption. A great example of this is Gowalla who only recently released a full API after being available for about 2 years.

For someone like myself who is seeking Internet fame and stardom the idea of being a platform underpins many of the decisions I make when developing a service. You see whilst I may think I know what people might want there are so many things that I just don’t think of when I’m elbow deep in my code. In fact about half of the features in the current version of Geon have come just from talking the idea over with my friends and people who’ve been in the business for some time. Keeping my service open means that should an enterprising user find something lacking they’re able to build it hopefully bringing more users to my service and giving them a little Internet e-cred.

Does this mean that every service that isn’t a platform is doomed to failure? Absolutely not. There are many things where an open API simply isn’t required like if the company themselves provides products that cater to their user’s needs succintly. Still the writing is on the wall for those who build things on the Internet and the more open your application is the more likely it will be picked up by the wider world. Google VP Andy Rubin said it best with the words “Open usually wins” and the recent decade of the Internet seems to agree with him.

Good, Good. Let The Geo Flow Through You.

After losing around 6 weeks of my life to a project I’ve only vaguely referred to as “The Plan” I promised myself a month off to catch up on my backlog of games, relax and then hit Geon hard with new features so that I could release it upon the world. That month is now closer to 6 weeks mostly because there were just so many good games out at the moment (I still have 4 to play through, ugh!) and I was thoroughly enjoying slacking off. I’ve justified putting off the development because in all my research around the web I haven’t found anything quite like it. That was until I happened across this:

Facebook is allegedly planning to roll out location sharing capabilities next month, once again playing catch-up to other services that have gained popularity thanks to location data. The rumor comes courtesy of anonymous sources who have been “briefed on the project” speaking to the New York Times, who said that Facebook will announce the feature at Facebook’s annual f8 conference in late April.

The company’s plans for such a feature have not been entirely secret—Facebook hinted at location features when it updated its privacy policy in November. Like other postings made to Facebook, location information will only be made available to the people you decide to broadcast it to.

Great news right? I toyed with Facebook integration a while back but I never got the authentication working right and after a night of tinkering I found that I’d have to do some pretty heavy handed guessing to get the data in the right places. With services like Twitter and Flickr having beautifully easy geo-apis available I wasn’t too keen to muddy the information feeds unless I could guarantee a certain level of actual geo accuracy. So Facebook introducing a true blue geo api means more work for me, but also yet another hook for potential users.

But further down the article there comes a list of four services that have built their success based around user data that has geographic information in them. They are:

  • Foursquare: A location based game where you get points for going to certain places and checking in. Not even in the same league as Geon and I’ve known about them for quite some time. It’s a cool idea though and definitely has a dedicated following, something which made me confident that there would be at least some demand for an application like mine.
  • Google Latitude: Google’s attempt at a geo-oriented social networking service. Yet another attempt to get some more information on their users for better advertising with the added benefit of being an awesome cyber-stalking app. I don’t know anyone who personally uses this (and frankly I don’t think anyone cares that much where their friends are all the time) but it would have some potential if they opened the api. It looks like they kinda sorta did but since I can’t find anything official from Google on using Latitude it’s a moot point anyway.
  • Loopt: Looks like Latitude coupled with the added benefit of showing you other things that are near you (like events and places to eat). Again its main focus is the geo-social networking market which is a dedicated niche, judging by their website traffic. Since the focus of Geon is information and not social networking I still feel like they’re in another ball field to Geon, but I would’ve gladly used their services if they had a public API.

The last, and the one that made me almost jump out of my seat, was Brightkite. The front page has a lovely little section down the bottom that said “What’s happening in Canberra” and was happily displaying images that were supposedly from my area. Considering that I’m in the business of information aggregation with a geographical bent (trying saying that 5 times fast :P) my first reaction was that dreaded sinking feeling anyone gets when they think they’ve had the most brilliant idea in the world only to find it’s already been done, packaged and sold to everyone who would want it. However diving deeper into Brightkite’s world I can see that yet again they’re focused on the social networking side of things and much less on aggregating information based on location.

This I believe is the selling point for Geon. Whilst I appreciate that social networking is all the rage these days (as was blogs before them, and badly hacked together personal sites on Geocities before them and so on) it’s not the vision I have for Geon. It will have some kinds of social features in there (like following your friends, although Facebook integration might render that moot) but my main goal is getting information and hopefully establishing 2 way communication between people who want information.

The upside to all these apps coming out (and subsequently kicking my ass into gear) is that there is real demand out there for something like Geon. Talking to people about it only goes so far and it’s always good to have that little bit of hope that your work will someday be appreciated by the wider world.

Maybe I scare too easily 😉