Business symbiosis

A lot of great Web 2.0 apps out there have released APIs to allow organizations and individuals to remix and integrate their existing applications with this new breed of web apps. APIs provide hooks into their applications, allowing data to be added, updated, deleted or queried by 3rd party programs.

An unexpected outcome of this, however, may be the creation of entirely different applications integrated closely with 3rd party applications (dependent on them). This has been anticipated by some web applications, and they’ve gone so far as to dub their API “a platform” (Facebook for example). What is the difference between an API and a platform though, and can/should applications be built on top of an API?

Back it up

Everything from project management applications, to online storage, to billing and invoicing applications have exposed themselves through APIs. Google does it with a large number of their products: we’ve seen a proliferation of great data visualization applications become available thanks in large part to the Google Maps API.

Up until now, this exposed access has led to some really great features within applications. You can import your client list from one application into another. You can display a map (or satellite view) of an address stored in your contact list. You can deploy an application on a social network and instantly have access to millions of users that will recommend it to their friends. With the exception of Facebook applications, however, there have been very few products that have taken it from the feature level to the platform level though and built applications that simply could not stand on their own.

Why not?

There’s always that fear that the company providing the original product will simply integrate what you’ve done as a set of new features. Instantly, you become irrelevant. It happens on the desktop all the time — Apple has often been criticized for “stealing” the function of a 3rd party app and simply integrating it into the OS.

There’s also the possibility that the originating company will change the terms of service for their API, and institute some restrictions that cripple 3rd party products. I don’t know the whole story behind FriendConnect and Facebook; but it’s a classic example of what could go wrong when you rely on another company for access to their product. Finally, they could just shut you down altogether with no notice because they don’t like what you’re doing. Unless you’ve got the time/money to get into a legal battle over it, you’re better off just moving on.

The advantages

There are some huge advantages to be gained from building off of an API though. You receive a massive jump start in functionality. In some ways, it’s almost like working with a framework on a massive scale; you don’t have to worry about a lot of application details because they’re dictated to you.

Depending on the company, you may have a massive PR/exposure boost right out of the gates. Some companies communicate with their users about any new developments, either internal or external to the company, that could have a positive impact on the way customers use their product. It’s in their best interest to keep the buzz going about their products.

Depending on the product, there is already a certain established “behavior” with customers. For example, users of various web-based applications will be used to a subscription model, and will be comfortable having applications hosted off-site. They’ll be comfortable actually paying for their applications (if they provide value).

Photoshop Plugins

I like to use the Photoshop plugin analogy here. A lot of companies make a good deal of income off of Photoshop plugins. It’s software that can’t exist without another piece of software (Photoshop) made by a different company (Adobe). Photoshop does a lot, but it doesn’t do everything — so various plugins have been released to fill in those gaps.

A lot of people have said that web apps will eventually replace desktop apps — is this another step forward in that space?