Twitter and the Power of Open, Integrated API’s
August 2nd, 2008
The executive summary of this post goes something like this: If you are trying to launch a service, product or other technologically “innovative” web presence, your idea either has to be absolutely stellar (which I will guarantee to you it isn’t), or you have to rip the covers off your technology and let your users decide how to use it themselves.
Lets take Twitter as an example. Their service statement is, in a nutshell, “We let you microblog”. They don’t do it particularly well, their online interface sucks, their features are extremely limited, and the service goes down so often that the “Fail Whale” has entered the colloquial as a term unto itself. And yet their users remain loyal, the service continues to be used, and it’s probably the most popular microblogging service out there.
Why? Realize that 80%+ of twitter activity goes through their API, not the website. In other words, the guys at twitter have opened up the doors to the kingdom and allowed the community to decide how they wish to consume the service. That same community has risen to the challenge, providing a plethora of applications, widgets and plugins that allow people to consume and update twitter where and when they want to.
Lets take a closer look at what actually happened here: Twitter provided an API… and let the community (whether individuals or businesses) suck up the cost of developing use cases and interfaces, do usability testing, write requirements and eventually write the software. In short they just made someone else pay for the entire involved and expensive front-end development process.
Not only that, but they also neatly sidestepped the need to perform strategic market analysis to manage their own growth. Rather than focusing on meeting the needs of a single segment (and expanding from there), they boiled the service down to its absolute essentials and let each segment figure out how they want to use it.
This approach isn’t particularly new. AOL IM and Jabber have been around for a while now, Google and Yahoo have been making their services available for years now, and most of all the world wide web itself is… well, a communication platform built on the premise of a server/client relationship.
The key is integration. There are so many services, websites and widgets out there today that only the truly exceptional will be able to stand independently on their own. All the others are fighting for search engine placement, social tags, advertising and community buzz to promote themselves, but are never likely to hit the big-time unless they can integrate themselves into their user’s pre-existing day-to-day activities.
Wouldn’t you call that the value statement of any new company? “Provide a service that improves a users day-to-day activity so dramatically that they can’t live without it”. Whether this is done via process improvement (Make a task more efficient / less frustrating) or activity enhancement (Now you can tag your pages while you surf them) doesn’t really matter, because the core task remains the same.
My favorite example of this phenomenon is Firefox, because it provides a platform where users can customize their browsing experience with all the ancillary services they want. Plugins exist for del.icio.us, livejournal, myspace, gmail, etc etc etc- all tasks and activities that may not be directly related to the core browsing activity, yet provide a sufficiently compelling experience enhancement that they’ve become part of day-to-day web surfing.
So in summary: If you’re providing a new service, put some serious thought into building an open API. If your services and your data (’cause that’s what it boils down to) are compelling enough, you stand a good chance of having the community take care of segmentation, marketing, and implementation.
Yes - this is similar to what Chris Brogan seemed to be saying in his post today - Why Twitter Still Wins. I love the notion of creating services, not sites.