It’s not only widgets, the entire mashup phenomenon that isn’t really working. I’m not usually such a naysayer, but for the sake of fellow developers I think perhaps I could call this one out. When I say widgets, I’m referring to widgets and mashups, both on the web and the desktop. I am also including Facebook applications (really, they fall under the widget category).
So why do widgets suck so much?
The first and often noted problem is that most are sheer novelty. Knowing who has read my weblog today or what the weather is next Tuesday is just sort of fun to know. Seeing a map of where in the world people are Twittering is something I only glanced at once. It’s difficult to maintain attention and grow your audience if they don’t really care to see your application every day.
From a business perspective, widgets are leeches. You are completely dependent on another service for your business. Your software is at their mercy. How is that different from being dependent on an operating system or on the web? For one, the Internet isn’t controlled by any single corporation. Developers love this because although there are certainly standards and limitations, these probably aren’t going to abruptly change so someone can make more money. Secondly, developers really have been wary for a long time about cross-operating system compatibility, but we also trust that operating systems developers care about maintaining applications that run on that operating system. It’s more of a symbiotic relationship.
Website developers, however, do care that people are making widgets for their site. It’s a different type of caring though. Website marketers see having an open API for their site as gaining unpaid labor. “Shh… don’t tell the developers that they’re just promoting our site for free.” The widget/mashup will never be as popular as the service it’s promoting and the service developers are all too happy to spend some time creating an API for the amount of publicity it will generate. The actual development of widgets/mashups only increases this publicity for them.
Speaking of unpaid labor, how are any of these widgets making their developers money? Licenses? Subscriptions? Ad revenue? Paid services? By building off another company’s platform, your company lacks complete control of revenue.
Widgets can be used as a promotional tool to draw attention and users to your software, but I’m wary of trying to make an entire business dependant on another, more compelling piece of software. I’d much rather be developing that compelling service. I can only see two cases where developing a widget/mashup could be a good thing for a developer. One, you are using it to promote a larger application and it’s not your only means of promotion. Two, you’re a student, in need of a portfolio item, or learning a new language. Widgets/mashups really shouldn’t be whole businesses and developers shouldn’t see a widget as their opus.
Then again, perhaps I just don’t get it. Maybe some developers really only have widget-sized ambitions. Maybe I’m just not one of them.