There are nearly 5,000 APIs. We can connect services and build just about anything. By "we," I mean developers. Regular people were left behind by what now seems like a passing fad, widgets. The web would benefit from a resurgence of widget-making and developers are just the sort to take advantage of the opportunity.
The most popular widget of all time is still going strong. YouTube's initial virality came from regular people embedding its videos into their MySpace pages. Widgets like this allowed anyone to share content across the greater web. It became a way to aggregate your actions across multiple sites. Widgets were available to anyone who could copy and paste. Widgets were agnostic to the blog software you used or even if you used blog software at all.
The widget became so prominent that VC Fred Wilson complained that they slowed down his blog, as widget after widget loaded in the sidebars. He later argued that widgets should be moved outside the sidebar and become part of the content. But that hasn't happened and I'm not sure that it's just because of slow load times.
Instead we've seen widget companies, like MyBlogLog, get acquired and forgotten. It may sound like I'm arguing for a bygone era, like I think we should all be hand-writing letters and sending them via Pony Express. We could still use widgets now and APIs could help developers seize the opportunity.
Why did they fall from favor? Perhaps it's developers becoming attracted to the technical superiority of APIs, or the influence of mobile, or the slow decline of MySpace in favor of Facebook and its stringent customization patterns. What we lost with widgets was a way for the web to come to you, rather than the other way around.
Recently I was helping my friend with his band website. In a previous version, he used a WordPress plugin that connected to the Songkick API to display upcoming shows. No longer on WordPress, my friend needed a solution to embed tour dates. Songkick doesn't have a widget, just an API.
Several years ago I might have made an opposite argument. APIs provide more flexibility than a widget--even the widgets that let you customize. In fact, I'd still make that argument, but I'd add that these serve different audiences. Telling my friend that Songkick has an API that does what he wants and more doesn't make him a programmer.
If most API providers are no longer making widgets, that leaves an opportunity for developers. Indie musicians are just one market, but if you don't think there is money there, ask Mobile Roadie or CDBaby. They solved technical problems for independent musicians and were paid for the service.
Sure, there are do-it-all mashup builder sites and those would be helpful here. But I think regular people look for a service that can solve their exact problem, not a bunch of anonymous problems they might have someday. Widgets took specific use cases, such as sharing your bands tour dates, and made them sharable wherever you wanted with a minimum amount of work.
As our API directory lumbers toward 5,000, we're doing our best to help developers make sense of the options. We should all be doing the same for regular people. Again, "we" means developers. Maybe you. And with luck we can also call you widget-makers.