Mashery: Syndication services for a world gone meta

Feedburner has shown that managing RSS feeds and providing usage statistics is a solid business. Now Mashery is going to measure how individual application programming interfaces (APIs) are being blended to create new Web services, making the popularity of a given API a potential foundation for business development and revenue growth. It's an idea whose time had come.

Where data comes togetherMashery's metrics will enable new approaches to pricing API access. are the new value flash points. Mashery, a San Francisco-based startup launched today by Oren Michels with funding from Josh Koppelman, is setting up business at the critical intersection. Dan Farber provides many of the details in a posting earlier this morning.

What I want to focus on is why this kind of business will be good for developers who take the time to build a valuable API and then market it.

An API used to be a great way to control value in the market. Think Windows API and you see what I mean: The Windows API was the foundation for any functions an application running on a PC might need to call—file management, windowing, graphics, etc. Without the API, nothing much happened on a PC, because so many features of an application had to be written from scratch.

Now so in the Web services world, where APIs are more like the ornaments that make a Christmas tree sparkle. The tree was already there, it is augmented by the decorations. Web services APIs don't lay an unassailable groundwork that locks in business, rather they offer to augment and modify data that's already available on the Web, adding functionality or algorithmic insights to make the data more useful, easier to access and manipulate—APIs do lots of things.

But not every API is a business, even though there are plenty of companies that are earning real revenue selling API access. No API is a business from the start. It must be promoted in order to find developers who will adopt it just to get traffic flowing. Even then, when there is traffic, it is not clear that every API developer should be thinking in terms of charging.

Mashery's service includes fee-based access, but I am intrigued by its free services, which are designed to let the startup API into a measurable game, where it can prove its popularity before the question of making money becomes the central issue. You see, if you can drive a lot of usage, it isn't necessarily the case that making a per-call fee is the best way to turn an API into a business.

With usage statistics, API developers may try some alternative approaches to revenue, such as performance-based compensation. For example, an API that concatenates Amazon historical pricing and Google AdSense performance metrics to provide placement recommendations for product offerings on blogs could exist entirely on fees earned by improving click-through or transaction rates. 

An API that mines local Yellow Pages listings to match business names entered by SMS users through their mobile phones to text a listing back to the handset could exist outside the traditional 411 service business, selling preferential results by ZIP Code, like an advertising company. It in turn could pass a small fee to the directory company that provided the data, all with measurement provided by Mashery so that all parties felt they were being fairly compensated.

Some APIs, however, will never earn a cent for good reason, they might just be useless. The ideas I've laid out could be bad ones, too, but the importance of measurement of the contribution made by each participant in the data value chain cannot be understated.

We're entering the Age of the meta, whether metadata or metarelationship, that depends on increasing contextualization of information. Most developers will add value to a product or service that they never physically touch or ever conclude a formal business relationship with the maker. Syndication of APIs will make this possible.