As third-party applications built on APIs become more important distribution channels, companies have good reason to look closely at their API usage limits. Case in point: Yelp last week announced that developers can now issue 25,000 requests per day to its API. Previously, the company limited requests to just 100 per day.
As TechCrunch's Josh Constine observed, "By freeing up usage of its API, Yelp becomes more appealing to developers looking to help people discover local businesses, and could persuade them to use its database instead of Foursquare or Google Places which were much more openly available until now."
Foursquare allows 5,000 unauthenticated requests per hour to its venue information endpoint and up to 500 requests per hour per OAuth token for authenticated requests. Google Places offers up to 100,000 requests per day to developers who verify their identity. While these are more generous than Yelp's new limits, given that Yelp is more popular than Foursquare and Google Places as far as online reviews go, some developers may find Yelp's API to be an attractive option despite the fact that it still allows for fewer requests.
Usage Limits: An Important Part of API Strategy and Management
Yelp isn't the only API provider that restricts API access based on usage. In fact, most of the major providers of public APIs impose usage limits to prevent abuse and ensure quality of service. In some cases, API providers even adjust their rate limits in real time based on anticipated demand. For instance, Twitter implemented dynamic rate limiting in 2010 in preparation for the World Cup that year.
In addition to the technical reasons API providers implement rate limits, there are also often business justifications for imposing them. Google, for instance, monetizes many of its APIs, including its popular Google Maps API. While Google doesn't automatically force developers to pay if they exceed the free limits, establishing clarity around usage levels that are free and usage levels that justify payment is arguably a good thing for both API provider and developer.
As Yelp demonstrates, however, usage limits are not a set-it-and-forget-it phenomenon. Nor should they be. As a company's strategy and position in the marketplace evolve, it makes sense to re-evaluate API usage limits and adjust them as necessary. Of course, increasing usage limits or decreasing the cost of paid access is easier than imposing new restrictions that irk developers and produce bait-and-switch accusations.
Twitter, for instance, has created a stir several times by implementing API policy changes that involved more restrictive usage limits. And recently, despite the fact that its new policy will affect less than 1% of the developers using its API, Foursquare found itself dealing with negative buzz when it announced that it would begin charging higher-volume API users.
The key takeaway for API providers: Pay attention to usage limits, and be thoughtful about them. There are enough instances of developers being burned by API providers to give them pause, so stability and clarity around usage limits is an important part of building and maintaining developer trust. And with the number of APIs growing, implementing sensible usage limits is crucial to attracting and retaining developers looking for the best opportunities.