If you are one of the 77 million monthly Uber riders, it’s hard not to marvel at the convenience from seamlessly finding, ordering, tracking and paying for cab service, all from a few taps on your smartphone. Behind this front-end convenience, however, lies a lot of real-time coordination between Uber and multiple technology providers.
When you request a ride on your Uber iPhone app, Uber invokes Apple’s Core Location framework to understand your location and then Apple’s MapKit framework to show your real time location on the map. Once the driver has accepted the ride request, Uber uses Apple’s notifications to advise the details of drivers and the estimated time to arrival. While creating your Uber profile earlier, you would have used PayPal’s Card.Io API framework to scan your credit card, which the app uses to process your payment through Braintree API.
While building the Uber app, the developers at Uber did not go about individually agreeing with Apple, Braintree and PayPal on how information will be queried, retrieved and used. Instead, Uber effectively outsourced its requirements for real time data, functionality and services to their APIs. Short for “application program interface”, APIs refers to an interface that allows the data and functionality of one software to be used by another software through a set of agreed-upon standards. API is not a new concept and has been around since the beginning of computing. According to programmableWeb, today there are over 18,000 APIs on the web.
The majority of these are network APIs, i.e. a developer can access information from a 3rd party system remotely, using either a private or public network (e.g. internet). They are increasingly being monetized by the providers, either on a per transaction basis or on a fees basis, leading to what is called an API economy. In this blogpost, we understand the 3 drivers behind the rapid growth and disruptive nature of the API economy.
First, even the most profitable firms do not have unlimited resources to focus on all business opportunities all by themselves. Opening up APIs can drive revenue growth as it enables organizations to reach new users, channels and markets. Bala Iyer and Mohan Subramaniam share cross-industry examples of API enabled revenue generation in their HBR blog:
“Salesforce.com generates 50% of its revenue through APIs, Expedia.com generates 90%, and eBay, 60%. Salesforce.com has a marketplace (AppExchange) for apps created by its partners that work on its platform; they now number more than 300. Expedia’s APIs allow people using third-party websites to tap its functionality in order to book flights, cars, and hotels. And APIs allow eBay to list its auctions on other websites, get bidder information about sold items, collect feedback on transactions, and list new items for sale — all of which give additional exposure to eBay items and increase revenue.
APIs also help accelerate innovation, as developers can piggy-back on multiple quality services at a manageable variable cost (vs. investing time and effort in building everything from scratch) and rapidly build higher level functionality for their customers. Consider the case of Google Maps APIs. According to rough estimates, it would cost about $8B to build an equivalent map product, and additional millions of ongoing annual expenses like updating images, monitoring inappropriate content etc. For a fraction of that cost, developers can use Google’s Maps APIs on a pay-per-usage cost basis and take advantage of innovative features like autocomplete, finding location, real time traffic information and distance measurement. APIs have exponentially improved developer innovation and, according to TechCrunch, there are over 9 million developers working on private APIs and another 1.2 million on private APIs.
Finally, while developers benefit greatly in speed, scale, scope and variable cost from APIs, the innovation in developer ecosystem sometimes inspires internal improvement of the core product - Twitter benefitted from superior user interface developed by TweetDeck, a third-party developer using its API. New technologies like AI and MLR require lots of training from high quality datasets before commercial readiness, and opening up APIs has been the preferred approach for Google Predictions, IBM Watson, Amazon Machine learning and Microsoft Azure cognitive services.
‘Build or buy’ has always a key strategic consideration while creating new products and businesses. In the API economy, one should expand that choice to ‘build or buy or API’!