Current Issue
This Month's Print Issue

Follow Fast Company

We’ll come to you.

Created for and commissioned by IBM
Created for and commissioned by IBM
How to Win in the API Economy

Ten years ago, any enterprise large or small that didn't have a website was not seen as a legitimate venture. Today, the same holds true for companies that fail to take advantage of application programming interfaces, or APIs—those often simple yet powerful bits of code that allow digital applications and services to communicate with one another. Leading companies in every industry are reimagining their business in code, and they are managing APIs as products and core business assets rather than IT projects.

Web and mobile users, of course, have grown reliant on APIs, interacting with them countless times every day. (Without APIs, the seamless way that Google Maps is integrated into your Yelp search results, for instance, would be unimaginable.) Here, Kyle Charlet, a Distinguished Engineer and z Systems cloud architect working out of IBM's Silicon Valley Labs, and Jason Gartner, IBM CTO of API Economy, discuss how today's API economy enables innovation at light-speed—while helping businesses transition from digital to cognitive ecosystems.

Like so many phrases associated with digital technology, "the API economy" seems to mean different things to different people. What does the phrase mean from your perspective?

Kyle Charlet: APIs allow apps and systems to talk to one another. There are consumer-facing APIs, which anyone who uses a mobile device is familiar with — whether or not that consumer knows what an API is — and there are what we might call inner APIs that an enterprise can use to call other applications within its own ecosystem. In an elemental sense, the API economy enables businesses to monetize their APIs and their services.

More than just money is exchanged through APIs, though. Right?

Jason Gartner: That's right. Something that makes the API economy unique is that various types of currency move between API producers and consumers, and not all of them involve a straight-up dollar exchange. There are times when a website will pay a provider x cents every time that site calls the provider's API. But other types of currency are exchanged in this space, too. For example, the Facebook login that authenticates users on so many other websites is an API, and the consumer of that API is any website that provides that particular login option. The currency exchanged is free advertising for Facebook, which also increases the value of its platform because millions and millions of users are now logging in from thousands of other sites every single day. No dollars trade hands, but there's definitely a value exchange taking place.

Jason, is the API economy helping to power the next big enterprise transition, from a digital model to a cognitive model?

Jason Gartner: Well, certainly the number of APIs that are being called by machine processes, as opposed to being called by a person on the other end of a desktop or a mobile device, is exploding. This means that a huge amount of cognitive activity is taking place between businesses without any direct human intervention.

Think about a loan process between a car company and a bank. In most cases, there's no longer a person involved in that interaction at all. The APIs are looking at insurance information, the applicant's driving record, the last three places he lived, social media activity, and all sorts of other data—not in sequence, but simultaneously. These are not simple binary, yes-or-no processes: They're cognitive. When they run on IBM's next-generation z mainframes with faster computation, stronger cryptography, and terabytes of memory to reduce latency, these transactions are quicker and more secure than ever. And it's all being driven by APIs.

With open source driving so much innovation, and with so many APIs coming to market every day, how do businesses know which APIs are for real and which will likely be forgotten six months from now?

Kyle Charlet: The API economy is about the APIs themselves, but it's also about the ability to quickly discover APIs that fit your business. Open standards make that pretty easy, and that's great. Of course, not all APIs are created equal. Those that are used the most are the most reliable. If you call three APIs to do the exact same thing, and two of them are down 10% or 20% of the time, and one of them works all the time, you're going to stick with the reliable one.

That's where a good chunk of the z Systems’ value proposition comes into play, with the reliability of the platform. With z machines' larger cache and software functions like z/OS Connect—which powers connectivity between the mobile environment and the mainframe's backend—you can drive APIs at scale, and they're not going to break down.

I want to ask you about the emergence of "microservices"—smaller, API–driven capabilities that are more easily managed than a single, monolithic application running business requirements. Why are microservices so hot right now?

Jason Gartner: It's important to realize that there is business logic written in code behind almost any API. Who manages that code and that API? Who scales it? Who upgrades it? For the longest time, one team—the team that built the monolithic application that you mentioned—was responsible for managing all the business requirements for that application. Today's business environment, on the other hand, requires that separate, loosely coupled teams develop and deploy products at a far faster rate than ever before. Microservices represent a more complex architecture than the big, monolithic model we used to embrace—but, paradoxically, that architecture is also more reliable than the old model.

For example, developers can tap into the assets and business processes that already exist in their mainframe environment, whether or not they have any mainframe skills, by incorporating microservices and APIs. Seamless application development, cloud integration, open tools—all of these can help power automated, adaptable, and, above all, agile capabilities.

No matter how fast they're built, enterprise-grade APIs still have to answer to multiple stakeholders—developers, IT admins, even marketers and salespeople. Are we nearing a point where development and deployment times for APIs grow so short that the reliability you're talking about is in danger of being compromised?

Jason Gartner: I don't think so. With StrongLoop, which IBM acquired late last year, we're working on our next-generation API capability. This will combine the four major activities that those stakeholders care about. In short, the new capability will create, run, and manage APIs, while also enforcing the standards behind those interfaces, whether they're built on-premise or on a hybrid cloud development platform, like IBM's Bluemix. This is a huge step forward, and by streamlining the API process without shortchanging security, our StrongLoop partnership will only strengthen our role as a central force in the API economy.

It’s worth remembering that up to 80% of the world's enterprise data and transactions already reside on or pass through z Systems mainframes. For businesses looking to stay a step ahead of their competition, tapping into that sort of information treasure trove—with industry-leading, next-generation capabilities—is the best way to win in the API economy.

For a more in-depth exploration of strategies to succeed in the API economy, click here.

Other Related Topics:

This article was created and commissioned by IBM, and the views expressed are their own.