With innovation, companies are playing with more pieces than they may realize.

Until recently, a business that wanted to spur growth into new categories had three primary options. The first is to leverage an existing core competency in a new market. Another is to be the first to out maneuver competition by anticipating a marketplace change. The third is to create new offerings to meet the unmet needs of existing customers.

The explosion of API offerings, tools, and formats hint at a new, fourth opportunity for business innovation. Rapid advances over the past several decades - the shrinkage and commodification of computers, the introduction of business software, and the rise of worldwide networks - have created a bounty of new, digital assets. Companies have found themselves increasingly saturated with data. Sharing that data creates value. Increasingly, the primary way to share data is via APIs.

Amazon remains one of the largest and most profound API innovation stories. According to legend, one day in 2002 Jeff Bezos, the leader of the online retailer, issued a company wide mandate stating that: * All teams would expose data and functionality through service interfaces. * All teams had to use those interfa ces to communicate with each other. * All interfaces, without exception, must be designed to be externalizable. * Anyone who didn't implement the above would be fired.

What started as an edict has since blossomed into a multi-billion dollar business, posed to someday exceed their traditional e-commerace business.

Amazon took their efficient and elastic scaling of computational resources, a core competency honed through ebbs and flows of seasonal shopping, and codified them. Once such useful building blocks were exposed, anyone who could consume an API could benefit from Amazon's cloud. This freed up additional time and resources for these companies to focus on their key value propositions.

Salesforce faced a similar dilemma. A decade ago they dominated the nascent online customer relationship management (CRM) space. But they knew that wasn't enough. Remaining only a website opened the possibility of Salesforce being disrupted by a more nimble startup; a situation where they would have the dubious distinction of being the "legacy" solution.

Rather than wait, Salesforce made the innovative business decision to expose key business processes through APIs. Rather than compete with the next big sales idea, they offered developers a platform on which to build and incorporate their own. Developers could jumpstart their application by being placed in front of the hundreds of thousands of customers that Salesforce already head. Salesforce gained the advantage of not needing to code everything themselves, allowing a community of developers try out ideas. In 2006, 40% of the requests to Salesforce's web servers were to their API. By 2013 that number had grown to 60%, across 2.5 million applications and 800,000 developers.

It isn't enough to start an API program and call oneself a platform, however. An API needs to be a continuation of a business's larger strategy. eBay was one of first large online ecommerace companies that offered an API. However, much to many's surprises, it did not offer the ability to search listings. At the time, eBay was most concerned with increasing the volume of sales on its webpage and, as a result, it's API only allowed it's users to post new items. It was only after it had established a reputation for having everything did it then add the ability to read from it's data.

Creating an API program doesn't just give developers the opportunity to kickstart their applications with a piece of proven functionality. The provider of the API wins by creating (or expanding) an ecosystem, increasing retention, and raising brand awareness. Some companies have gone so far as to becoming API-only, where the developers are the customer. Payments processing company Stripe and communications provider Twilio both are meant to be consumed primarily as a service - the API is the business.

Before the first line of API code is written, companies must first analyze their digital assets. This data can generated from a wide array of physical items (commonly now referred to as the Internet of Things), each producing a stream of performance information that is useful for refining processes. It can be the culling and refinement of data from across previously isolated data silos to create new forms (the Big Data movement). It may be the codification of a core competency into a service, as was Amazon's case. Data may even come from new relationships; recognizing the value of their data to create actionable insights, companies may trade with interested parties who are able to reciprocate.

Next, companies need to establish the vision of the ecosystem they wish to be built and the role of their platform within it. Does the company wish to enter a new market? Is the purpose of the API to extend their company reach? Understanding the larger company goals and how a platform, exposing digital assets via APIs, supports them is essential.

Most importantly, businesses need to begin with the realization that an API is a strategic and architectural choice, not a technological one. The work in creating a successful API initiative is independent of the technology stack within an organization. It is only after taking inventory of the core functionality or digital assets to be exposed and the platform that a company would want to see that technology considerations should be considered.