3 Best Practices for Enterprise APIs

By Chris Marsh, Principal Analyst, Enterprise Mobility, Yankee Group — March 24, 2014

The confluence of mobile devices, distributed cloud services and SaaS applications is changing the very rules of the game for enterprise IT and requires the rethinking of  what products and services are offered to customers, users and partners.
 
More mobile apps connecting to more data sources need more enabling (and flexible) APIs. And just as much as a mobile strategy needs to be cross-disciplinary, so does any thinking around APIs. Yet, few companies are changing fast enough to keep pace with what are significant implications for the architecture, deployment and management of applications.  
 
So what does this confluence create?

  • It ages the typical understanding of an enterprise application itself with modern mobile applications needing to connect to multiple rather than single or a few data sources.
  • It makes redundant the traditional three-tier enterprise application architecture.
  • It requires a new way to build and manage APIs.
  • It degrades the value of SOA infrastructures and traditional web APIs used to facilitate transactions between these applications and core backend systems.
  • In particular demand is growing for a way to bring scale to the build and management of APIs for the long-tail of internal and customer workflows companies will look to mobilize beyond those most strategic of business processes.

API Strategy
In this time of growing technology flux and increasingly distributed data, enterprises need an API strategy. With few companies having a properly thought out mobile strategy, even fewer have a stance around how and when to use APIs to provide access to the mobile applications they might need to consider enabling.

The considerations behind building out an enabling API library are as much organizational and even product focused as they are technical, and are a cross-disciplinary responsibility for the following reasons.

Mobility’s immediacy raises user expectations. The immediacy of the interface between a mobile user and their applications has rapidly enhanced expectations about access to data and the richness of the experience delivered by integrated data streams. APIs need to become much more granular to provide only the data needed for experiences that are contextualized to the user’s device, OS and application requirements.
 
Enterprise data has value only if more connected. Data locked in siloed enterprise applications and rigid SOA-based transactions is degrading in its value to the enterprise. Increasingly the connectedness between different data sets is where its value will lie.
 
Technology will bring down organizational silos. Technology is the chicken and organizational change is the egg. More flexible and scalable ways of building, deploying and managing APIs to deliver richer and more connected experiences will help bring down organizational siloes which have been compounded by siloed enterprise applications. Product managers, developers, security and risk professionals all have a common stake in API evolution.
 
“Data as inventory” needs new measurement. The continuous immediacy of the mobile experience freeing data from the internal systems housing it in fact turns it into inventory. This requires enterprise IT to be measured in a new way—from evaluating the costs of internal systems to determining how continuous returns can be generated by recycling this “data inventory” to iterate new products and offer more compelling mobile services.
 
All the above creates business model disruption. Users will demand rich multi-API enabled experiences, API publishers will exploit new revenue streams by monetizing high value data streams, platform vendors will provide a new middleware of flexible backend services and data orchestration.

Enterprise API best practices:
  • API-fy internally as well as for customer experiences. An API is a way to access data for both internal stakeholders as well as for customers. For example a retail store opening its data to sales, marketing and product teams each of which use it to enhance the overall value of the data.
  • Consider where internal data might add external value. Not for all companies will exposing data externally be a good idea, but for some it might open up new revenue streams.
  • Think of yourself as a node not a hub. As systems and data become more distributed modern APIs become not about the unwieldy dependencies of application-to-data source silos, but require companies to think about themselves more like a nodes, facilitating user access to data while not forcing them to journey to you to get that data.

POST A COMMENT

comments powered by Disqus

RATE THIS CONTENT (5 Being the Best)

12345
Current rating: 5 (2 ratings)

MOST READ STORIES

topics

Must See


FEATURED REPORT

Mobile Risk: Security Is Not a Game

IDC predicts 2 billion mobile devices will be shipped by 2017, while Gartner expects a 26 billion Internet of Things installed base (excluding smartphones and tablets) by 2020. With more devices, more machines, more connectivity comes more risk.