For the last decade, businesses have fretted over the increasing amount of data that is to be stored, analyzed, and made to generate sufficient value to justify the cost of its collection and storage. But we are nowhere near the data growth plateau. The Internet of Things in particular will keep us climbing the curve of data volume and velocity. In consequence, businesses of all sizes are cultivating monumentally large data lakes that are expected to generate value.
API-driven development is one strategy for effectively putting all that data to work. At its simplest, an API of the type we’re discussing here forms a layer between the data store and client applications. Client apps consume, handle, and present the data to users. Many different apps can use the same API, perhaps combining it with data from other APIs and performing application-specific analytics.
One of the main problems with large quantities of data is the need to expose it to external services so that it can be consumed by the business. The usefulness of data is typically not limited to a single business unit, which means it must be made available in an easily accessible form that can be consumed by business units with a range of different interests.
For example, customer sentiment data is valuable right across the business. Product designers can leverage data for insights into new features and enhancements. Marketing teams can use sentiment data to segment and target campaigns. Sales teams can shape their strategies and pitches based on conversations users are having. The same data, but different use cases necessitating different user experiences and presentations.
API-driven development has obvious advantages in cases like this. The developers of the API itself needn’t concern themselves too closely with the business logic of the applications: their function is to expose data in a way that makes it easy to discover and consume.
Each department that needs to make use of the data is then responsible for implementing the business logic in client applications. To continue with our examples from above, marketing teams might leverage the API combined with a mapping API and a demographics API to identify geographic clusters of positive sentiment in high-income areas, segments worth targeting during future marketing efforts.
APIs have the potential to go well beyond their status as a development methodology, as has been noted by Deloitte’s analysts:
“An organization’s core assets can be reused, shared and monetized through APIs that can extend the reach of existing services or provide new revenue streams. APIs should be managed like a product — one built on top of a potentially complex technical footprint that includes legacy and third-party systems and data.”
API-driven development is not without challenges. An API is only valuable if it meets the needs of business units, and if those business units are both prepared to use the APIs and aware that they exist. A siloed approach to API-driven development is likely to stall as developers build APIs that no one in the company uses or is even aware of.
That said, API-driven development is increasing in popularity as businesses seek to generate value by building APIs and applications that expose the data lakes and insights derived from analytics to business units thirsty for information.
About the Author
Karl Zimmerman is the founder and CEO of Steadfast, a leading IT Data Center Service company. Steadfast specializes in highly flexible cloud environments, robust dedicated and colocation hosting, and disaster recovery.