APIfication is a business decision

21.7.2017 -

Software developers have always shown an interest in dividing responsibilities into different areas and in packaging and distributing their tasks. To rush development, they like to utilise solutions created by others. How does the so-called “APIfication” tie into this and why is it worthwhile to invest in it?

Modern software development consists largely of piecing ready components and data sources together. Packaging tasks and responsibilities, so-called componentisation, adds the concept of application programming interface (API).

APIfication means offering functionality and data as a service for systems to use. An API hides the actual implementation and reveals only the simplified interactive model to the user.

In generalised terms, there are two models for APIfication: proactive and reactive. The proactively offered interface helps in identifying the target audience’s needs and in attracting new users for your own solutions or data. This means actively creating new demand. The reactive model, in turn, has external initiatives and needs. Other parties have already observed the value of the functionality or data for their own business and wish to utilise it.

In both cases, it all comes down to a business decision with the following underlying question: “Does offering the capability or data to others as a service yield more benefits than responsibility, maintenance costs, and work?”.

Individual developers make APIfication-related decisions at the code level daily to make their own work, or that of their colleagues, easier. Those who own the software or product will perceive the issue in slightly larger contexts and may incorporate a revenue generation model. The decision to invest in an open interface may be strategic, made by the executive group of a single organisation, or made by a grouping of organisations. It may also be a political mechanism to regulate society.

Benefits of APIfication

Why is APIfication beneficial, then? Here are some concrete advantages with brief reasons.

  1. Use self-service for integrations
    The aim is to reach a point where, for example, an online store wants to display order delivery status to the user, and then all the store developers need to do is find the correct API in the developer portal and implement it in their own solution. No other parties are needed in the implementation. This can be reached by creating a good data architecture and by requiring that the systems employ productised interfaces.
  2. Place importance on software developers
    We must require that systems offer humanly understandable, well-defined, and easily adoptable interfaces. The API producer must support other developers with good documentation, working code examples, a community of peers, and even by providing software development support. It goes without saying that error situations and recovering from them must be well-designed and automatable.Once a well-chosen team has authority over both service content and implementation-related issues, the motivation to create high-quality solutions increases. The interfaces are productised to make them better available for self-service purposes.
  3. Simplify and reduce the number of surprises
    The spectrum of technology is constantly growing, thus APIfication is becoming increasingly more essential in the creation of solutions that are as simple as possible and support business. Especially in relation to Web API management, today’s developers have a unique key with which usage can be tracked and various restrictions can be set; it is also possible to message with developers in advance and avoid surprises in change situations.
  4. Boost the innovation cycle
    Sometimes people can, waving an interface description around, rush into measuring interest towards an idea even before the decision to invest in the service itself has been made. When software functionality and data are introduced to a larger group in a controlled manner, the number of ideas increases. Among these one may find new, fresh combinations and perspectives of unforeseeable popularity.
  5. More protection for operative systems
    Especially with data query interfaces, spending the company’s operative systems’ computing capacity that is often licenced at a high cost makes little sense. Status changes for transactions or entities can also be published. Data services built on cloud services or open source technologies can receive these status changes and provide forwarding query interfaces that are independently and affordably scaled in relation to the number of queries.

Naturally, one must always consider the concrete benefits in the context of one’s own company, and create an API strategy to support the operations. What is essential to realise is that the architecture, speed of development, quality, and organisation all reflect one another.

By investing in a micro-service architecture, autonomous delivery teams, and user-friendly interfaces you can boost service development and scalability. Your implementations will benefit from the increased product-oriented business thinking.

Marko Pitkänen

Marko Pitkänen

I work at Solteq as a Head Architect. My speciality is system integrations. Currently I am tackling issues related to modernising the methods of system development and value production. In my spare time, I like to exercise outdoors and often take my family with me on competitive orienteering trips, among others.

Read all articles

Subscribe to our newsletter

We won't spam, promise!