Openness drives innovation. The world over, governments, institutions and corporate are opening up access to their data and core systems to allow for the creation of new services that will add value to their consumer base. Innovators on the other side are looking at what is being made available and creating mashups by combining data sets or simply relooking the relationships between data from one particular system and adding for example the element of location to bring about a different perspective.
Access to data is given through a set of protocols and you may hear the term API thrown about within developer communities or by your in-house IT team. API stands for Application Programming Interface, and simply put, it is a specification for data interaction between two systems allowing them to talk to each other.
Data is considered sacred as it forms the backbone of any business and therefore you must consider some baseline best practice if you are looking to make an API available, whether the reason be to monetize it or simply provide an avenue for innovation.
Data protection ranks highest in my books when looking at opening up data. No one wants sensitive customer data laid bare as a result of a poorly thought out data protection strategy. The starting point of this is to know what elements of your data are to be considered sensitive, what are the international standards in regards to format and finally looking at encryption, threat detection and masking. In some instances, for example government open data, care has to be taken when making available data that has been stripped off of personal unique identifiers, as one may layer a different dataset and suddenly the person behind the data is revealed.
Ease of use and management comes second as it will determine the uptake and evangelization of your API among developers. Self-service tools that allow for quick deployment and testing of your API will go a long way in enhancing adoption and innovation and reducing simple support issues that many arise in the absence of this.
Scale and availability are critical if you plan to monetize your service. Whether a service using your API is considered “mission critical ” or not, you must plan from a capacity and availability perspective to ensure where possible 99.9% uptime and seamless scale should traffic spikes occur. This planning will also inform the packages that you make available to that will guide provisioning of various accounts to mitigate backend overload or better manage “rouge” or “inefficient ” users.
Finally you must plug in operational analytics to help understand usage, service patterns, trends to improve service quality and customer satisfaction.
API means business, here’s to hoping you do too.