The term Application Programming Interface (API) became attractive to organizations in the Web 2.0 age. Since then, API Economics has become the central pillar driving the digital and information revolution in various industries. With the help of APIs, enterprises can integrate internal and external resources and complement service scenarios with partners to develop new businesses and meet market demands. As a result, product iteration is expedited, solving a series of scenario-based problems to improve the user experience. It enables disruption with minimal investment in resources.
An API is a collection of definitions, rules, and guidelines for building application software. It is a set of predefined functions to serve as channels for communications and data exchanges between applications. Enterprises only need to provide standard API services, as developers and partners reorganize and integrate them with their applications, web pages, and smart devices to create new services. A good API enables developers to develop computer programs by providing all the building blocks under one roof.
API Economics refers to economic activities generated from partnerships established between enterprises through APIs. By exposing data, business processes, and other services and resources through APIs, organizations are creating new and innovative business platforms. API Economics is no longer a simple concept, with many enterprises applying it in their commercial activities. It is not only about generating value through API but also about locking in partners to innovate business models.
With the emergence of cloud computing, mobile internet, big data and machine learning, a major revolution in business models is underway. To continue strengthening core competitiveness and expanding customer base, enterprises need to break the closed system and work with partners to support more business scenarios and to meet the increasingly fragmented user demands. It helps in providing mobile, web, and other client interfaces as a flexible layer of channels on top of API.
Cross-platform and stateless APIs can serve as channels connecting systems. As a result, the service capability of business systems is opened up to a limited degree. The upstream and downstream of enterprises can be linked to establish close ties with partners and form an economic community for expanding business scenarios.
Many traditional enterprises use APIs to open up their services. For example, Interbank Funds Transfer, the third-party ticket query services, and the traffic inquiries of transportation bureaus are all achieved through APIs.
Organizations should open up data and service APIs to a limited degree. As a result, they can exploit new opportunities without changing existing models. However, the battle of 'API-ization' is challenging. Not all companies can stay on track during the 'API-ization' process, with obstacles to development a constant threat. Therefore, organizations need to plan meticulously and deploy uniform and standard management techniques to develop API Economics in an orderly fashion.
Businesses should keep the following things in mind while creating API economies.
●Follow the best practices for development throughout the API life-cycle
●Allow easy access for developers to find, research and test your APIs
●Stimulate adoption by indexing your APIs
●Enable users to search for APIs using custom classifications or through advanced search capabilities
●Set alerts or notifications whenever your APIs are impacted
●Monitor API performance, and track API traffic to form metrics for monetization
First, you should consider the system situation of the enterprise, analyze the market and plan the API scheme. In this phase, you should study the commercial objectives and feasibility of opening up APIs, the opening-up approaches, and revenue models.
There are three forms of APIs in terms of usage scenarios. You can adopt them in combinations as well as in individual practical applications.
1. Private API: APIs used inside the enterprise. Private APIs link internal systems, mobile phone apps, and smart devices using standard internal interfaces. The advantage lies in the fact that uniform interfaces can be used to achieve interaction among all systems, reduce inter-system coupling, and make the system architecture more flexible.
2. Partner API: Enable businesses to connect with partners and achieve in-depth cooperation. They facilitate software communications and integration as well as coordinated business development for both sides.
3. Public API: Enterprises publish their data and functions for free or in a fee-based manner so others can develop new products through the APIs, promoting business innovation and influencing expansion.
An API contains three subjects, the API provider, the API consumer, and the end user. API providers deliver services to API consumers who, by integrating more API innovated services, provide more functions to serve the end users in a deeper and broader way.
●API provider: enterprises or individuals opening up the APIs
●API consumer: enterprises or developers who use the APIs opened up by the API provider to create new products or services
●End user: users who use the products or services provided by the API consumer
In actual API operation, different API types and various phases of the enterprise require various modes of operation. Below are the four current API business models:
1.API as a product
Enterprises encapsulate their business capacities into APIs, and consumers pay for their usage of the API by the number of calls or the frequency of calls. As a result, the business capacity gets monetized. APIs in the form of tools or services mostly adopt this model, such as weather services, image and face recognition, and text message services.
2.API as an extension of products or services
In this business model, APIs are called for free. API providers integrate products of their consumers and partners to provide services in a deeper and broader manner to enhance product competitiveness. E-businesses and resource-type APIs usually adopt this business model, such as Salesforce.com, and Workday.
3.API as a way to develop partners and link up enterprise upstream and downstream to promote products. API providers that adopt this model usually open up APIs to promote their products to partners or third parties for integration.
4.API as a means to provide value-added services to partners
Developers enjoy value-added services and enhance products through the APIs. The revenue is mainly from advertisements, such as YouTube, Facebook, and Foursquare.
API business models differ a lot from each other depending on the different scenarios. API providers establish ecosystems through APIs and enterprises should select an appropriate business model based on actual situations.
You need to identify the API revenue sources when opening up APIs according to the set business targets.
●Free model: APIs are offered for free calls. Enterprises adopting this model fall into the following two situations. Enterprises provide APIs to developers/partners for free to achieve diversified development of products, expand influence and provide in-depth services to end users.
APIs are configured into multiple charging dimensions to attract development. Basic functions are free, while advanced ones charge fees and free users are encouraged to become paid users. Charged by the number of API calls or frequency of API calls. This is tested in public APIs. Developers can buy a call number quota or call frequency quota. For example, if you pay $100, you can call the API for 1,000 times, and the allowed call frequency is 1 call/minute; if you pay $1,000, you can call the API for 10,000 times, and the allowed call frequency is one call/second.
●Profit sharing model: APIs are made available to a limited degree to partners. Partners are encouraged to sell products through APIs, and partners divide product profits in fixed ratios. This model is suitable for resource-oriented enterprises, such as developing distribution and agency models.
To make your system service-oriented and 'API-ized', you need to manage a large number of APIs. API providers and consumers need to incur a significant management and communication cost, which may constitute a bottleneck for the enterprise to develop API Economics. Here's a few tips:
●Achieve full lifecycle management of APIs - Complete the entire process from API development, testing to API launch and removal, and realize monitoring on API running conditions with timely warnings in case of exceptions.
●Prepare a sound API service discovery mechanism - an approach for consumers to discover APIs should be provided, so the consumers can timely find newly added or updated APIs.
●Provide API documentation and SDKs - API instructions should be made available to developers, including request and response descriptions. Multilingual SDKs should be made available for various consumer scenarios so that users can call the API with simple operations. Also, the SDK and API documentation should be automatically generated to ensure users get SDKs and API documentation in a timely and accurate manner, reducing API development and iteration costs.
While bringing convenience, APIs also introduce risks. You need to impose corresponding user authentication, permission control, and traffic restriction means among others to secure the services and prevent data leakage. Also, you need to implement the protection mechanism against attacks, such as traffic cleaning, traffic identification, and anti-DDoS IP addresses to ensure stable services. The security against anti-attack protection is indispensable no matter whether there actually are attacks or not.
Identity authentication is required to secure requests and protect APIs from malicious calls and for APIs with high-security requirements. This mechanism can protect requests from being tampered with during network transmission, make requests time-efficient and prohibit requests from being repeatedly used. Every request should carry identity information. API identity authentication may come in many forms, such as HTTP Basic, API KEY/APP KEY, Oauth, openConnectid, and JWK.
Here are some qualities APIs should have.
●Well-designed: good APIs are always well-designed. Designers should utilize logical abstract ideas and consider the user scenario to design a set of fully-functional, easy-to-use and extensible APIs.
●Scenario-oriented: the usage scenario of users should be studied to create APIs aligned to the users' interests.
●Cross-platform: the designed APIs should support different terminals such as Android, iOS, and WebService.
●High cohesion and low coupling: an API should provide complete function, and the service correlation among various interfaces should be minimized. An API shall not change with the change in another API, or exist with some other interfaces as the premise.
●Scalable: postponing functional implementation is the advantage of being scalable. New functions can be added later as needed. If a function is not required, it can be excluded.
●Data size control: an API should not return too much data. A large data size not only complicates the processing but also imposes high stress on data transmission, leading to slow responses from the client. The large data size can be attributed to the unclear division of interfaces.
●Uniform format: all APIs should follow a consistent parameter format and style. For example, all the API parameters should be either separated by commas, or in arrays; or all the date parameters should be either in YYYY-MM-DD format or the YYYY-MM-DD format.
●Permission control: a complete system should be in place to ensure data security and manage API access permissions and data permissions.
●Request encryption: important APIs should adopt encrypted transmission, such as the popular HTTPS. However, HTTPS may compromise the performance a little, so that you can take alternative measures in line with actual conditions as necessary.
An API service should be stable to assure users and promote regular use. The stability of API services is vital. But the reality is that you will keep being nagged by various problems. You need to monitor the API operating conditions to always improve the user experience. Monitoring can provide information to directly observe the API running conditions and users' habits, providing a basis for API Operations & Maintenance (O&M) and business development. Here is a rundown of what you should be monitoring:
●Number of calls: by grasping and analyzing the number of calls to APIs, you can understand the consumers' preference (which APIs are more user-friendly and which are not), and service loads so as to allocate resources appropriately. You can also learn if your APIs are properly designed, such as whether the user needs to call the same API multiple times during a single business operation.
●Error rate: error rate is an important indicator appraising API performance. Causes for high error rates include:
○Substandard back-end services, thus multiple errors are reported
○Over-complicated API design that is hard to understand
Error rate is an important indicator measuring the system health. Different countermeasures should be taken to address various errors.
●Error distribution: a good monitoring system should display the error details for figuring out the error cause, such as:
○If an incoming parameter has a high error rate, the API design should be adjusted to optimize complicated incoming parameters and avoid comprehension deviations.
○If an API is called by too many users, pushing the traffic beyond the limited quota, the service capability should be expanded, and the traffic flow allowance should be adjusted.
○If a service suffers severe timeout issues, the code efficiency should be checked to identify whether back-end service capability should be increased as appropriate.
●Warning: a warning and notification mechanism should be in place for important indicators, such as a sudden increase in traffic flow, a soaring error rate, and slow service responses. The issue should be handled immediately to avoid faults.
●Principles:
-Only set warnings for key indicators. The principle of "the more, the better" doesn't apply.
-Carefully analyze the warning threshold values. The principle of "the lower, the better" doesn't apply.
-Use concise titles and descriptions for warnings. The principle of "the more detailed, the better" doesn't apply.
-Warning text messages and e-mails should be used in combination.
●Traffic control: when you open up APIs, the request frequency should be monitored in various dimensions such as newly added APIs and users according to the service capability and business scenarios, to ensure service availability and prevent malicious calls.
○API traffic control: the businesses are layered based on the terminal service capability and business importance to impose differentiated traffic control over different APIs and ensure continuity of important activities.
○User traffic control: users on the API are segregated to avoid congestion and competition for resources.
API Economics refers not only to the opening up of service capability for others but also to the adoption of mature and stable third-party APIs to enrich system functions. This not only facilitates lightweight system code and accelerates development progress, but also gives developers more time to handle issues in their areas of expertise, which is more valuable than repeatedly building mature functions that are already available from others, such as payment, weather, and image recognition services. Other companies have provided standardized services, so it is not worth the effort to redevelop them again.
API Economics puts forward a new business model to help enterprises quickly respond to market demands at a low cost, establish an enterprise ecosystem, and spur integration of capabilities across industrial chains for new economic forms. The degree of API openness will become a core indicator weighing the enterprises' competitiveness. It is up to organizations to leverage the huge opportunity of API economics, and the potential is for all to see.
Three industries you didn't realize are benefitting from big data
2,605 posts | 747 followers
FollowAlibaba Cloud Community - December 28, 2021
Alibaba Cloud Community - February 4, 2022
Alibaba Cloud Community - March 9, 2023
Alibaba Clouder - September 28, 2018
Alibaba Cloud Community - February 3, 2022
Alipay Technology - November 6, 2019
2,605 posts | 747 followers
Follow
Metaverse Solution
Metaverse is the next generation of the Internet.
Learn More
VPC
A virtual private cloud service that provides an isolated cloud network to operate resources in a secure environment.
Learn More
OpenAPI Explorer
OpenAPI Explorer allows you to call an API through its web interface or WebCLI, and view the entire process.
Learn More
API Gateway
API Gateway provides you with high-performance and high-availability API hosting services to deploy and release your APIs on Alibaba Cloud products.
Learn MoreMore Posts by Alibaba Clouder
