Overview – A best practice approach to the API lifecycle accelerates digital transformation
As we all know, APIs are key to digital business. When developed to expose IT and business functions, we can use APIs to:
- unlock value from data
- launch innovative products
- enter new digital ecosystems
- comply with regulation
- build successful partnerships
That’s why large enterprises are focusing heavily on how to design and develop APIs with consistency and scale. Of course, it’s more than APIs in reality (think Services and Events etc.), but we’re using APIs as an umbrella term in this post . There’s also a big push to reuse, extend, and/or modernize (where appropriate) existing APIs. A robust portfolio and mature API strategy are key enablers to accelerating digital transformation.
In a previous post we answered the question “what is API lifecycle management?”, but today we’ll be focusing in on the following aspects of the lifecycle:
- Tactics to utilize so that your API lifecycle ensures scale, speed, and consistency
- The roles involved in managing (and partaking in) the lifecycle
- The 4 key API lifecycle phases (plan, design, build, run) – and why we should all be focusing especially on what’s happening upstream of CI/CD
- Benefits of this approach
Tactics for a best-of-breed API lifecycle
Most organizations have spent time and resource standing up infrastructure that supports part of the API lifecycle, including API gateways, API management platforms, and other distributed environments. However, many are noticing a gap in the lifecycle and supporting architecture (that which is upstream of deploying the API to CI/CD and run).
That gap is most apparent when we have development teams manually designing and building APIs, that are code-specific to one architectural standard, runtime system, or application framework. With this approach, time to market for new APIs is slow and cost inefficient. We also see a lack of consistency and alignment to governance rules and standards, plus lots of API duplication and a lack of reuse.
Digital leaders are using the following tactics for a differentiating approach to the API lifecycle:
- Aligning APIs to business needs and enabling business roles to collaborate across the lifecycle: this recent Forrester report states that an IT-led API business strategy is in fact the opposite of an API business strategy
- Automation of manual and error-prone tasks: e.g. governance checks and validations, and artifact generation based-off the API specification
- Organizing APIs (and tracking them through their evolution) in a holistic API catalog: where they are discoverable to the entire organization
- Integrating the API lifecycle with upstream and downstream platforms: to create an end-to-end framework and best of breed API architecture
A focus on the planning, designing, and building of APIs is key to leveraging these tactics for successful API programs (more on these API lifecycle stages later).
Who manages the API lifecycle?
We commonly think of the API lifecycle being managed and executed exclusively by IT, with API developers taking the limelight. But, for a killer API strategy that helps deliver the thousands of high quality APIs needed to encapsulate business and IT functions, our API lifecycles should enable both business and IT roles to collaborate and participate.
That way, you’re helping to align APIs to business goals, and ensuring that your API designs are consumer-centric. Check out this other post for the 11 roles other than developers that want to use APIs, and how to enable them.
Let’s now take a look at the 4 main phases of the API lifecycle when we have these tactics and roles involved.
API lifecycle phases – Plan, Design, Build, Run
The API Plan stage ensures we can quickly deliver new capabilities without creating duplication and poorly designed artifacts. Consistent APIs cost less time to implement, and they decrease the workload of developers, while increasing the quality of the applications using them.
Business stakeholders, Product Owners, and Technical Analysts are able to ideate on fulfilling new digital business processes using existing and reliable APIs. Collaboration and prioritization with development is easier and more accurate with end-to-end views of the lifecycle and portfolio. And customer journey maps can be planned and populated with reusable or target state API candidates via business friendly views of our API catalog. Quickly discover and reuse existing API Specs or register to design new and consistent APIs.
- Initial Service definitions
- Prioritization of new or updated APIs
- Analysis on reducing technical debt (including consolidation)
- Reporting on expected API Product KPIs vs. actual (based on plan and design integrated with consumption)
The API design stage is when API architects, developers, or technical analysts define the functions of the API and how it will be represented in its technical form.
The design stage ensures risk management and governance is followed through the different stages of the lifecycle with a view of API specification versions, approvals, and usage.
Designs also ensure we can avoid new and unnecessary API construction/duplication.
- A set of API standards will be defined as governance rules that can be applied to the technical design of the API. This will provide feedback to the designer about how well their design meets the governance rules set by Wells Fargo.
- Specification(s) framework which acts as a representation of APIs, Services, and Events.
In the API build stage, developers build their payload for the API. We can use built in templates and guides to automate the alignment to governance, standards, frameworks, security, and deployments. This automates much of the build process, reduces development friction, and ensures all artifacts are consistent and governed.
When building new APIs, developers should implement and define the technical details (payloads, headers/parameters, NFRs, mappings, response codes, etc.). For existing APIs that are already deployed, developers modify technical details or extend new version(s) as needed. Developers can also apply defaults set by the organization as a way to expedite the API build process and reduce any guess-work around technical requirements their API may require.
- Reliable and consistent industry standard specifications to develop against
- Visibility of API deployment status through CI/CD pipeline (including any failure notifications)
- Visibility of what is in the different Run and application environments
- Visibility of what is in or out of governance throughout lifecycle stages and environments
For the API run stage, our APIs are deployed to gateways with roundtrip details and history of APIs across the entire lifecycle. And with bi-directions integrations, when an update is made in runtime it can be pushed back into our catalog (and vice-versa).
The API gateway provides a single point of entry for all APIs and handles common tasks that are used across a system of API services. Common tasks such as authentication, routing, rate limiting, monitoring, analytics, policies, alerts and security.
- API in runtime environment
- System of record for the API in same place as design-time artifacts
Once APIs are deployed, we can monitor them using runtime API management platforms. Discovery and subscription of available APIs can be performed in a variety of API solutions. For more on that, check out this blog on API catalogs vs developer portals vs holistic catalogs to help you choose which is right for your use cases.
Benefits to this API lifecycle approach
At the highest level, this approach to the API lifecycle ensures delivers on the promises of API-first. Enabling bridging the gap between legacy applications and new cloud based apps, rapidly adapting applications to support new digital business processes, and reaching economies of scale that seamlessly and economically support additional users, data flows, and transactions.
We also reap these further benefits:
- APIs as strategic business assets instead of one-off integrations
- Accelerating time to market and reducing costs of developing new APIs
- Collaboration of resources through all lifecycle phases
- including reducing developer friction and freeing them up for big brain thinking
- Standardized governance, frameworks, patterns, and validation to ensure API consistency, compliance, and completeness
- Visibility, discovery, and recombination of APIs to support digital initiatives
For a deeper dive on this topic, we recommend reading through our API lifecycle whitepaper – you can download it for free here.