MACH Aligned for Retail: API-First

Retailers must constantly evolve to meet growing customer expectations and remain competitive. Both their internal- and external-facing applications must be developed using principles that promote agility and innovation, moving away from siloed architectures.

As discussed in the first article of this series, the MACH Alliance promotes the development of modern applications through open tech ecosystems. MACH is an acronym that represents Microservices, API-first, Cloud-native SaaS, and Headless. MongoDB is a proud member of the Alliance, providing retailers with the tools to build highly flexible and scalable applications.

This is the second in a series of blog posts focused on MACH and how retail organizations can leverage this framework to gain a competitive advantage. In this article, we’ll discuss concepts relating to the second letter of MACH: API-first.

What is an API-first approach and why is it important?

An application programming interface (API) is a set of routines, protocols, and tools that allow applications, or services within a microservices architecture, to talk to each other. APIs can be seen as messengers that deliver requests and responses.

Applications built around APIs are said to be API-first. With this approach, the design and development of APIs come before the software implementation. Typically, an interface is created that is used to host and develop the API. The development team will then leverage the interface to build the rest of the application.

This methodology enables developers to have access to specific functionalities of external applications or other microservices within the same application, depending on their needs. It promotes reusability because functionalities are interoperable with mobile and other client applications.

In addition, applications developed with an API layer in mind can adapt to new requirements more easily because additional services and automation can be integrated into production when new requirements arise, therefore remaining competitive for longer.

The role of API-first in retail

APIs play a crucial role in deeply interconnected systems that need to interface with other internal applications, third-party partners, and customers — all key areas when it comes to developing powerful retail applications. Think about how an e-commerce platform connects to the different systems making up the purchase process, such as inventory management, checkout, payment processing, shipping, and loyalty programs.

The use of APIs is deeply interlinked with the concept of microservices. Software and data need to be decoupled to enable retailers to meet ever-increasing requirements, including omnichannel and cross-platform integration, seamless experiences across physical and online stores, and the ability to leverage real-time capabilities that enable differentiating features, such as live inventory updates and real-time analytics. APIs can be seen as a bridge for loosely coupled microservices to communicate with each other.

Besides enabling a microservices architecture, an API-first approach offers the following additional benefits:

• Avoid duplication of efforts and accelerate time to market. Developers can work on multiple frontends at the same time, being confident that functionalities can be integrated by embedding the same APIs once ready. Think of multiple development teams working on an e-commerce web application, mobile portal, and internal inventory management system all at the same time. An API enabling the placement of a new order can be seamlessly leveraged by the web and mobile application and fed into the inventory management system to aid warehouse workers. Bug-fixing and feature enhancements can happen simultaneously, avoiding duplication of efforts and allowing new capabilities to be released to market more quickly.

• Reduce risks and operating costs. An API-first approach enables system stability and interoperability from the beginning because API efficiency is placed at the center of the development lifecycle and is no longer an afterthought once the application or functionality has been developed. This approach reduces the risk for retailers and saves money and effort in troubleshooting unstable systems.

• Enable new opportunities and scale faster. A flexible approach revolving around APIs provides more opportunities when it comes to integrating and refactoring the way different client applications and microservices communicate with each other, allowing retailers to improve and scale their IT offering in a fraction of the time. This approach also changes the way retailers can interact with external partners and do business with them since they can be provided with the tools to easily integrate with the retailer’s offering.

• Achieve language flexibility. Effective retailers need to have the capability to adapt their digital offering to different regions and languages. The plug-in capabilities of API-first allow developers to offer language-agnostic solutions that different microservices can integrate with, leveraging region-specific frontends.

What is the alternative?

The four MACH Alliance principles combined (Microservices, API-first, Cloud-native SaaS, Headless) act as a disrupting force compared to the way applications were built until recently. Adapting to a new technology paradigm requires effort and a different developer mindset. But what was there before?

From an API-first perspective, it can be said that the opposite is code-first. With this approach, application development starts in the integrated development environment (IDE), in which code is written and the software takes shape. Development teams know that they will need to build an interface to be able to interact with each function of the code, but it is seldom a priority; developing core functionalities takes precedence over the interface where those functionalities will be hosted and accessed.

When the time comes for the interface to be developed, the code has already been defined. This means the API is developed around existing code rather than vice versa, which poses limitations. For example, developers might not be able to return data the way they want because of the underlying data schema.

Bottlenecks can also occur as other teams requiring the API will need to wait until the code is finalized to be able to embed it in their underlying applications. Any delays in the software development lifecycle will hold them up and delay progress. Although a code-first approach might have worked in the past, it is no longer suitable for dealing with highly interconnected applications.

How MongoDB helps achieve an API-first approach

Simply lifting and shifting monolithic applications to a microservice and API-first architecture will only provide minimal benefits if they are still supported by a relational data layer. This is where most of the bottlenecks occur. Changes to application functionalities will require constant refactoring of the database schemas, object-relational mapping (ORM), and refining at the microservice level.

Moving to a modern MACH architecture requires a modern data platform that removes data silos.

The MongoDB developer data platform provides a flexible data model, along with automation and scalability features to adapt to even the most challenging retail use cases and to multiple platforms (e.g., on-premises, cloud, mobile, and web applications).

MongoDB Atlas, MongoDB’s fully managed cloud database, also provides capabilities to manage the data layer end to end via APIs, such as the MongoDB Atlas Data API. This is a REST-like, resilient API for accessing all Atlas data that enables CRUD operations and aggregations with instantly generated endpoints. This is a perfect answer to an API-first approach, since developers can access their data using the same principles leveraged to connect to other applications and services.

MongoDB’s Atlas Data API provides several other benefits, allowing developers to:

• Build faster with developer-friendly data access. Developers work with a familiar, REST-like query and response format, no client-side drivers are necessary.

• Scale confidently with a resilient, fully managed API that reduces the operational complexity needed to start reading and writing your data.

• Integrate your MongoDB Atlas data seamlessly into any part of your stack — from microservices to analytics workloads.

This article has provided only a sample of what can be leveraged via MongoDB’s APIs. The MongoDB Query API provides a comprehensive set of features to seamlessly work with data in a native, familiar way. It supports multiple index types, geospatial data, materialized views, full-text search, and much more. In the next part in this MongoDB and MACH Alliance series, we will discuss how a cloud-native SaaS architecture can enable full application flexibility and scalability.