What happens when you want the guts of a system—data models, business logic, and machine learning capabilities—without the baggage of being boxed into the platform’s standard user experiences or simplified workflows?
APIs and widgets provide flexibility to extend a platform, which may suffice if you have a few extensions or simple integrations. But what if business requirements force you to bolt on many workflow customizations, complex real-time integrations, and significant design customizations?
The added code may become complex to support, and you might think it’s better to customize a solution from scratch. But developing an ERP, CRM, CMS, e-commerce, search, or other complex system from scratch is daunting and expensive.
What are headless architectures?
Advanced development teams may find a middle ground using platforms designated as headless architectures. These platforms have API-first architectures that deliver a complete back-end system with databases, business logic, and integrations. They may provide basic user-facing UIs, but the expectation is that development teams will fully customize the customer-facing front-end user experiences and integrations using the platform’s APIs or SDKs.
Gordon Allott, president and CEO of K3, says, “Headless architectures harken back to the old [Jeff] Bezos legend and message to his employees, ‘Build what you want, how you want. But it must include an API that encompasses all communication.’ ”
Here’s a summary of the famous Bezos mandate, which many people attribute to Amazon’s success in launching AWS.
A good option for CMS and search
Some headless architectures provide back-office tools, but the front-end customer experience is custom-built using the platform’s APIs. For example, a headless CMS may provide tools for creating and publishing content, whereas the development team custom codes the customer experience using a JavaScript framework of their choice.
Employee- and customer-facing search is another area where many organizations use headless search. The search platform typically provides the back-end tools to integrate content sources, manage taxonomies, develop search indices, tune search relevancy, and configure recommendation engines and other machine learning algorithms. Instead of using the search platform’s UI, development teams use the headless architecture’s APIs to build webpages, mobile apps, and components that integrate with software as a service and other platforms.
Arvind Jha, senior vice president of products at Newgen Software, says, “Headless architecture is all about freedom that one gets from an interface perspective. Large enterprises with higher IT maturity are adopting a headless content approach to get the best of both worlds—lightweight UI frameworks and API-based content services.”
Why headless architecture?
Customization flexibility, especially around a customer-facing user experience, is one reason a dev team may want to leverage headless architecture. Kashyap Deorah, founder and CEO of HyperTrack, says headless architectures are also used to connect complex workflows spanning multiple systems.
“When multiple systems with distributed ownership unify to power a workflow or experience, consider headless architectures,” Deorah suggests.
Deorah gives an example: “The fulfillment workflow of an e-commerce order might involve a different system that manages the cart, warehouse, courier selection, driver app, and proof of delivery. Each system might have a different owner, scale differently, and run on a different platform, yet they must all come together to power an end-to-end order fulfillment experience for the customer.”
It’s not just workflows across multiple systems that drive the need for headless architectures but also the growing number of user devices. We went from supporting web interfaces to mobile-first architectures; today, more organizations support UIs in watches, cars, and home assistants, and companies want to be ready to support metaverse experiences.
Amit Patel, senior vice president at Consulting Solutions, says organizations optimizing experiences across multiple devices should consider headless architectures. He says, “If you want to provide a true omnichannel experience, then a headless architecture is a perfect choice in that it allows for your digital content to be seamlessly delivered to users across multiple customer touchpoints—on desktops, mobile, smart watches, smart-anything—without regard for underlying platforms and systems.
Flexibility in user experience design, complex platform orchestrations, and omnichannel experiences are three reasons development teams may opt for headless architectures.
Composable systems build on headless architecture concepts
Headless architectures provide one level of front-end user experience customizations, but they don't address middle and back-end flexibilities. The next level of modularity and decoupling comes with composable systems that allow organizations to cherry-pick functionality using different modules from different platforms.
Commerce is a compelling use case, especially for businesses with multiple lines of business or operations in different geographies.
Composable commerce architectures are open, flexible, and business-centric solutions that optimize different e-commerce experiences. These solutions extend beyond headless capabilities and allow for decoupling orders, payments, catalog, inventory, and other modules bundled in e-commerce solutions.
Jason Cottrell, CEO of Myplanet (now Orium), shares several other examples of how business complexities lead to benefits using headless architectures and composable systems. He says, “Brands selling in regulated environments—launching a direct-to-consumer line alongside their wholesale business or tightening omnichannel integration between web and store—are just a few examples of the complexity composable commerce is well positioned to support.”
Note that composable commerce offers flexibility, and modularity provides choice, but swapping in one building block for another is generally not free. Composable commerce is not Legos, at least not yet. Much work is needed to create truly open standards.
Seeking flexibility at every layer
Architects use layers to define how services and applications are developed, so here’s one way to think of the connection between microservices, headless architectures, and composable systems.
- Microservices enable dev teams to build, deploy, and orchestrate workflows more easily by reducing dependencies and automating the delivery of small, atomic capabilities.
- Headless architectures enable the decoupling of back-end capabilities from front-end experiences.
- Composable systems help organizations leverage best-in-breed modular capabilities from multiple platforms or service providers.
Additional architecture considerations include how to implement shift-left security practices, when to consider multicloud architectures, whether to build capabilities with no-code or low-code paradigms, and what types of data architectures optimize performance.
For simple apps, the best advice is to stick with simple mainstream architecture. But when you face security, scale, and business complexities, there are a growing number of options to optimize a flexible architecture.