Timing matters

Inventor and technology futurist Ray Kurzweil once said, “Inventing is a lot like surfing; you have to anticipate and catch the wave at just the right moment.”¹ Bottom line: Timing matters when it comes to technology innovation.

In 2008, Gartner published a research paper that introduced the concept of supply chain convergence. Supply chain convergence is about the ability to observe, manage, orchestrate, and eventually optimize end-to-end (E2E) processes that span traditional functional and application boundaries.

Back in 2008, we felt that supply chain organizations needed to do a better job of orchestrating and synchronizing processes, subprocesses, and activities across functional domains such as planning, customer service, sourcing, warehousing, transportation, and manufacturing. In a perfect world, E2E processes would span all domains and application silos, creating flawless information and transactional flows.  

Most organizations want to support E2E supply chain processes. The challenge is that their technology portfolios consist of many independent, loosely integrated applications that are often provided by different vendors. Those applications have various levels of maturity and are based on different technology architectures, often because users built or bought their applications at different times, for different needs, and with little thought given to how they connect with and support an E2E process.

Many companies integrate disparate applications by simply passing data back and forth—but this is not supply chain convergence. Rather, navigating a supply chain application environment in such a way is similar to playing rugby blindfolded. Sightless players run up and down the field tossing the ball—hopefully to their own teammates—before being wrestled to the ground. It’s hard to know what’s going on and coordination is nearly impossible.

In the supply chain management (SCM) application world, the ball might be an order moving from a customer relationship management (CRM) system to an enterprise resource planning (ERP) system to a warehouse management system (WMS), and so on (see Figure 1). Because orders are simply thrown from one system to another, it is very hard to orchestrate the E2E process in one direction let alone bi-directionally.

So, while our 2008 hypothesis was sound, companies were not ready or able to pursue supply chain convergence at that time. The timing wasn’t right. Gartner has revisited this concept repeatedly over the years, but even today, most companies struggle to systematically integrate E2E processes in the fragmented supply chain functional and information technology (IT) environments that are still prevalent in most organizations.

But the time for supply chain convergence may, finally, be upon us. Some progress has been made. Companies have done a good job when it comes to optimizing vertical functional processes that are aligned with their applications portfolio. For example, a WMS does a good job of coordinating the work within the four walls of the warehouse, and a transportation management system (TMS) can optimize the mode and carrier selection process for multimodal shipments. The challenge, however, lies with orchestrating and optimizing horizontal processes that cut across functional and application silos. While cross-functional application integration is doable, it is complex, and true process synchronization across applications and functional domains remains challenging for many companies. Until recently it was impractical, if not impossible, to coordinate activities across all functional domains without some form of coordination technology.

However, the need for coordination across the supply chain has become increasingly more important. Supply chains have become more distributed and outsourcing more pervasive, meaning that network complexity has increased. At the same time, product complexity has also increased. And then, enter COVID-19. The global pandemic and its lingering effects have showed how fragile global supply chains are and have forced companies to rethink how they support end-to-end processes.

The rise of microservices

Remember, the timing of technology innovation matters. When convergence was first discussed, many assumed that the solution was easy: just buy all your supply chain applications from a single vendor. This approach seems logical until we dig deeper into how supply chain applications are built and deployed even within large application suite providers. There are notably different architectural models for delivering supply chain applications. These can broadly be categorized as application portfolios vs. platforms (see Figure 2.)

Application portfolio vendors typically offer multiple functional applications that might share some elements or an integration bus but are mostly standalone applications with their own unique process and data models. There often are redundancies between functional applications (for example, replicated master data), and the vendors have not rewritten their applications in a common shared architecture. Portfolio vendors have often, but not always, grown through acquisitions, yet have chosen not to re-architect and rationalize their solutions.

To move towards convergence, portfolio vendors typically try to address this challenge by layering some type of analytical or orchestration capability that spans their vertical silos. A common name for these is control tower.

Application platform vendors, on the other hand, start with a common architecture, and all applications are built on a shared technical foundation—from the data and process models all the way to the user experience (UX). These vendors are often on the forefront of modern microservices architectures, which arrange application capabilities as a collection of loosely coupled, messaging-enabled services that are fine-grained while the protocols are lightweight. 

These architectures remove most, if not all, redundancies, and functional capabilities (such as picking, carrier selection, or order promising) are rendered once and shared across the platform. Technical instrumentation such as rules engines, monitoring, configuration, and extensibility is also often shared. These vendors typically build their platform solutions organically from the ground up.

Platform vendors address convergence via composability. Capabilities are broken down into reusable microservice components, which can then be assembled or “composed” to build the E2E process. For example, a simple order-to-cash process might be composed by associating an order service, an order promising service, a picking service and forward picking replenishment service, and a parcel carrier rate shopping and selection service. 

With the emergence of composable, microservices-based applications and the rediscovered mission-criticality of supply chains, convergence is now becoming a reality. Today’s composable microservices architectures can support composite processes that bring together subprocesses and activities from specific domains. Users can then merge these into a larger, converged E2E process.

To get to supply chain convergence, supply chain organizations must work closely with their IT partners to adopt a cross-functional application strategy and platform that allows them to horizontally model, orchestrate, and synchronize E2E processes. Also, as they seek to buy new supply chain solutions, companies should increase their emphasis on an application technical architecture that supports composability. Until they have such an architecture in place, companies with heterogeneous supply chain application portfolios will likely have to focus on analytical solutions that can at least span multiple functional boundaries.

Note:

1. Ray Kurzweil, The Singularity is Near: When Humans Transcend Biology, Penguin Books, 2005.