Why Heterogenous IT Fuels Modern Business

As a digital strategist, I spend a lot of time talking to customers and potential customers. Much of that time is spent becoming intimately familiar with customer challenges and problems, so we can work together to solve those problems.

Frequently, these challenges stem from a tendency, often rooted in industrial-era principles, toward over-specification. This tendency is often at odds with the principles that drive modern business, creating tensions that only perpetuate these challenges as enterprise systems evolve to accommodate new business ecosystems and take advantage of cloud services and other new technologies.

As we transition from these aging, overly-specific, and overly-proprietary tactics to new paradigms based on concepts such as platform strategies, ecosystem growth, and demand aggregation, our perspectives on technology must change as well. This article will focus on two specific changes:

• IT must be seen as a catalyst for strategic agility, not as custodians of servers, networks, and other infrastructure.
• Enabling IT’s strategic role in the business often requires enterprises to embrace a wider variety of solutions and manage increased complexity.

From the Industrial Age to the Internet Economy

Many enterprise strategies remain grounded in a corporate heritage that goes back to the industrial revolution, if not earlier, and a fixed, rigid dynamic between suppliers, distributors, and consumers. Industries such as large pharmaceuticals and financial services have traditionally been naturally structured around this classic supply aggregation model, for example.

In the Internet economy, we see a very different dynamic, one in which — instead of procuring raw materials, improving them, and bringing them to market — Internet-based platforms orchestrate between suppliers and consumers.

In addition to difficulties rooted in traditional corporate methods, many companies face additional challenges because of the ways their industries have evolved over time. For instance, many telecommunications and cable TV companies evolved through a series of mergers and acquisitions as they drove to extend their footprint and offer services to more people. The bevy of legacy applications, systems, and resources these companies consequently amassed can be an ever-growing albatross, a heterogeneous expanse of operational overhead and associated costs. These brittle architectures cannot deliver the agility modern internet economy dynamics require.

In both of these scenarios, the key antipattern is over-specificity. Companies developed proprietary “stacks” of legacy solutions that were tuned to specific problems in specific contexts but that are not equipped for today’s more fluid economy. These over-specific approaches provide neither the speed and agility that businesses now need nor the flexibility to participate in new digital ecosystems formed by other companies, developers, and consumers. This is why a new approach is so necessary.

Consider Pitney Bowes, a customer of my employer, Google. Pitney Bowes has long provided companies of all sizes the ability to efficiently utilize mailing and shipping for their business needs, including mass mailers, customer billing, general correspondence, and marketing campaigns. As correspondence and commerce now span both physical and digital channels, Pitney Bowes is making their solutions available on open platforms.

By making its capabilities, such as location intelligence or payments, accessible via well-managed APIs, Pitney Bowes has positioned itself to orchestrate shipping, billing, logistics, and other services and resources between end-user customers and the enterprises that need to reach them. The company is transforming the mailing industry by offering a suite of cloud-based location intelligence and customer information products to create new ways of doing business.

Digital Transformation: Embracing Heterogeneity and Managing Complexity

As the preceding example indicates, modern enterprises continue to focus on proprietary strengths, just as they did in the industrial age — but executing on those strengths requires support for an ever-increasing variety of business requirements, use cases, and support systems.

This shift toward more diverse solutions and greater complexity can open a chasm between expectations and delivery capability (more features, more flexibility, in less time), and lead to quicker, more graceful market exits when ventures fail. This is the chasm across which IT leaders and business leaders frequently spar — and it’s the chasm that digital transformation attempts to bridge.

Few businesses today invest in a five-year time horizon for a new venture; because windows of opportunity open, close, and change faster than ever before, even one year is a lifetime. Nevertheless, in order to be agile — and Agile — businesses must unite their focus on short-term market demands with a longer term view on technology decisions and their impacts.

This longer term view is essential to preserving the flexibility needed to continually serve customers. Understanding the experiences of the developers and customers that engage with your company is a critical way — if not the only way — to make your products and services compelling. In a perverse transposition of Conway’s law, it is often a variety of technologies that enables organizations to develop systems that are a mirror image of their users, facilitating a value exchange by meeting users and developers where they are, regardless of how the legacy systems were built.

Building a heterogenous network — one that meets users and developers where they are — would have been largely impossible 15 years ago. Most systems were not inherently modular, and even if they were somewhat modular, they did not interact well with third-party systems, so infrastructure was largely homogenized. Services Oriented Architectures (SOA) and Enterprise Services Buses (ESB) were introduced to help couple the systems, and provide a heterogeneous approach — but they had a major unintended consequence: tight coupling meant treating all systems as one system. Unfortunately, the coupling between systems was so rigid it meant treating those systems as a single monolith, shackling design, development, testing, and deployment to multiple complex components.

Luckily, companies today have more options. For example, many enterprises use APIs to manage lightly-coupled services from which they granularly build applications — a completely different paradigm. Having explored why diverse solutions that support modern demand aggregation are increasingly critical, in my next article I will explore how APIs have supplanted approaches such as SOA, and how APIs can enable companies to assemble a diverse technology stack of cloud assets, mask complexity, and deliver value.