Some history: After the Internet went mainstream in the early-to-mid 1990s, networking and infrastructure specialties became mission-critical. Designing, implementing and operating IP networks, client-server computing systems and storage (including the newly invented storage networking technologies) were major challenges. Staffs swelled, and IT managers struggled to cope with challenges around hypergrowth, particularly when it came to rapidly rolling out infrastructure that could scale exponentially.

Meanwhile, telecom deregulation flooded the market with new players touting exotic technologies and business models—and the emerging competition brought prices down sharply, for those who knew how to negotiate. (Throughout the late 1990s, telecom prices dropped an average of 20 percent year over year).

After the dot-com bust, the economy moved into a period of more moderate growth, and the watchwords for infrastructure teams were consolidation and optimization. IT folks worried less about fast, scalable buildout, and more about optimizing operations and reducing total cost of ownership. As these efforts proved successful, IT departments began divesting themselves of infrastructure engineering expertise.

With outsourcing, we're seeing the natural conclusion of that evolution: IT departments are increasingly handing off their networks to managed services providers. All indications are that the trend will continue for the foreseeable future (the growth in demand for managed services is hockey-stick steep). Add to that the ongoing trend towards relying on the Internet to connect remote and distributed workers—as a full 81 percent of the IT professionals I work with say they do—and it becomes clear that telecom and networking is increasingly about crafting credible service-level agreements, negotiating rock-solid contracts, and effectively selecting and managing providers.

It may seem like these developments mark the end of the career of the "network engineer." Not at all. As new technologies come on line, there's plenty of network engineering—and architecting, design and implementation—left to do. The trick lies in understanding the distinctions between technologies that have become commodities (and can therefore be outsourced) and those that are emerging as strategic differentiators.

Some examples: Unified communications (including presence and conferencing) is revolutionizing how far-flung organizations interact. UC is all about gluing together disparate systems and services—something networking folks have, by necessity, gotten good at. So networking teams should be taking lead in UC architecture, strategy and implementation.

Mobility's another example. Many firms are considering offloading support for mobile devices and strategies altogether, and simply asking employees to provide (and pay for) their own cell phone services. Big mistake. Mobile devices and services aren't just portable phones—they're the computing and communications platforms across which next-generation applications will be delivered—and need to fit in with an overall strategic architecture.