Western Union delivered its final telegram on Jan. 27, 2006.

Make no mistake, Western Union was the Google Talk or MSN Messenger of the 19th century -- telegrams traveled coast-to-coast in less than one day, with the "last mile" covered by couriers who hand-delivered Western Union's signature yellow envelopes.

Alas, the telegram was ultimately doomed by the telephone, obliterated by e-mail and finally dropkicked by text messaging.

What was once iconic had long ago become anachronistic. Tom Standage, who wrote a book on the history of the telegraph called The Victorian Internet, suggests that instant messaging is the true heir of the telegraph.

His eulogy -- "The telegraph is dead. Long live the telegraph" -- reminds us that, for all the hyperbole about technological revolutions and the extinction of all that existed before, history suggests evolution within broad categories and that eventual displacement is more likely in the natural order of things.

That is an important lesson in responding to recurring questions from public employees and private technology companies alike about the next new thing. In answering those questions, analysts tend to chase mandates (homeland security or compliance) and call them trends; or turn promising technologies (Web services or wireless) into magic elixirs that can cure all governmental ills in the hope that one pitch will capture the political imagination.

Yet a reasonable argument is that the much ballyhooed, market-shaping killer apps are only reshaping the edges of categories that have remained constant over time.

Consider an analysis of more than 188,000 requests for proposals (RFPs) from state and local government captured by the Center for Digital Government since the end of the public-sector revenue recession in 2002.

Almost all IT-related market activity initiated by government fits in four conventional categories: hardware, software, networking and telecommunications, and services. With a nadir in the single digits, growth has averaged 22 percent over the last four years. Networking and services are each represented in one-fifth of RFPs while hardware and software each commanded one-third of procurement activities.

There is inevitable ebbing and flowing among categories over the years, and the mix within each category has also changed -- and will again.

A recent snapshot reveals that the hardware category divides into four relatively equal chunks -- desktop PCs (28 percent); servers (24 percent); storage (20 percent); and laptops, tablets and portable devices such as PDAs, cell phones, pagers and other wearable computers that are not place-dependent (22 percent).

In the software category, commercial-off-the shelf products account for 26 percent; popular, three-letter initial software (CRM, ERP, ECM) remains in single digits, not surprising considering that the big-ticket items only come around once a generation or two; and custom-developed software accounts for 10 percent.

While custom-developed code is only a small part of the software category, developing and implementing it helps drive 40 percent of the services market, which is second only to maintenance at 41 percent.

The average of 47,000 RFPs per year clearly indicates that political subdivisions are busy doing something, and doing so consistently over time -- although there is no clear line of delineation between RFP activities intended to prop up old, anachronistic processes, and those that displace them in favor of a public service delivery transformation.

Samuel Morse, inventor of the Morse code, sent the first telegram on May 26, 1844. He had a sense of reverence over displacing the Pony Express with an early network that could go on forever.

That first message read, "WHAT HATH GOD WROUGHT?"

Paul W. Taylor  |  Contributing Writer