THE NETWORKOLOGIST

Business Justifies The Expenditure

by Patricia Schnaidt

A wall of new technologies confronts us. The underlying network and computing infrastructures are deployed widely enough that we can start building more sophisticated applications and begin to worry about the slow spots in our networks.

Yet, how do we justify those applications? As the effort moves from connectivity to communication, we are faced with the challenge of building a justification for the bevy of new applications and networking products ahead. That justification must also have a business, as well as a technology, rationale.

A raft of new applications dangle the promise of productivity. The Web betroths us a diamond mine of valuable information if only we had the time to find what we wanted. Computer-telephony integration lures us with phone calls that follow us across the country as well as a tighter link between our desktops and our phones when we're in the office. Desktop videoconferencing hopes to make faceless phone conferences as antiquated as rotary phones. Object-oriented component software vows to slim down our memory-stuffing applications and enable us to work more flexibly and with our favorite pieces of different applications. ATM, Fast Ethernet and switching promise to make the underside of the network sufficiently speedy so that we can use all of the productivity-enhancing applications.

The vendors spin their stories so that these applications and infrastructures appear as risk-free as installing NetWare and 10BASE-T. Sure, the new applications of technology will show the greatest delta in productivity, cost or other measures of success, but the hype surrounding the product often outweighs what the product can ever achieve.

Vendors typically market their products by emphasizing their new features, but how many more features can you pack into a word processor before more than two percent of the user population could possibly find a practical use for them? How many more bits can a switch forward while holding latency and errors to a minimum before we hit wire speed? How many more SNMP MIB variables can a management station collect when the users are complaining about application performance? Don't get me wrong--improving the functionality of networking and computing products is essential to the progress of using computers to facilitate business. Yet the focus on the features and hype can distract us from the potential of real business improvement with this technology. We're losing the trees of business reasons in a forest of spec sheets and marketecture.

Companies must be enticed to implement new applications and their supporting technologies, and fewer will be lured by new technology. Instead, they'll be drawn in by improvements that support the business. It's a tough request for a technology-based industry. We've harvested the low-hanging fruit of worker productivity with PCs, desktop applications and LANs. Now we must figure out how to squeeze the next improvement from a far more complex application, groupware, and do it across the enterprise. For companies to implement the new raft of applications--Web applications beyond simple browsers, videoconferencing, computer telephony, multimedia--there needs to be a strong business case.

Worker productivity isn't moving lock step with the overall expenditures in technology, which makes additional technology expenditures more difficult to justify. Is the expenditure difficult to justify because the real business improvements are not there? Do we just not need the next great thing? Perhaps as we sque eze the productivity funnel, there's less to be gained. Is it difficult to justify because we don't have the language and the business models to quantify the improvements? Or perhaps the metrics and language must change from a product orientation to a service orientation. Are we trying to quantify the unquantifiable? We struggle to find the answers.

A Tale of Two Panels At the Enterprise Management Summit in Dallas, some very experienced network managers and I will be dissecting what's wrong with network management at a panel called "Making Enterprise Management Work: A Real World Perspective" on Wednesday, October 25. At Comdex Fall I'll chair the discussion of "LAN Switching Options" on Tuesday, November 14.

Patricia Schnaidt can be reached at pschnaidt@nwc.com.

October 15, 1995