Sam Walton, the late CEO of Wal-Mart, believed that if someone's efforts were not directly helping Wal-Mart to service its customers, those efforts should be stopped. He had a public relations staff and a human resources staff that he cut severely because they weren't helping with the end goal of serving Wal-Mart's customers. He would ask individuals what they did in their jobs to serve the end goal; if he didn't like the answer they were gone. Although a bit rough around the edges, Mr. Walton's approach is making inroads in the software systems engineering approach.

Engineering software solutions often involves large investments in time - gathering requirements, determining the optimal solution through small proof-of-concept applications, and careful analysis of existing tools and technologies. What's wrong with this? Simple. Other than requirements gathering, most time spent on projects is not spent with the customer. As most project leads will attest, most project time is billed to project management, or simply "overhead." I'm not suggesting that project management is unnecessary, but it needs to be seriously overhauled.

Customer attitudes and their economic realities are changing. Not long ago I billed a long-time client 50 hours for a small project I was contracting. After seeing the invoice, she remarked that we only spent "10 or so" hours with her. What gives with the extra 40? Of course, about half of the additional hours went to programming. The rest? You guessed it - project management.

The origins of system engineering go back over 100 years. Thomas Edison once said: "Invention is 1% innovation and 99% perspiration." This can-do attitude fueled the American industrial revolution. This "perspiration" approach found its way into today's model of developing IT systems. Edison's European rival Nikola Tesla cringed at Edison's needle in a haystack approach. Why spend 99% of your time going through straw? Why not invent a system that can detect magnetic properties and zero in on the needle? Tesla was correct.

Not long ago, the industry applauded the long hours of developers. Over late-night take out and countless pots of coffee, teams of developers would toil into the wee hours writing, designing, and debugging systems - and billing the customer. Attitudes are starting to change.

Nowadays, long hours are no longer a congratulatory feat. Managers are asking why are all those hours necessary. I am one of them. Quite simply, Edison's perspiration costs money. A development team staring into the same monitor at midnight is nonproductive and expensive. The customer gets billed for time that was not directly spent on him.

The problem is not an easy one to solve. Overengineered systems are expensive; underengineered system are even more expensive when you add time for rewrites and debugging. The point of this is to remember that to build a system that is beautiful from an engineering perspective is not as important as a lesser system that may be uglier but cheaper. The solution is to develop off-the-shelf system engineering practices. While our software is custom, the process engineering cannot be. It's just too expensive.

What the American IT industry is doing is not unlike what the airlines are doing. With increased competition and tight profit margins, most large air carriers are adapting their business model and spinning off low-cost carriers. Chances are that your shop operates as a full-fare carrier like United. What you need to do is operate like, um, Southwest. If you have not started spinning off a streamlined low-cost version of yourself, your days are numbered.

Competition is global and it's not going away. In this election year, you'll soon hear rhetoric about protecting American jobs. Don't let this sway you. Your shop competes directly with shops from Belfast to Bangalore. Get used to it. The irony is that this competition is good for the American economy - eventually. Economists don't like to use the term "eventually." Because eventually we'll all be dead.

It's important to understand that IT offshoring is only part of the huge picture of the convergence of global markets. In earlier times, goods and payments passed across regulated national borders, but production facilities, knowledge, and culture remained within those borders as a national competitive advantage. At present there is an almost instantaneous movement of ideas, information, key people, and capital across national borders. Productive capacity can be quickly established anywhere there is an economic advantage. It is possible to have high technology, high productivity, high quality, and low wages. This is a transnational phenomenon. It is occurring across nations where regions are working together to define markets and make investment by international companies most attractive. Europe is moving toward a trading block. International borders are becoming invisible to trade.

To remain competitive, successful companies are relocating design and production wherever it is advantageous to produce goods and to maintain relationships for selling goods. Businesses that do not follow these economic imperatives are likely to wither. This is the information age and technology has shortened the time to cross oceans. The American programmer can compete in this environment. Think globally and create a streamlined version of yourself. Become the Wal-Mart of software development. You can even hire me as your greeter.