Many companies are launching major information systems projects: some are Year 2000 retrofits, others are Internet and electronic commerce solutions or major enterprise resource planning projects. A large amount of money is being spent. The question I hear being asked is, "How much money should we be spending?" The question seems to be answered either by benchmarking a company against others in the industry or just an edict from senior management that the total budget for IS should be a certain number. Neither approach is correct.

The correct approach is to answer two questions:

• How much to fix the plumbing?
• Where can information systems help us make more money?

The same logic applies to information systems at a company. Year in and year out you will have to spend money maintaining your current (legacy) systems. The relentless development of new technology will force this. Software will have to be upgraded to new versions, hardware will have to be replaced, and new network and communications systems will have to be upgraded and replaced. All of these expenses will not give a return in the traditional accounting sense. All they are going to do is keep your systems running. If you don't spend the money, your systems will slowly (or not so slowly) degrade.


Buying new systems
As your systems get older you will spend more money on maintenance — just like your house. The question then becomes: When do you stop spending money on maintenance and spend it on new systems? The answer is when you reach a point that the amount of money to maintain the system starts reaching a level that could be used to finance a new system.

To finance new systems the question has to be: "How can information systems help us make money?" The answer to that question is simple — information systems either help you take costs out of your current business processes or they help you improve your value to your customer by giving your customer a new product, service, etc.

Taking costs out is simple to measure. You should have process flowcharts and metrics that show what it takes and costs to do things today. Charts and metrics showing the "to be" are prepared. These "to be" costs are then compared to the current process and costs savings are identified. The decision process is straightforward — the money saved justifies the expenses.

The second way to make money is to find ways to give your customer better or new services that increase your sales and market share. You are taking risks here because you will never know with certainty how the customer will react to your moves. But in today's more competitive, global marketplace, you have no choice but to continually bring out new services or tools for your customers. If you don't, you are on the path to failure.

Example: Dell Computer set up a Web site where customers can order computers. There was a lot of skepticism, but Michael Dell decided to move forward. The result? The Web site has been very successful both in getting Dell out to a larger market and in generating hard sales.

Example: A number of companies are putting their parts catalogs and product information on Web sites so customers can get information on products immediately. The information is always up to date, presented in an exciting format, and customer tracking can be done. Some of these sites have been successful while others have failed.


Spending the bucks
The fact that can't be changed is that you are going to spend money on information systems. Spending money on fixing the plumbing isn't glamorous and doesn't win anyone friends in executive positions, but you are going to do it. You are going to spend money in trying to offer new services and products, and you will constantly be trying to take cost out of your processes. The question then becomes how well you manage these activities. Well-managed information systems projects and processes make money — poorly managed ones cost plenty of money.