'Lightyear' Breaks Decisions Down to Their Simplest Parts By Michael Schrage

Lightyear is a horrible name for a very interesting program that's just hit the market.

But what's in a name, anyway? I didn't much like Lotus 1-2-3 either, fearing it would spawn names like Manna 4-5-6 and Ambrosia Plus. But who wants to argue with success?

I certainly don't, and I think Lightyear has a genuine shot at being one. Costing a cool $495, Lightyear is the finest example to date of what we like to call "decision support" software.

While database programs, word processors and the reams of integrated spreadsheet software that crowd computer store shelves are nice and useful, they are fundamentally information jugglers. They are tools that let you do a better job of dealing with data. They support your decision-making efforts.

What they don't do is let you explore the process of decision. What are the biases and values that go into the decision criteria? How are they weighed? What do they mean? How does the decision trade off between them? These are precisely the questions Lightyear is designed to address.

Most programs represent a new way of doing old things, like filing data or typing reports. Lightyear, produced by a Santa Clara, Calif.-based company with the same name, offers a new way to do new things. It is a tool that lets you model, albeit crudely, the decision-making process.

Specifically, it gives managers a disciplined way of examining the key variables that go into making a decision. That's important because a manager should not only know how a decision is made, but be able to explain the process to others. Take a variety of examples: site selection for a new plant; a new marketing approach for an old product; hiring a new employe or investing money. Lightyear provides a framework for dealing with each of them. It may not look pretty and it's certainly not profound -- but it's potentially quite valuable.

Lightyear allows you to fashion a decision model around the variables you think are important. Consider the case of investing money.

Lightyear asks you to list the alternatives. There are stocks, bonds, money market funds, art, gold, silver, etc.

Next, consider -- and weigh the priorities -- of your key investment criteria. Let's say they are risk, return and liquidity.

Skip back to your alternatives and -- numerically, verbally or graphically -- rate them according to these criteria: chancy, risky, sure thing, etc.

Then decide what sort of rules or guidelines should govern your investment approach. Suppose return on investment must be at least 10 percent. And, if an investment is very risky, then the return should be at least 25 percent. Maybe risky investments should not be more than 20 percent of the portfolio.

Play out on the screen which investments meet your guidelines and which ones do not. You also can see how each individual investment ranks according to rules and criteria. You can compare specific investments and evaluate the quality of different investment models.

If all this sounds suspiciously like those "expert systems" with quasi-artificial intelligence, you're absolutely right. What Lightyear does is offer a quick and dirty way of letting you create your own expert system that shows you what your decision-making biases are. That can mean a great deal when you're deciding whether to hire someone or acquire a company.

Don't dismiss the importance of a structured approach to decision making when explaining to others how a decision was made. The beauty of this approach is that you often can find the logical inconsistencies in your thought processes that lead to bad decisions (or at least poorly thought out ones).

There is no question in my mind that, regardless of whether Lightyear survives as a company, this kind of decision-support software will become a fixture on managers' desks. It's a software way to introspection. Needless to say, (though I'll say it anyway), you will see this kind of software linked to Lotus, DBaseIII and project managment software as part of an integrated decision support environment.

Reliable rumor has it that Lightyear fits quit nicely on the IBM PC AT's TopView operating system so data from one window can be cut into the Lightyear decision model. Or several models can be displayed simultaneously.

John Couch, formerly of Apple and one of the savviest people in the business, is Lightyear's president. Couch sees all sorts of vertical market applications for his new program, ranging from personnel planning to McKinsey & Co. consulting packages based on Lightyear. He's counting on at least 25,000 sales during the next year.

While Couch insists that Lightyear's stiff pricetag won't drop, don't be surprised if he licenses less robust versions of the program to lower-cost software houses. In fact, there will be a lot of programs like Lightyear on the market by the end of spring.