flux@microsoft.com

n a meeting the other day, a former engineer now in the marketing department was describing an idea. The idea applied a new technology to a different market. Asked why he wanted to do this, he talked on and on about how the company had experience in the technology and wanted to get in this new market. There was no mention of a problem that would be solved, let alone a mention of the potential customer. This is classic engineer-think. Solutions are cool, and are readily provided by competent engineers. It’s the problems, or rather recognition of the problems, that is hard. Let’s look at a few solutions in search of a problem.       A classic solution to a non-problem is the Network Computer. The idea of the NC is to have a generic machine that can be manufactured by anyone and that will run generic software from a variety of vendors. That sounds noble, but just when in the last 10 years has interoperability between hardware vendors and software vendors been a problem? The Wintel architecture—the combination of Windows® on top of a personal computer derived from the old IBM PC/AT design—already solves this problem. Dozens of hardware vendors produce PCs that are Wintel-compatible, and hundreds of software vendors produce software for these machines. Folks may not like paying for Microsoft® Windows licenses, but the same situation would result if the world were populated with NCs and folks had to pay Sun for a Java license on each machine.       So what problem does the NC solve? It’s an attempt to remove Microsoft from the PC equation. Not that this is wrong; any company should try to beat its competitors, Sun included. The question is, what problem does the NC solve for consumers? While there is talk of total cost of ownership, the real goal of the NC is to replace the license fee for Windows with a Java license fee. That’s not much of a help to Joe Consumer.       Another widely pushed "solution" is the smart card. That’s the credit card-shaped device that is supposed to hold electronic cash that can be deducted from the card when you buy something. Here is a classic example of solving a non-problem. Today, consumers can pay for their purchases either by cash, credit card, or ATM card. Cash, especially the U.S. dollar, works anywhere on the globe—no wires, no fees, no worries. Credit cards add flexibility because they let you purchase items even if you don’t have the cash on hand to pay for them. The other big plus with credit cards is that if you lose one, your losses are limited to $50 by federal law. ATM/debit cards add one more option of payment.       So what problem does a smart card solve? The problem of how banks can get more money out of your pocket. All that electronic cash in a smart card must be backed up with real cash in a bank somewhere. While you’ve got that smart card loaded with cash, the bank is collecting interest on the real cash back at the bank. With millions of potential users, banks stand to realize billions in potential profits.       The Internet provides an entirely new vehicle for providing solutions with no problems. A couple years ago in Flux I reviewed software that provided a three-dimensional environment that you could wander around and talk to other "people" logged on to the same server. So what? Essentially, this product was a glorified chat room. It was flashy, but it didn’t solve any problem for the user. The only problem I could figure this software solved was how to get money out of a venture capitalist’s pocket.       There are times when a solution finds a problem. Take the case of the Apple II. It was created with some ideas in mind: word processing, games, and so on. But the problem that the Apple II solved was to make the task of running the numbers easier for businesses. Running the numbers is where some low-level accountant or manager changes a few base numbers and has to recalculate all the other columns to get to the bottom line, profit/loss number for the company. VisiCalc automated all that. The original IBM PC repeated the same success, this time with the relatively easier to use Lotus 1-2-3.       The Internet itself was a solution awaiting a problem. Originally sponsored by the Department of Defense to connect university and national laboratory computers, what we now know as the Internet sat around for years waiting for a problem appropriate for the solution. The breakthrough was the creation of Mosaic, the original hypertext browser that used the Internet as a text delivery medium. The problem that the Internet and Mosaic solved was to provide access to information to the masses. In effect, the Internet has become the world’s largest public library.       It’s all well and good to develop a new technology without a specific solution in mind; just don’t expect to make money on that technology until it solves some problem. People don’t buy technology; they buy solutions. This maxim is even more important when working on Internet solutions since the technology is so new and interesting solutions are so easy to come by. So next time you come up with a cool new application, ask yourself: just what problem does the application solve?

From the November 1998 issue of Microsoft Interactive Developer.