A recent Activemedia study found that only 31% of all commercial Web sites are profitable, and most of those marginally so. Frankly, I was surprised the number was as high as it was. A brick-and-mortar organization that rented the most impressive office available, decorated exquisitely, and then just waited for the customers to walk in would undoubtedly end in failure. Yet this seems to be the model on which most corporate web sites are based.

Development resources are spent making sure Web sites use the latest technology and incorporate as many high-tech baubles as possible. The theory seems to be that if the site is flashy enough, thousands will drop in to have a look and the site will be a success. This is an odd measure of success. Having a highly trafficked office or store is not the same thing as having successful business. The cool Java applet you commissioned might draw thousands for a few weeks, but unless that traffic translates into sales or savings, it was probably a waste of money.

I'm not putting down the effort that goes into producing some of the flashier sites. The creative programming and design work that goes into the best of them are astounding, and these sites deserve every design award they receive. It's just that many of these sites seem to be designed to win awards rather than to attract customers. Creative use of ActiveX or non-standard Netscape plug-ins may get you mentioned in magazines and talked about on the Net, but they will be unusable for more than 80% of your potential audience.

Some sites actually boldly proclaim "Designed for <browser> 3.0 or later." Perhaps this is a designer's way of saying "our site uses the latest technology", but what it says to a potential customer is "We couldn't be bothered to design and test a site that would work with your browser, so download and install the browser we use and come back later." Trust me, they won't. Web sites must be designed to look great on all browsers -- even with graphics turned off. This may mean that you are using last month's uncool technology, but in the end, it's not what's cool, it's what works.

1) Set written goals

Know why you are you putting up a web site. Do you want to sell a product? Build a brand? Reduce operating costs? Enhance communications? Improve customer support? Get feedback on your products? If you don't know what you want, you are not going to get it.

2) Design to achieve your goals

Every element of your Internet presence should be considered solely in terms of how it will help achieve your goals. Putting a product database on the Web will not win awards, but it may go far to achieving the sites goals. Great content and stunning design will help you achieve your goals only if they are created with those goals in mind.

3) Devise Metrics

Come up with concrete metrics to determine the success of your site. I generally don't recommend using hits or number of visitors as a measure of success. How many units do you expect to sell? How much of a load can be taken from telephone customer support? How long will it take for the site to become self supporting?

4) Monitor and modify

Once the site goes public the real work begins. Listen to your visitors. Examine your log files to learn how people move through the site and modify it to direct them towards the most important content. If the goals are not being achieved, modify either the site or the goals.

5) Return to step one and repeat for as long as the site stays on the Web.

Building and maintaining a Web site does require special skills, but in the end, success depends far more on effort and planning than on technical expertise. In future articles, we'll look at how a few innovative Japanese companies are using the Internet successfully.