The price can be high when you order up a system before you assess your business needs

Given our optimistic beginnings, it was hard to believe. But after a $250,000 investment in technology, our systems were so exotic we had to jump through hoops just to print a letter. And for 18 grim months, we were besieged by frustrated clients who complained that we never answered our telephones. Little did they know that our state-of-the-art telephone system was so bug-ridden that on some days our phones never rang.

The source of the problems? Separating the systems from the business and hiring an outsider to set the systems up.

When we founded the Netherlands Design Institute, in 1993, our hopes were boundless. The city of Amsterdam had given us a 10,000-square-foot canal-side building, and the federal government had thrown in $1 million with which to outfit it. Our mission was to bring together company managers, technical and cultural experts, and designers to consider future scenarios concerning the implications of new technology for their businesses. We were determined to keep our in-house operation small, so we turned to technology -- internal networks, a shared database, teleconferencing, E-mail, on-line discussion software, and the World Wide Web -- to get the job done.

Thus, it was in the context of a business that did not really exist yet that we set out to build "the systems." That was blunder number one. We ended up separating the systems from the questions that should have defined them, namely: What is our business? How can we do it better using technology? What software and equipment does a given project require? And because my knowledge of information technology was limited to my experience on a personal Macintosh SE, I asked an outside expert to set us up. That was blunder number two. The result: Nobody in-house really knew how the systems worked -- or cared about making them better.

We had to rush at the beginning because the builders were already on-site. So we quickly specified cabling for up to 80 workstations (part of our business concept was that outside people would use our facilities to work with us on projects). We made our local Apple dealer happy with an order for a network comprising 15 desktop computers linked by five miles of cable. Then we added a $50,000 state-of-the-art digital telephone system, ISDN connectivity, and boxes and boxes of software that, according to our consultant, was the most up-to-date available. In order to do simple tasks -- printing a letter, for example -- we had to configure the network software and the shared database without delay. The consultant had to guess, from hurried conversations with me, what the system was supposed to do.

We ended up with a fine-looking front end to a system whose inner architecture none of us understood. The telephone system was indeed the latest model -- but after we stopped wondering why the phones were so quiet, we spent our time showing a team of telephone-software engineers where to find the recalcitrant control box. Other costs took longer to make themselves felt. The database, for example, was almost too easy to use. After a year it contained a few thousand names and addresses (in the plan), but it also offered more search combinations than there are atoms in the universe (not in the plan). And though we created a very popular Web site (, it was so hard to maintain a flow of new content that we had to relaunch the whole project.

How much did all this cost? Who knows? I suspect we lost several person-years of productive time by approaching things backwards. Ideally, we should have asked the right questions at the beginning. We didn't, so we paid the price. But that price may turn out to have been an investment -- in the form of on-the-job (albeit unplanned) training. Today we run the system ourselves. Staff members "own" bits of the information environment. The Web-site team, for example, is responsible for the content, the design, and the programming of our page, and everyone looks after his or her own bit of the database. And we don't spend a penny on new software or equipment unless a project demands it. Are the lessons we learned worth the cost of the blunders? You bet they are.

* * *

John Thackara ( is the director of the Netherlands Design Institute in Amsterdam, The Netherlands.