Software Doesn’t Have to Crash

If we’re so smart, asks Alan Cooper, why is our software so dumb? His manifesto: “Most software needs to be spanked.”

According to Alan Cooper, the problem with software is not the way it works but the way the industry that makes it works. “There’s a fundamental problem with how the software business does things,” says Cooper, 46, who started his first software company in 1976; who invented Visual Basic and later sold it to Microsoft; and who now heads Cooper Interaction Design, a 30-person firm based in Palo Alto, California. “We’re asking people who are masters of hard-edged technology to design the soft, human side of software as well. As a result, they make products that are really cool — if you happen to be a software engineer.” Small wonder that Cooper Interaction Design has chosen the manifesto “It’s a fact: Most software needs to be spanked.”


Why such drastic punishment? Because designers and programmers aren’t just building flawed products. They’re also punishing users — and, in the process, doing social and economic harm. “We’re building what I call ‘software apartheid,’ ” Cooper says. “We’re in the process of creating a divided society: those who can use technology on one side, and those who can’t on the other. And it happens to divide neatly along economic lines.”

Building great software, Cooper argues, requires embracing a new approach to design, one that distinguishes between three challenges: conceptual design, interaction design, and interface design. Conceptual design shapes what a product does. Interaction design determines how it behaves by examining what users are trying to do with it. Interface design affects how it looks and feels. Cooper’s firm deals with the second challenge — how computer products interact with users — yet it doesn’t employ a single computer programmer.

Cooper’s approach to design is winning lots of converts, including clients such as Sun Microsystems, Coca-Cola, Compaq, and Dow Jones. And it’s already being applied to products such as image scanners and computerized medical devices. A few years ago, Cooper Interaction Design began working with Logitech International SA, a leading manufacturer of mice and other PC peripherals. Logitech asked Cooper to design software for its new Peacock scanners. The purpose of such software is to help users edit and manipulate digitized images. But Logitech worried that poor design caused users to spend more time on reconfiguring controls than on editing images.

The new design simplified the software so dramatically that users now deal with just three image-manipulation functions — and need not fuss with a single hardware-management function. “I think the phrase ‘computer-literate’ is an evil phrase,” says Cooper. “You don’t have to be ‘automobile-literate’ to get along in this world. You don’t have to be ‘telephone-literate.’ Why should you have to be ‘computer-literate’?”

And as computing power weaves its way into every aspect of life, poor design threatens to make everything harder to use. “A revolution is about to occur,” Cooper warns. “Computer chips are in everything. I have a digital camera that behaves like a computer. I have a cell-phone that doesn’t behave like a phone: It behaves like a computer that makes calls. Computers are becoming an integral part of daily life. And if people don’t start designing them to be more user-friendly, then an even larger part of the population is going to be left out of even more stuff.”

Sidebar: Less Is More by Design

Cooper Interaction Design has been working with Varian Associates Inc., a major electronics company (1998 sales: $1.4 billion) that manufactures computer systems for medical and industrial instruments. One division of Varian, which builds nuclear magnetic-resonance-imaging machines (MRIs) for medical and geological purposes, wanted to simplify the software that controlled the MRIs’ huge magnets (some of which are as big as a compact car). Cooper’s preliminary research identified three main user groups for a nuclear MRI: scientists who devise methodologies for using the instrument, researchers who use it to develop experiments, and technicians who use it to perform those experiments.


Each category of users has a distinct need. But the Varian MRI’s existing interface was designed to serve all three categories at the same time. “So it was an incredibly complicated program,” says Cooper. “It followed the mechanics of what needed to be done — but it didn’t follow the way the human mind really works.” Cooper’s solution was to create a separate, appropriate interface for each set of users.

The MRI redesign underscores an important principle behind the Cooper approach: Less really is more. Cooper improves applications not by adding features but by removing them. “In the software industry, if you can add a feature, you do it,” says Cooper. “But too often, features obscure the purpose of a program and make it harder to use.”

Learn more about Cooper Interaction Design on the Web ( or contact Alan Cooper by email (