Washington Apple Pi

A Community of Apple iPad, iPhone and Mac Users

electric pi


The Inmates Are Running the Asylum: Why High-Tech Products Drive Us Crazy and How to Restore the Sanity

© 1999, Paul J. Chernoff

Washington Apple Pi Journal,January/February 2000, reprint information

The Inmates Are Running the Asylum by Alan Cooper. Indianapolis: Macmillan Computer Publishing, 1999

Would you trust a computerized refrigerator to order food for you? I wouldn't. My life is already filled with difficult-to-use high-tech devices. My entertainment center's remote controls are too complex. I can program my VCR, but most people still cannot set the time. A best-selling word processor prohibits me from manually numbering paragraphs. It seems that whenever you combine a device, such as a clock, with a computer chip, you end up with something that is more difficult to use than its uncomputerized counterpart.

Alan Cooper is on a mission to make computerized products and software easier to use. He finds the problem is poor interface--or interactive--design, not technology. In The Inmates Are Running the Asylum, he presents "the business case" for good interactive design. He shows how bad things are, why they are bad, the benefits of good design, and an approach for successful interface design.

Cooper prefers the word "interaction" over "interface." He feels that interface design encourages programmers to think that a user interface can be slapped on for superficial aesthetics after they write the bulk of the code. Interaction design is much deeper and starts with the user's goals. Interface design is about making a feature look good; interactive design starts with figuring out the user's needs. Good interactive design results in easier-to-use products that meet consumer needs and cost less to create. Many non-programmers create computer products, often in the guise of spreadsheets and databases and should be concerned about the design of what we produce.

Cooper has been on both sides of interactive design. He is a programmer, and his projects include SuperProject and Microphone II for Windows. He is the father of Visual Basic and now heads Cooper Interactive Design. He has the utmost respect for programmers, being one himself, but points out that they rarely make good interactive designers.

While Cooper is not a Macintosh fan, this book is of value to Macintosh enthusiasts. The Mac rests on good design in hardware and software. Mac fans expect well-designed software with a superior interface. That is not to say that Macintosh software should slavishly follow the Macintosh interface guidelines, although we do prize consistency, but that it should be elegant and user-oriented.

Why Is There So Much Bad Software?

Why do we put up with poorly designed computer products? For the same reason we are impressed by a dancing bear. We are so impressed that the bear can even dance that we don't care if it can't dance well. Consumers put up with an awful lot of garbage because they are impressed that computers work, even if they are almost impossible to use. We either apologize for the sad state of affairs and defend the computer for what it can do (this describes most programmers and "power users") or just grit our teeth and carry on. Very few people try to fix problems.

Software publishers and developers usually deny there is a problem. Interface design is often an afterthought and done too late to change how the software or computer devise works with the user. Interactive design needs to be well under way before coding is done. Too often the interface is designed by people with no knowledge of interface design, which is why Cooper is trying to convince people heading companies of the need to hire interactive designers and make them important to the product design and development process.

Currently, programmers usually do design work, and it is a rare programmer who is good at both coding and interactive design. Cooper considers Bill Atkinson and Andy Herzfelds, programmers of the original Macintosh, to be "probably the two most talented, creative, and inventive programmers ever." However, Cooper feels that the General Magic device they created failed because it was "engineered and not designed" (emphasis in original). A professional interactive designer is needed for every project.

However, most programmers think they know best and refuse to bow to anyone's demands. Too often they design for themselves and not the typical user. They often love complexity and look down on people who do not. This works for games, where the user wants to discover the rules of the game on his own and enjoys being overwhelmed with choices, but not for someone who wants to set his or her alarm clock or write a memo. The most telling story is of the Amazon.com 1-Click interface. Jeff Bozos, Amazon's president, presented the 1-Click design to his programmers. The programmers insisted on inserting a confirmation dialog box, resulting in a 2-click interface. Even Amazon's president had a difficult time convincing the programmers to remove the confirmation box.

There is a need for professional interaction designers. Programmers are concerned that if someone else does interactive design, they won't have anything interesting to do. Cooper argues that interaction design is only a small part of software design. Interaction design is limited to the interface, so programmers still do a great deal of design work.

Another common problem is prototypes turning into products. Prototypes are only supposed be created for demonstration and proof of the concept. But sometimes they work too well, and someone decides they should become the final product. There are several reasons for this such as an executive deciding that the prototype is ready for commercial sales after viewing a dog-and-pony show. Often someone decides to base the final product on the prototype in order to "save time and money." In most cases, the prototype is made out of spit and bailing wire and does not have the right foundation or architecture for a full-featured system. In reality, it is cheaper and quicker to throw out the prototype and reprogram the actual product from scratch.

Which comes back to the one main reason interactive design is not done: people believe that there isn't time for design--which is often viewed as fluff--when in reality, a good design saves time and money.

The rush to produce software often results in substandard products. In most cases, we see badly designed software competing against badly designed software so that there is no lack-of-sales penalty for producing crap. Cooper argues that a well-designed product will beat a product that seems to dominate the market. Poorly designed software endears little loyalty, which is why Netware, a well-engineered product with a lousy interface, lost much of its market share to alternatives including NT and "even AppleShare." Apple, in contrast, was able to hold onto its many customers during its dark days out of loyalty inspired by good design.

How to Design Good Products

While Cooper did not write a "how-to" book, the last part of the book describes his approach to interactive design. He says that it is important to design for the user, but there is a problem. You never know who your user is. Rather than spending endless hours and dollars on market research, he recommends creating a detailed persona and designing for it.

A persona is a hypothetical person. It has a name, family, and motivation. Cooper provides excellent examples from projects he has worked on including the Logitech Scanman. A general rule is, avoid designing for a broad audience and try to help solve a real person's problem. Witness: wheeled luggage was designed for flight crews but became a product popular with the general public.

Apple's iMac might be a case where designing for a narrow audience produced a successful product that went beyond the target audience. The iMac was designed for the first-time computer owner, yet most purchases are made by people who already own a computer. If Apple had tried to make a Mac model to attempt to please everyone, it probably would have pleased no one. Many design decisions (such as eliminating the SCSI, ADB, and serial ports) simplified the machine. Compare the number of ports on an iMac to those on the eOne, a Wintel iMac lookalike. Which offers a more confusing array of ports?

Designers should not blindly copy other designs. Many design conventions do not work well. Many users find dialog boxes annoying and often don't read the warnings they display. A test was done by taping $50 bills to the bottom of a chair and having a person sit in the chair and run a computer program. During the test, a dialog box appeared on the screen telling the person to take the $50 bill. In most cases, the $50 bill remained taped to the chair. Programmers love dialog boxes, but most users don't read them. Use things that work.

The key word is design. Don't just respond to a bunch of user requests for new features. Design a product that solves problems. This means that you must understand what the user is trying to do and the larger goals.

Apple and Microsoft

Inmates is not a study of the personal computer industry, although Cooper criticizes Apple and Microsoft. He only refers to them to show the importance of good design, and any discussion of the two companies is peripheral to the book. He feels that both companies discourage other companies from innovative interface design and then blatantly violate their interface rules in their own products.

Microsoft has a very costly design--or nondesign--process resulting from attrition. Good design eventually manifests in a Microsoft product after 3--5 generations. The source of the problem is that Microsoft is the home of arrogant programmers. While Microsoft does employ professional interface designers, programmers are in control of product design, and one result is that products are designed without much concern for interactivity. Good products eventually evolve after a few versions have been released to the public. Microsoft can afford this costly and inefficient interface design process owing to its large bank accounts. Unfortunately, most other software companies emulate Microsoft's wasteful design process and often fall by the wayside.

Cooper has less to say about Apple: "And that devotion to design and attention to the details of interaction have created for Apple a customer loyalty that borders on--and frequently transgresses into--fanaticism." Otherwise, he feels that "Apple's technological prowess is good, but not great. From a capability point of view, Apple is no better than Microsoft in innovation." Apple's blunders of the past decade would have killed any other company, but its use of design earned it incredible customer loyalty. "The Mac fulfilled the user's needs only as well as Windows ever did, and in many cases less well, but the fulfilling of needs isn't the vital ingredient in market success." Apple's design-inspired loyalty caused many customers to ignore superior solutions. He feels that Apple is fixing its problems and will become a company worth paying attention to again.

This brings out the only contradiction in the book. If Apple offered superior interactive design, why does it not make, in his opinion, a superior computer? Cooper says Windows computers are superior to Macs in most cases. So isn't the superior design of the Macintosh worth something to him? Is Apple's good design only of a shallow nature, did Apple screw up so badly that great design doesn't cover its faults, or did Microsoft Windows eventually catch up to the Macintosh during Apple's bleak years? Or does this opinion reflect Cooper's background as a programmer? I could make a case that Apple's superior design did not make up for other deficiencies in the Macintosh, but he does not fully explain his case. While this issue is not important to his business audience, it is of interest to we Mac people. And while it is peripheral to his book, if he brings up how the Mac's design inspired user loyalty, he owes it to the reader to explain how the Macintosh's design does not translate into a better product.

I also think he misses an important point. Apple does not force any software company to follow its interface guidelines. The true force behind design consistency in Macintosh software comes from the users, and Macintosh users are more demanding of good interaction design than are Windows users. When a company makes a stupid change in how an application works, such as making -W do something other than close a window, I hear users complain. I think that Macintosh users, while accepting the dancing bear too often, are more likely to select software based on a superior interface than on what is popular than are Windows users.

So why does Inmates use the command cloverleaf () as a section separator? I guess it came from some Mac-loving book designer who wanted to show his or her blind, fanatical loyalty to Apple.

What of Consistency?

Cooper never directly brings up the issue of consistency. I get the impression that he looks at each software package on a case-by-case basis and does not value consistency of use. This is not a fatal flaw because a persona could be created that values consistency. But it is odd that he does not bring up this issue except in attacking Apple and Microsoft's' design guidelines. I wish he had addressed this issue.

For example, he praises Kai's graphic products because feels they address a very narrow audience that values complexity and the game-like attributes of functional discovery. I have heard people from this audience complain about Kai's idiosyncratic interfaces. People complain about how they force you to throw out everything you know about how to use software and not about their complexity or the sense of adventure in using them.


Inmates is an important book for anyone interested in computers. It is entertaining and enlightening. Since many Macintosh users value good design, it gets us to think more about how much better our software should be. I now have less tolerance for software that does not remember that I want my calendar to default to monthly view and not daily, darn it. At the very least, we can be more demanding about the products we buy and think about design more. For product creators, Cooper's approach is sure to result in superior products.

Cooper's book has made me realize how lucky I am to have worked for a company that designed products before creating them. I have worked on projects that were started before the design was even started, because management was in a hurry, at a huge cost and many, many missed deadlines.

Inmates is not an interface cookbook. Cooper's argument for good interaction is irresistible. Coupled with a clear, simple, and entertaining writing style, Inmates is must reading for anyone working in technology fields or interested in improving computer products. Despite being a "business" book, Inmates is extremely readable and valuable to consumers and programmers.

Paul Chernoff deals with everything computer related for The Washingtonian Magazine He is lucky to be doing most of his work on a Mac. He is thankful to have a wife who edits his articles. At home, he balances time between Mac and family. He can be reached at paul.chernoff@tcs.wap.org.

Return to electric pi

Revised January 9, 2000 Lawrence I. Charters
Washington Apple Pi
URL: http://www.wap.org/journal/