The Inmates are Running the Asylum: Why High-tech Products Drive Us Crazy and How to Restore the Sanity [Paperback]

The manner in which Alan Cooper points out problems with many high tech products is thoughtful and insightful. The book contains many descriptive examples and entertaining anectodes to illustrate the problem of "dancing bearware". His case for the necessity of "interaction design" is convincing. Overally the book is thought provoking and educational. So why only three stars?

His accusation of engineers being the root cause of the problem is badly misguided, with a silly generalization of programmers as a whole. I develop software professionally for a living, and I certainly do not consider myself or my peers "techno-jocks". I do not look down upon end users any more than I would expect an M.D. to look down upon me for lack of knowlege about medicine. In the organizations I have worked in, I have seen that developers have the task of interaction design UNWILLINGLY thrust upon them due to miserable product specifications coming from sales and management. I have also seen useless gadget features come from sales and management more often than from engineers. From my experience, these things alongside unreasonable project plans and "we can fix it later" attitude on the part of managers have resulted in awkward products many customers dislike.

Also, the book was too self-referential. In some portions, it appeared that the author was advertising his own company.

It's a shame the "inmates running the asylum" theme and self-advertisements were over-emphasized. Aside from these things, this is a good read for both high-tech managers and engineers.

In my experience with system design, it is rarely the engineers who add the "extraneous" features. We're a lazy bunch and like to design to spec. It's the non-technical people...the marketing department, the customer reps, who blather about the software doing this and that and the customer bites. The customer thinks they get all these great features, but when the technical folks try to explain why it's a bad idea, managment says "Just put it in, we already promised them."

Besides, who says you HAVE to upgrade?? Most people upgrade because they believe they need all the 'new features' the next version has. I'm sure you've realized that nobody is fixing bugs in these new versions...ahem..windows..ahem...

The most fundamental and consistent error throughout the book is the idea that usability, failure to meet requirements and lack of an adequate design phase are new phenomena, as consequences of this era's computer technology alone.

This simply isn't true. If it were books like "design for the real world" written by Papanek over 30 years ago would have been unnecessary, Three mile island wouldn't have happened, and no one would ever misdial a telephone.

Sadly Cooper does not present proper evidence for a 'new' problem, preferring an informal and anecdotal style and, in doing so, extrapolating his entire argument from false foundations. He also sees the need to invent a whole unnecessary set of jargon to use, with fairly woolly and subjective definitions.

There are constant inappropriate references and analogies to other forms of engineering (particularly building), their methods and traditions.

"In the industrial age, engineers were able to solve each new problem ... they made bridges, cars, skyscrapers, and moon rockets that worked well and satisfied their human users. .... But unlike the past [computer] things haven't worked so well. "
Is he implying there were no problems before? Tay bridge, Tacoma Narrows, Ford Pinto, Challenger shuttle, Soyuz-1 and Soyuz-11. All suffering from dangerous design flaws (and not isolated) and none of them had anything to do with computers.

By ignoring the reality of past and current failures in (non Software) engineering Cooper quickly leaps to the conclusion that we "... have encountered a problem qualitatively different from any they confronted in the industrial age".
Errr, no. One of the first things we learn in engineering is how much of our wisdom has come from analysing failure and disaster fully, objectively and with academic rigour.

"When engineers invent, they arrive at their solution ... [it] will always be a derivative of the old beginning solution, which is often not good enough. "
Eh? Brunel? Stephenson? Even Santiago Calatrava doesn't shy from the title 'engineer'. Even in our beloved computer field, engineers and scientists abound; John von Neumann, Berners-Lee, Wozniak and Jobs. All brilliant in their day, and derived of what exactly? Not only a questionable assertion but grossly disrespectful and immodest from someone whose claim to fame is prettying up other peoples' work. These were and are the Engineering geniuses, and Cooper clearly doesn't understand engineering enough to see the differences between invention, innovation and merely doing what the budget allows.

In terms of descriptions of what a UI needs to be to qualify as usable, Cooper totally glosses over important concepts such as context. He ignores any Cost-Benefit analysis of designing and building this "no-training, no-maintenance system", blithely asserting that achieving that software (mirage) will reap all rewards. No proof, again, let alone an attempt to prove it would even be feasible.

The problem in programming is not that programmers are ill equipped or unprepared to solve the problems (though some may be), it is that no-one is demanding it of them in a coherent fashion.

Programmers are still being pushed to add 'features' buttons, wizards, gizmos and gadgets of little purpose because marketeers know they need to be able to print it on the box, and that is needed to generate the revenue.

Some programmers have the mindset he characterises, they are hardly very influential. Lack of proper requirements gathering, design, and industry-wide experience of very late, swingeing specification changes cause the problems. Programmers aren't to blame, even anti-social ones, the marketeers aren't, or the pushy ill-informed managers, the customer isn't either, but, at the same time, we all are. What we see is the consequence of nobody really knowing what they want, still less clearly stating, but everyone wanting to stamp their influence on the end product. Nice conspiracy theory Cooper, but it is nonsensical in the real world.

All the evidence sadly refutes Coopers Business Case. Products which demonstrate brilliant consideration of their target users fail miserably to make an impact (or a profit).
Look at the few of case studies of his own consultancy work he is able to offer;
1. A piece of support software for Logitech to bundle with their page scanners. = Logitech got out of the scanner market some time ago, didn't help their sales obviously.
2. Drumbeat web authoring. Well reviewed in its industry journals but scored poorly for ease of use. Elemental Software was bought out by Macromedia, Drumbeat was discontinued shortly after.
3. His in-flight entertainment (IFE) system (Clevis, et al.) for Sony Trans Com. Bought out by a competitor, Rockwell Collins, 2 years ago. Their new IFE will now be run, in their words, "on an industry standard Microsoft windows platform", Coopers system is not their flagship at all.

Now I am not going to say I think Cooper's advise for UI design is poor, or that his design methodologies are wrong. I think he is right in most of what he asserts there. It is just all based on flawed reasoning and syllogisms, and furthermore, most of it is not ground-breaking or even new ... there are plenty of good books out there discussing usability and making recommendations which are far more realistic and thoroughly presented than Cooper's descriptions of how he runs his consultancy. And he still has to demonstrate examples of where applying those principles won't cause you to crash and burn.

Cooper is presenting arguments firmly directed at those who are outside this industry and relying on their ignorance of what goes on. He plays on Technophobia and peddles misinformation. He very cleverly characterises programmers as having something to protect and a reason to be put on the defensive by what he says, in doing so appears to be trying to pre-empt responses and criticisms from technically informed readers. This has (as can be seen from the mudslinging here) unhelpfully stifled debate on his assertions. As Cooper is clearly intelligent and experienced enough to be aware of the flaws I identified, I can only conclude he was having a wry giggle with this book.

The book's populist slant and claim to have "the solution" are very appealing to some, and almost guaranteed its success. Sadly, it contributes little of use to a known and serious set of problems.