There was a third book this year which describes a methodology for Interaction Design. It comes from the Guru of Goal Oriented Design, Alan Cooper. Unfortunately, it is not a "How to" book but more a plea and justification for adoption of software development processes which will lead to improved Interaction Design. Cooper's earlier book, "About Face" is on the UIDesign.net recommended reading list. Was its successor about to deliver sanity amongst the nonsense of software development?

[ Note from January 2000 : Over the course of the last 4 months, I have adopted the techniques of this book extensively into my own work. I have used the techniques at several large brand name clients all with excellent results. I have developed an approach for Goal Oriented, User Centered UML development which uses the techniques from this book and merges them with those of Larry Constantine's Usage Centered Design, finally tying in Rational Unified Process at the technical development level. As such I now consider this book to be one of the two most important books in the field published in 1999.]

The inmates are running the asylum

Why high-tech products drive us crazy and how to restore the sanity

by Alan Cooper, SAMS, 1999

I ordered this book, the moment I heard about it. A new book from Alan Cooper really was big news for UI Design. That was over 6 months ago. The fact that it took me more than 6 months to review it says something. As soon as it arrived, I realised that it was very much in the same mold as Don Norman's, "The Invisible Computer". So it ended up on my shelf to be read as a lower priority.

Cooper explains in the preface that this is not a "How to" design UI book because he had been persuaded that he needed to address a higher problem - the business case for improved UI Design. Hence, he has written a highly political justification for his own (and mine) occupation. So if you are a programmer out in the trenches fighting that battle against delivery dates, then this is not the book to help you.

Cognitive Friction

Cooper has been responsible for introducing several terms into the field of UI Design and he wastes no time introducing another one - Cognitive Friction. He starts with a series of anecdotes about the influence computers are having on other things. His conclusion that ultimately the computer is anything but invisible. Indeed, everything you cross with a computer is always a computer. These anecdotes serve as the justification for the rest of the book.

Cognitive Friction is a posh term for all that hassle and frustration that we all encounter when using software controlled equipment. Cognitive Friction is a measure of the difference between your psychic thought control (the mental model) of what's happening against the actual performance which naturally doesn't match. Experiencing cognitive friction makes people feel stupid or angry.

Cooper continues Fred Brooks like, through the woes of the software development process and concludes that the real miracle is that it works at all. He compares it to a dancing bear. The miracle is that the bear dances. However, if you needed a dancer, you wouldn't hire a bear. A great analogy. The term "dancing bear ware" is used throughout the rest of the book to describe bad software.

The next terms to be introduced are the Apologist and the Survivor. Two types of user, the first group will apologise for not being easily compliant with the Dancing Bear Ware. The second group merely try to survive the experience of Dancing Bear Ware with the minimum of impact. A survivor is the type of user who will resort to pen and paper just to get the work done.

The true cost of a Dancing Bear

The second section sets out to detail the true costs of bad software with poor user interaction. Chapter 3 is a competent summary of much of what Fred Brooks, Jerry Weinberg and Capers Jones have all said before but it needs to be said again and again and this time it comes from a UI perspective. Chapter 4 describes some typical Dancing Bear type software products. I felt there was a missed opportunity here to have said more about the adherence to standards and guidelines which often lead to many of the problems he describes. Chapter 5 spends a lot of time to make a single point - Design is the key element when trying to improve software controlled products. Without good design you have another Dancing Bear.

The cause of the problem

Chapter 6 makes the second and perhaps the most important point in the book. The industry is in denial. Denial that what it is producing is not good enough. This echoes so strongly the whole reason d'etre for UIDesign.net and I found myself in full agreement.

Chapter 7 introduces another term to the UI Design field and its perhaps the most controversial part of the book. "Homo logicus" is Cooper's term for that element of the population that just love logical challenges and thrive on using technically difficult tools. You all know these people - they are the ones who argue that the command line is the best form of User Interface. The "Who needs an IDE when I have a command line and vi".

There has been a recurring theme throughout the book that the problem with software products is that they are designed by programmers. At times this gets almost vitriolic and is bound to alienate a lot of readers. Chapter 7 seeks to explain why many programmers make bad designers. However, I found the argument narrow in its execution, focusing as it does on individual psychology and engineering construction issues and failing to open the debate into social interaction and process issues which are integral to the final solution.

That concludes the business case for better design. At 120 pages it is roughly half of the book. The second half seeks to give us the solution.

The solution

Part 5 is in some ways merely an in-depth marketing brochure for Cooper's design firm in California. Many of the case studies are available at his web site. However, it is worth reading because in amongst these sample cases, there are yet more gems of wisdom and more new language for UI Design.

Personas are Coopers term for what others might call Actors or Users. Cooper makes the perfect observation that you should not design software for everyone but merely for "someone". Why? Because that someone is almost certainly not unique. You use the Persona technique to accurately describe that imaginary "someone". Then you design a product to make their life easier. Simple! Brilliant!

Defining Personas is the key to User Analysis for mass market products or web sites.

We are re-introduced to Goal Oriented Design and Cooper observes that Tasks are not Goals and shows us how this is so.

He goes on to define Polite Software and explain why good products have Polite Software. Then he offers us the four types of Goal - Personal, Corporate, Practical and False. I thought that this was ground-breaking and thought provoking and I may well re-write my UI Analysis paper to better incorporate this thinking. This section offered us the first real in-depth "how to" stuff from this book. The Rules of Software Politeness coupled with the Goal Definitions represent Cooper's approach to User Interface Analysis. This really was practical help in a book swimming with superficial high level analysis.

Chapter 11 entitled "Designing for People" is a series of User Interface Design strategies presented in almost anecdotal format. There is no real catalog and you have to read carefully and take notes in order to extract out all of the points. Another term introduced here is Use Scenario which sounded like another name for Goal Oriented Use Case. The need for a Domain Dictionary and Glossary is emphasised and there is a good exploration of when to invent a new widget and how to challenge existing widgets to see if they will do the job.

Chapter 12 harps back to the "Design is the solution" theme and looks at some often proffered suggestions which don't really work. Some of this really needed saying, such as Usability Testing is a poor alternative to design, and Style Guides don't solve the problem. Visual Design and Focus groups also go to the knife. Some of this is very strongly worded. Its Brave!

Chapter 13 finally offers us Cooper's process for the management of a product from concept to delivery. If this chapter had a flaw then it didn't take into account marketing and the needs of the sales channel to provide a "complete product", each stage in that channel having to add some value to the overall completeness. However, from a developers standpoint it makes a lot of sense. The observation that the software industry could learn a lot from the movie industry which frequently delivers very very large projects against tight timescales to a budget with agreed functionality. If the movie business can do it then why can't we?

Indeed!

Summary

[Jan 00: Having started to use the material in this book in my own work, my opinion of it is changing. It really is a thought provoking book which alters the state-of-the-art in Interaction Design. I use Personas in my own work all the time. I now consider this an essential text. However, my remaining summary is undimished...]

It is fairly and squarely aimed at the Silicon Valley High Tech Product Hot House. Its relevance to corporate IT shops the world over is somewhat diminished. At times the language is vitriolic and antagonistic. The book alienates much of the audience that it seeks to influence - namely Silicon Valley programmers. Cooper should know better than this and his editor should have provided better guidance.

However, the book makes a good case and significantly extends the vocabulary of the Interaction Designer. Cognitive Friction, Personas and the Goal types are really profound improvements in the field.

Overall the book suffers from being a "bit of this" and "a bit of that". It isn't an in-depth "how to" book and yet it is often too technical for the business audience. I really struggled to decide on the overall rating. Like Being Digital, if it makes it onto Airport lounge bookshelfs as a paperback then it may well do the world a favour. Without that I'm undecided.

Recommendation

4/5 - If Cooper writes a "how to book" then it may well get 6/5 but this one just isn't deep enough.