Christophe did not only write a very thorough review of "The Design Of Business" by Roger Martin. He also gave me the book ("You have to read this."). The last time he did this, it was for "Business Model Generation", which I cannot praise and recommend enough, nor see the end of what I learned/I am learning from it. So, unto "The Design of Business".
As you already have a good idea of the book structure (if you do not, go read it there. Done ? OK !), I got the perfect excuse to go immediately to the "opinion" part of the review (which is fine for me, as its the part I'm the more comfortable with).
To begin with the bad, I'm not really convinced by what I think is the "case" of Roger Martin: to present a "Design Thinking" technic to be applied, at equal distance between artistic chaos and mathematical predictions. Where the book in good in many ways (which I'll describe in a moment), it fails (for me) at describing/identifying this specific approach in a convincing way. For most "vision based" books, it would probably have been a show stopper (or, as I'm almost unable to not finish a book, however bad it can be, enough to make it fall definitively in the "bad" category). Yet Martin's book keeps all its value despite this major flaw, for different reasons.
First, Roger Martin is a very engaging author: his book is as easy to read as a fiction, his voice is one talking from a personal experience which transpires every page, and his discourse is motivating, enthusiast and yet humble.
Second, the description that Martin gives of the state of the industry regarding process and design approach is nothing short of stunning (if not surprising): the systematic preference for reliable, well-known, proven approach over more daring or audacious one, the processes that kill innovation in companies that praise it, the focus on incremental improvements that evacuate all chances of a major breakthrough, described by Martin, contribute to create a vivid, laser-focused if quite negative picture of many companies today (including several I've worked for/with).
Third, if the concrete tools and technics that could be described under this "Design thinking" approach stay a bit blurry, the examples given (which composes most of the book) are really convincing, if not easily reproducible. At the minimum, they are a stunning proof that the sad industry state described is not irremediable at all, and a vibrant warning to all companies founders or stakeholders: the reliability pitfall is there, and if you do not pay attention, you'll fall in it, as have so many people before.
Fourth, the author manage in a (short) last chapter to give some advice, formally directed at those in "normal" companies, without the power of the CEO or Department leaders featured in his examples. Not everyone can compare himself (well his situation) to Steve Jobs or Jeff Bezos. This chapter is probably the best of the book, shedding some light on an intriguing but elusive method, making me regret he did not make it longer or more detailed.
In addition to all those points, and as a final argument for the yet unconvinced, Martin manages the "tour de force" to stay moderate even when advocating his approach. While he harshly criticizes the over-emphasis on reliability, far from rejecting it altogether, he recognizes it as a useful and necessary principle. He only promotes to balance it with more innovative, less proven ways, showing that his recommendations are not those of an extremist in a moderate world, but more those of a moderate man in a world that has becomes extreme.