Of all the book publishers out there, one of the most experimental and interesting is O’Reilly. They do many smart and audacious things.
Recently, in an experiment that was inherently appealing to me, Todd Sattersten began writing a book entitled, “Every Book Is a Startup.” The book is being published in small chunks, with each available via e-reading (ePub, Mobi, or PDF) at a low price. The premise and promise is that feedback from early readers will shape subsequent and more complete editions, until a “finished” book is achieved and printed.
It’s an interesting approach, one that Sattersten compares in an interview to the agile approach to product development startups prefer, in contrast to the waterfall approach of traditional publishers:
For a startup, the first iteration of the agile process is the minimum viable product. What is the absolute smallest feature set that can be introduced to the market so that the company can gather feedback from real customers? The initial release emphasizes learning and iterating, adding what is needed as the customer base grows.
Sattersten also notes that book publishers “launch more products than any industry on the planet and somehow we can’t seem to learn anything from that.”
However, as befits a 0.1 version, Sattersten has internal conflicts to his arguments, some of which are glaring. Fundamentally, the mistake he’s making is to believe that a form will easily yield to a process.
Sattersten notes that authors often use the agile approach to writing. This is their process. Authors us an agile process, and have done this for decades, by first publishing short stories, serializing chapters, or blogging their content, taking the feedback and ideas from those activities, and using them iteratively to create a final work. Other forms are used to accomplish the agile process.
It’s not an insignificant fact that most authors work this way. The only ones who typically don’t are those who write on spec. But most writers typically publish early drafts in one way or another, and use feedback to hone future elaborations, revisions, combinations, and extensions.
Books don’t work this way. Books are a form of finished content. Authors can be agile, and historically are. They have had many ways of being agile, and blogs have added one more. Books are expected to be fairly polished and final expressions of an author’s work, whether that’s an e-book, a paperback, or a hardcover.
Returning to the notion of the “minimum viable product,” we again encounter Sattersten’s assumption that a short PDF or e-book is a minimum viable book product worthy of purchase. The problem with this is that the book is a finished form. When you call something a book, a complete book is the minimum viable product in the market of books. A few blog posts or short essays strung tentatively together is not.
Authors and publishers have some time-honored and some new ways to be agile. Agile is a process choice. Conflating the notion of agility with the book form is not something that seems likely to succeed. Write a few blog posts. Publish one of the chapters as an essay in an online magazine like Slate. But don’t try to sell me your incomplete book.
That’s not agile. That’s clumsy.
7 Thoughts on "Treat a Book Like a Start-Up? Only By Confusing Process With Form"
Kent, I am surprised to find you, of all people, objecting to a paper form moving in the direction of digital fluidity. Completeness is not a feature of thought, so why should it be a feature of expressed thought? The concept of being a book may be about to change, because it now can. This sort of concept change is a standard feature of technological revolutions. Is Wikipedia not a book?
I’m all for digital fluidity. My complaint here is that we have ways of being fluid already, so why conflate that with a non-fluid form? Especially as a sales approach.
And I don’t think Wikipedia is a book. It’s too big and too capable of evolving. It’s something new, something else, and maybe something better.
Wikipedia is exactly what the designers of the encyclopedia had in mind, a basic compendium of the world’s knowledge. (I used to collect and read encyclopedias, plus teach the history of science and technology from them). Wikipedia is Hegel’s dream (he pioneered the science of concept change). But you think we need a new word for it. We don’t and this is concept evolution. Wikipedia is just an encyclopedia.
The funny thing about language is that it tries to keep up. How else can we say what is true?
>But don’t try to sell me your incomplete book.
>That’s not agile. That’s clumsy.
I believe that a lot depends on the kind of incompleteness — on where the “holes” are.
I believe that a minimum viable book should covers the basic “arc” of the message. There may be some examples, references or more elaborate explanations missing, but the book contains the core message. In other words, it has a “beginning”, “middle” and “end”.
A minimum viable book may lack some editing and polish, and one could expect a few typos here and there, but not to the point of distracting from the message.
I am experimenting with this idea myself, and wrote such a book. It’s 72 pages instead of, say, 200 or so pages for a “proper book”, but it covers all the bases – and it’s free (may be that’s one of issues.) Based on the feedback, I believe it’s succeeding. I am getting great feedback and ideas and people are getting the core message of the book.
I’d would have been happy to provide a link to my minimum viable book here, but I have refrained from doing so because I don’t want people to think that I am writing this post to market my book. I just wanted to defend the “Every book is a startup” idea because soooooo much time and paper is wasted in writing and printing books for which there is no market.
Having said that, Kent makes several good points, and I believe that the level and type of incompleteness is key to what would be acceptable and not taking advantage of the reader.
Thank you for an interesting counterpoint,