Search-engine optimization reshaped the craft of a good headline. Will Amazon's book promotions have a similar effect on novels?



We all know that people who run websites have a hatful of little tricks intended to give their sites more prominence in searches. (For an SEO primer, see here.) One of the curious things about SEO optimization is that it works by altering webpages so that they market themselves: that is, instead of creating ads external to the thing advertised, you re-shape the thing itself so that it's easier to find and more interesting and attractive to link-clickers. And if we can do it with webpages, why not with, say, books? Why shouldn't books undergo whatever ongoing tweaking they need to be as successful as possible?

Three years ago Steven Johnson wrote a piece for the Wall Street Journal in which he predicted that the writing and publishing of books would come to require something like SEO optimization:

A world in which search attracts new book readers also will undoubtedly change the way books are written, just as the serial publishing schedule of Dickens's day led to the obligatory cliffhanger ending at the end of each installment. Writers and publishers will begin to think about how individual pages or chapters might rank in Google's results, crafting sections explicitly in the hopes that they will draw in that steady stream of search visitors.

Individual paragraphs will be accompanied by descriptive tags to orient potential searchers; chapter titles will be tested to determine how well they rank. Just as Web sites try to adjust their content to move as high as possible on the Google search results, so will authors and publishers try to adjust their books to move up the list.

What will this mean for the books themselves? Perhaps nothing more than a few strategically placed words or paragraphs. Perhaps entire books written with search engines in mind. We'll have to see.

Since this piece was published, there's less general confidence that everything will be findable by Google (or its search-engine rivals) -- in part because of Facebook's attempts to create its own complete and walled-off internet ecosystem. But Johnson's prophecy that the digitization of books will change the writing and editing of them is starting to come true in some ways that he didn't quite anticipate.

For instance, this article in London's Telegraph reports that some novelists who hope to sell a lot of Kindle books are constructing their plots so that a major cliffhanger occurs about 10% of the way into the book. Why? Because that's about how much of a book Amazon allows readers to download as samples. These writers (there will be more and more of them) try to calculate precisely when to insert that Oh-my-goodness-what-happens-next moment, so that it occurs at just the instant when ...

End of this sample Kindle book 

Enjoyed the sample? 

Buy Now

An irresistible invitation.

I imagine another possible development. It's delightful how my Kindle remembers where I left off reading a book, so that whether I'm on the Kindle itself or my phone or a tablet I pick it up at just the right location. But of course that means that Amazon knows how far I've read into any given Kindle book, and therefore knows the precise point where I stopped reading something that I never finished. That's very valuable information indeed to editors and publishers (and of course Amazon is now in the publishing game).

Here's the value: if a significant percentage of readers are running out of steam at the same point in a book, then perhaps the text needs to be sent back to the author for tweaking. A second edition -- or a third, or a fourth: there's no necessary limit to the iterations -- can perhaps fix the problem, which can get more readers to finish the book, which can get the book higher ratings on Amazon, which can lead to higher sales. Unhappy readers of the first edition can be informed that there's a freely downloadable New and Improved Version, which may induce them to give the book another try.

In short, the "release early, iterate often" model of software development could end up being a big part of the future of book publishing. Of course, writers have always been responsive to their audiences: consider Arthur Conan Doyle, who got so sick of writing Sherlock Holmes stories that he killed off his beloved detective, only to be convinced -- by big bundles of cash waved in his face more than by the grief of Holmes fans -- to bring him back from the dead for many more stories. But these new developments promise something different: micro-marketing, even nano-marketing, and endless iteration. Future conversations about books may less often begin with "Have you read X?" than with "Which version of X have you read?" Which will be rather weird.

We want to hear what you think about this article. Submit a letter to the editor or write to