• Print

Getting Google to notice your ebook

3 book SEO tips and why metadata and book covers matter to Google.

TOC 2011Now that Google has launched its ebook store, it’s time for book publishers and authors to learn a little SEO (search engine optimization).

If you sell shoes or cellphones online, you already know SEO, or you’re out of business.

Until now publishers have played in a different pool — more Amazon/Barnes & Noble than Google — but Google eBookstore suddenly gives booksellers a reason to at least wade into SEO.

Because now if Google can find your book, it can sell your book, a few different ways: via its online eBookstore, or in apps for the Apple and Android platforms, or through one of its independent bookstore partners. (See also: How SEO relates to the Google eBookstore.)

But getting noticed by Google is not easy.

I discovered this when I tried to shelve my experimental ebook, “Ebook Publishers to Watch: 2011,” cover facing out, on the Internet. I put it on Amazon, I uploaded it to Scribd. I created a companion website, with a shopping cart. I sold a few dozen copies, pretty much right off the bat without doing much of anything. Not bad.

But I couldn’t help noticing that as far as Google was concerned I was barely visible.

No, worse than that. My ebook wasn’t even the top result on Google when I entered the exact book title into the Google search bar. It was beaten out by a mention of the book on TeleRead. That was sobering. I started thinking: Was I neglecting some obvious SEO techniques? Should I be choosing keywords? Optimizing chapter titles? Posting an ebook sitemap to Google and Bing? Are there emerging best practices for an ebook author?

Google’s book view

The good news is that even before the Google ebookstore, Google takes books very seriously. I heard this directly from Sergey Brin himself, in a conversation we had after a small press get-together last year. He told me that Google co-founder Larry Page has been interested in scanning and indexing books since the company’s early days. We also talked briefly about his footwear, although I digress.

More recently when I spoke to Matthew Gray, lead software engineer of Google Books Search Quality, he reiterated that visibility of ebooks and books is a priority for Google

“Our goal is making all the world’s information universally accessible and useful, and we believe that a lot of the world’s information is in books,” he said. “So it’s important to us to make that information available. If you need to know something about a disease, or a travel destination, there’s good chance the best information is in a book.”

For publishers who are members of Google’s free Google Books Partner Program — 35,000 publishers have joined since it launched in 2005 — every book’s content is indexed and made available in the search engine’s universal, blended search results. An Internet searcher can usually scroll through about 20 pages of text around the search result, depending on the publisher’s preference. Books that are out of copyright are 100 percent available to Internet searchers. For books with hazy copyrights — the books that are covered by the proposed Google book settlement — Google serves up much shorter snippets of text around the search result.

In all three cases, books are “discoverable” on the search service. For example, during the financial meltdown of late 2008, Internet users searching for “economic crash” discovered “The Great Crash of 1929” by John Kenneth Galbraith, first published in 1955, one of thousands of books on the backlist of Boston publisher Houghton Mifflin Harcourt. Normally, Galbraith’s book would have faded from public attention, but the book’s contents were a good match for the search query, and it benefited from the visibility on Google.

Ebook tools and best practices will be examined at the next Tools of Change for Publishing conference (Feb. 14-16, 2011). Save 15% on registration with the code TOC11RAD.

Metadata and market signals

But what about new books and ebooks? How does Google determine which new titles, and the more than 15 million books that have been scanned, float to the top of its search results pages: in the web search box and in the ebookstore.

The challenge, for Gray and other Google engineers on the Books project, is that the best known component of Google’s algorithm for determining the the value of a web resource — the number of links to it by others — does not apply to books and ebooks. Although it is possible to link to a selection in certain books on Google Books (here’s a hyperlink into the aforementioned Galbraith title) people don’t generally create links to the contents of a book or ebook. So linking is not a reliable indicator of quality.

My conversation with Gray took place before the launch of the eBookstore,
but it safe to assume the approach has not changed.

One strategy that Google employs is to tap into the book industry’s “rich tradition of metadata.” Gray mentioned that Google considers author blurbs, subtitles, synopses, reviews, and author biographies to be valuable sources of information about a book or ebook.

Today, much of this metadata travels with books in an ONIX file. ONIX, which stands for ONline Information eXchange, is the XML-based standard that contains more than 200 data elements: from author name and title, to book reviews, author photos, and excerpts. Google allows its book partners to provide ONIX feeds, Gray said.

Google also looks at what Gray referred to as “market signals:” how often a book has been reprinted, web searches, recent book sales, the number of libraries that hold the book, etc.

Google’s book search algorithm incorporates more than 100 “signals,” and those signals change constantly. The goal is that all these signals add up to a single, simple result: “That the best way to get a book ranked high on Google Books is to write a really good book,” Gray said

And when the signals point to an obvious target of a book search, Google Books now confidently displays one extra-large result: a super-sized book cover of the title. For example, if Google thinks your search terms indicate that you’re looking for Malcolm Gladwell’s “The Tipping Point,” it will feature that at the top of the page rather than any of the half-dozen other books with the same, or very similar, title.

Google Book Search result
Google uses signals to push obvious results to the top of its listings. In this case, Google assumes a query for “tipping point” means the searcher is looking for Malcolm Gladwell’s book rather than other books with the same title.

3 best practices for getting Google to notice your book

Despite all this signal sniffing, Gray said there are a few practices that authors and publishers can follow to increase the likelihood that a book or ebook comes to Google’s attention.

1. Use descriptive titles and chapter headings — Gray said an approach that favors “cleanliness of information” will make it easy for Google to find relevant content in a book and serve it up to the inquiring mind on the other side of the Google search box.

For example, if a book on the Internet has a chapter on the history of the web, it would be much better to use the title “History of the Web” than simply, “History.”

“We’re going to do the best we can,” he said, “but more complete chapters titles will help us out.”

2. Create quality content outside the book — The content you create around a book can also make a difference. As an example Gray cited the bestseller “Freakonomics,” by Stephen Dubner and Steven Levitt, pointing out that the first result from a search for the title is not the book, but the New York Times “Freakonomics Blog.” That comes ahead of the book’s web page, and the Amazon listing.

Presumably the frequency of the blog updates, the authority of the New York Times, and the number of inbound links to the column on the Times’ website boost the newspaper site’s ranking over the original book.

Yet the column’s prominence also helps the book’s ranking on Google.
As more book publishers use the web to augment the content between the covers, this kind of synergy is likely to become more common. “This blurring of the line between books and other content is something that I expect to happen more and more,” Gray said.

3. Book covers matter — One significant piece of metadata that Google has discovered, Gray noted with wry irony, is the book cover.

“A book cover is actually very rich metadata,” he said. “People associate a cover style with a particular author, or a particular series, and they respond to that. They recognize the cover and say, ‘That’s what I’m looking for’.”

Gray’s advice to book and ebook publishers: pay attention to the covers.
“We have observed,” he said, “that people actually do judge a book by its cover.”


tags: , , , , ,

Comments: 7

  1. The problem with Google’s continued reliance upon citation analysis is that it has been proven to be scientifically unreliable by the Math Union.

    So as Google expands its PageRank paradign into new search verticals without carrying forward the original IR-based scoring that made its classic Websearch so good, the quality of their vertical search declines.

    Metadata, as Google knows so well, can be inflated and manipulated — and essentially is in every aspect of marketing, including book distribution. So you’re saying they may be looking at and relying upon a signal that they supposedly don’t trust in Websearch.

    Maybe it’s time for Google to stop trying to prove that PageRank works (it doesn’t — otherwise, they wouldn’t have to filter out up to 90% of all Weblinks these days) and just focus on developing good content analysis metrics.

  2. While it’s still key getting Google to notice you with the new Facebook profiles you enter a book you are interested in and you can display that icon on your profile for everyone to see.

    The issue is that how do you influence these results as there is far less transparency than you need as an author.

    I agree with Mr Martinez it’s not best to rely on a progressively changing algorithm for a business model…

  3. I think Google has a long way to go insofar as getting their searches right for books or ebooks.

    For instance, if anyone is searching for my book The Joy of Not Working, the first Amazon webpage that appears is for an old edition of the book that is out of print. The same is true for my international best-selling book How to Retire Happy, Wild, and Free. This has been the case for The Joy of Not Working since 2003 when the latest edition was released. I think that people at Google should have got this right years ago by having the Amazon webpage with the latest edition appear.

    What’s more, Amazon compounds the problem. When a prospective customer arrives on the Amazon webpage for the old out-of-print edition of The Joy of Not Working, there is nothing on this Amazon webpage indicating that there is a newer edition of this book that is still very much in print and for sale by Amazon. I have trying to get this rectified for years by both prodding my publisher Ten Speed Press and by directly notifying Amazon. There is no doubt in my mind that this has cost sales of thousands of copies of this title over the years.

    At least for How to Retire Happy, Wild, and Free, I was able to get Amazon to indicate on the webpage for the old out-of-print edition, that there is a newer edition of the book available for sale on Amazon. But this took over 3 weeks of my e-mailing Amazon to as many e-mail contacts as I could find.

    In short, these are two obvious book marketing problems that both Google and Amazon should have rectified a long time ago, wouldn’t you say?

    In fact, I am amazed that people such as me should even have to point out these problems given how much high-priced help they have hired. It is because of problems such as these that I will rush in to have my How to Retire Happy, Wild, and Free or 15 of my other books published (which have sold over 650,000 copies worldwide) as an e-book by either Amazon or Google.

    Ernie J. Zelinski,
    Author of How to Retire Happy, Wild, and Free
    (Over 125,000 copies sold and published in 9 languages)
    And The Joy of Not Working
    (Over 250,000 copies sold and published in 17 languages)

  4. Good points on the change-ability (there must be a better word) and non transparency of Google’s algorithm. It’s definitely not worth betting a business model on it. And it is frustrating trying to figure out how much attention to spend on SEO. But there is an emerging consensus on the rock bottom basics that are worth paying attention to. Google’s new guide is helpful as a primer. http://bit.ly/fRjrhO

  5. Jacques St. Pierre

    The author said above: “… For books with hazy copyrights — the books that are covered by the proposed Google book settlement — Google serves up much shorter snippets of text around the search result.”

    The author is in error, probably because he relied on information from Google and Google has been lying outright. Of the 14 Million+ books scanned by Google in the book-scanning project, the vast majority are in-copyright and Google knows it. The scanning project is the biggest single book piracy project in history, includes millions of criminal copyright infringements. There is nothing “hazy” about the copyrights; except for the tiny number of out of copyright books, ALL the books illegally scanned/stored/etc. by Google are clearly in copyright and Google’s action is clearly both criminal and civil infringement.

    Why aren’t Google executives in prison?

  6. What stands out for me is I have to rewrite my book chapters to accommodate Google’s way of thinking. There goes my creative and well thought out chapter titles.

    I am not sure I like the increasing need to do everything the way Google says lest we not be noticed or found. I am all for some conformity but this is starting to get to be too much–and considering Google hasn’t been around long, I can see that we are really leading up to a day when we all are seeking to know and understand the truth according to the Google bible. IN many ways, we are there now.

  7. Google has so many features now that don’t seem to get much press. I’ve heard of some of its product and comparison features, but I this is the first I’ve heard of something specifically for books. Is this one of those things that pops up in normal search or just when using the book search? And how many people actually use that part of the search engine?