The Content of Things, and SEO

Michael Jensen | 03/30/2009 | Digitization

SEO is the acronym for “Search Engine Optimization,” the technical term for “making your pages juicy for search engines.”

Search engines like Google are just programs that operate on sets of rules. These rulesets, and the program code that implement them, are generally called “algorithms”—and they determine the “relevance” of a document, in the eyes of the search results.

How does a search engine bring your book’s catalog page to the first page of its results?

If I could tell you with certainty, I’d be a jillionaire, but only for the next moment. Google’s algorithms are intentionally varied, week to week and day to day, in their constant struggle to do three things: improve the user experience, foil the cheaters, and make money for themselves and others with advertising.

Cheaters are the companies that build what’s been estimated to be more than a third of the volume of Web pages—pages that are produced just to promote other pages, or to just be magnets for pages with ads. Any joker with a bit of programming expertise can produce 10,000 pages in a weekend, each with 1000 words from the dictionary, and each with a bunch of advertising on them, or a bunch of links to a particular page to raise its “rank” in Google.

Google wants to ignore those pages—they have no value, irritate us, and make us try Yahoo instead. The challenge is that there are valuable pages that look kind of similar to the above.

Improving the user experience and foiling the cheaters are part and parcel: Google wants us to find the thing we want, the most truly relevant information, on the first page, ideally as the very first item, and not have a bunch of dreck returned that we end up going to. It’s in their interest to ensure this, to continue to make a ton of money from of a billion tiny transactions.

So how do we position our Web pages to be juicy to Google? By leveraging what we have: quality.

Google is unique in having Google Book Search to work with. What they have done is harvest the cream of the publishing crop, the stuff that was considered worthy of publishing risk, and considered worthy by libraries. That means they have millions of books’ worth of text they can process, to learn the language of substance and quality.

They know more than anyone about Web pages—their statistical analysis has a resource set unmatched anywhere else—and so they can learn a great deal about substance, quality, context, and purpose.

We, as generally small publishers, don’t have the resources to hire SEO experts, to hire Web specialists, to commission consultants. But we do have quality content, written by professional writers in one form or another.

Our quality comes in our use of language, in our titling of things, in our human involvement in the content of our books, our Web pages, our promotional material. And in our understanding of what constitutes significance, in the publishing world.

These elements are likely to be ever-more key, as Google refines its algorithms. It’s in their interest to locate our material when someone wants to find, well,  something like our material. It’s our job to be sure that we are “as much what we are” as possible.

This goes against traditional, mechanical “SEO,” I should note. It’s important to have key terms in your text, key terms in your HTML titles, links to other pages, links from other pages, more text than images, and the like….

But I’m trying to think two years down the road, when Google’s relevance systems are even more refined, and better informed than they even are now. In two years, they’ll have learned even better how to attend to grammar, and author intent, and user habits, and salesmanship, and mechanical SEO; they’ll have learned even better how quality material tends to structure itself, and point to each other; they’ll have learned even better what constitutes textual quality.

A catalog page isn’t just an advertisement for a book; it’s an invitation to humans—and importantly, to algorithms—to understand your intent. Humans want to know all you have to tell; algorithms are increasingly understanding what humans want to find, and how to avoid what’s been designed just for algorithms.

So try to use those elements of quality consciously, as you think about your Web pages. Try to include all the blurbs, reviews, and supplementary substantive text about your books, on the catalog page for that book. They were written by quality writers. Have a “sample” page that has much of the catalog page’s content on it, as well as a chapter from the book, to be sure you have substantive content for the algorithms to process. Give those algorithms what they will be increasingly be hungry for: quality prose.

Comments [0]

RSS Feed

Boosting the Canadian Books Catalog

Michael Jensen | 03/27/2009 | Digitization

Back in the early 90s, two colleagues and I initiated what would become the AAUP Online Catalog, beginning as a Gopher site and evolving into a Web page for books published by presses belonging to the Association of American University Presses.

The idea had many goals: collecting our diverse array of books in a centralized location would facilitate discovery (this was pre-Google); allowing Presses who did not have technical staff to begin to tiptoe into the Internet; promoting the scholarly output of the member presses; and lots more.

It was a success for its time—though there were substantial roadblocks, because most small- to mid-sized presses still didn’t have databases, or still were led by directors who hoped “this internet thing” was “just a fad.”

There was a small listing fee per book (under a dollar), depending on the amount of work required to standardize the data, and many presses couldn’t figure out if this was a marketing expense, or overhead, or what. Nonetheless, a majority of the AAUP membership had books in this “union catalog,” and traffic was substantial (for the time).

And then along came Amazon, and Google, and the desire for individual presses to have direct control over their own Web look and feel. The drive to have a collective presence waned, and the project was mothballed.

Interestingly, discussions are beginning again within the AAUP community about reviving it, especially as now most of the presses have fairly straightforward standards-based data feeds, cover images are already online, etc.—the whole process would be much more simple, and the benefits fairly clear. The scholarly publishing universe is fairly small in relative terms, and the kind of collective promotion made possible by a collective online catalog seem increasingly obvious to many—especially if the price is low).

The AECB publishes Rights Canada, its joint paper-based catalog, and has an online catalog/database. What its online version does not yet do (based on my explorations) is contain all the books from its member publishers, nor link directly to the publisher’s page, nor link to any “buy now” sort of tool, nor routinely include descriptive copy, nor present much in the way of “related titles,” except in one particular category.

All of these would take a bit more marketing attention from the member publishers, and would necessitate a different pricing model from the organization for online inclusion, not to mention some increased effort by the AECB.

Links to the publisher’s book-specific Web pages (provided by the publisher), and publishers’ active participation in enriching the online catalog, could encourage ebook purchases, and facilitate Canadiana in general. It would not replace anyone’s own Website, but could be a traffic-driver and promotion tool for every member’s publications.

Comments [0]

RSS Feed

Open Access to Francophone Developing Countries

Michael Jensen | 03/26/2009 | Digitization

Since 1994, the National Academies Press (who publishes the reports of the US National Academy of Sciences, the National Research Council, the Institute of Medicine, and the National Academy of Engineering) has made its 180 reports per year “open access” (in the sense of free, read-every-page online access), via the NAP web site.

Further, since 2004, we have been making full-book PDFs of all of our reports freely available (in English) to developing countries.

On the one hand, this seems absurd: why give away PDFs for free? Isn’t that ensuring that there’s no market for our publications?

On the other hand, it is very rational: we are required by our institution to be equally committed to *dissemination* as we are to *financial sustainability*—even though we’re expected to be self-sustaining through sales.

We have made our publications available to developing countries freely because we have control of our systems. We have servers we control (http://www.nap.edu); we have tremendous traffic (~1.5 million visitors per month) that we attend to on a daily basis. Because we control our own servers (as opposed to using a Digital Asset Distributor, or other aggregator), we can control the options available to users.

We use a service called GeoIP (at a nominal expense) to discriminate between developing and non-developing nations. We recognize, via the IP (Internet Protocol) address of the Web browser (and GeoIP’s database that correlates IP address to country), the locale of the requesting browser. If it’s in, say, Senegal, then we adapt our Web pages’ presentation in response to that fact. If the reader’s in Senegal, then we say (in essence) “You may download the full PDF of this publication, if you give us your email info.”  If it’s from France, we say (in essence) “PDF, epub, or Kindle, US $29.95.”

Part of the logic behind this choice is a rational response to that 50% “dissemination” mission we were given. The other is that we know that the developing world is a *really tiny market* for us. The cost of goodwill, in any developing country, is nearly as little as the cost of a space ad in a journal.

The lessons for Canadian Francophone publishers seem to me evident: take advantage of this period (2009 - 2011) to build brand, build goodwill, build a list of email addresses of people who care about your kind of publishing.

The economic drivers are clear: currently, scholars in most developing countries still have to pay for Internet access “by the minute” – either out of their own pocket, or that of the university. Further, the economics of the countries means that most book purchases are made by an individual, not a department or university; they are rarely affordable.

This means that, by hardly threatening our bottom line (since these countries provide 0.1% of our sales), we can, almost risk-free:

a) provide valued content to valued readers;
b) make some friends in terms of brand;
c) develop an email list of those non-US friends;
d) build a market for later e-formats, which some might pay for
e) virally promote our enterprise among those with whom this publication is shared.
f) encourage, if even slightly, purchase of a print book

The key, to my mind, is recognizing that the next two to three years are, in overall terms, *just the beginning* of a worldwide marketplace for ebooks. In that sense, our job as publishers is to nurture that market for the long term.

Canada, after France, is the country with the largest Francophone publishing enterprise in the world.

The overall French-speaking population is greater than 300 million worldwide, with only 60 million in France, and another 8 to 10 million in Canada.

What are the drawbacks to spending the next two to three years *building a market* for less than 0.1% of our potential income, by providing free access (with email) to readers in Francophone developing countries?

I think the drawbacks are small indeed, and the benefits legion. I’d say the same thing is true regarding English publications as well, but that’s a larger market, and might require more daring.

Comments [0]

RSS Feed