Archive for March, 2009

The Great Unbundling of Content

Wednesday, March 11th, 2009

There is no question that Web searching has changed the business of publishing and marketing content. The old watchword among publishers that “Content is King” has been replaced by the web2.0 phrase “The User is King”. Just as iTunes allows music buyers to obtain a song unbundled from a record album or CD, so too does Google enable any Web searcher to very quickly identify and download articles on nearly any topic of interest from any of thousands of sources.

Google and iTunes weren’t the first to deconstruct publisher “packaging”. The unbundling of full text content first emerged in early online services like Dialog and LexisNexis. Initially, this was a great deal for publishers, since a single article from a newspaper that originally cost 25 cents on the street could be sold thousands of times online for as much as $3 dollars a pop. In effect, researchers in the early days of the online industry were willing to pay a premium to use technology that could help them find that “needle in a haystack” article they needed. However, the rapidly declining cost of technology, combined with publishers’ desires to use that technology to better serve their audiences, have driven the price of an individual article, in many cases, to zero or close to it. Publishers still sell “bundled” content –print journals and institutional subscriptions are still alive and well– but the increasing number of consumers have been trained by Google to think that many articles can be had for free.

This ability to obtain an individual article, song, or other piece of content has broad implications for both users and businesses. For users, while they are no longer required to buy more than they may actually want, they also lose the opportunity for discovery. How many times have you bought an album or CD because of a popular single, only to realize there are lots of other great songs now in your possession. iTunes and the like would argue that they create additional avenues for discovery through their ‘related artists’ and ‘recommended songs’ capability – and we agree, but more on that later. For content owners, this unbundling trend is of course very serious as it can cannibalize the sale of journals, books and CD’s. The strategic question therefore is whether to ride and somehow maximize this unbundling wave, or resist the wave and potentially cede the end-user relationship to the platform providers, in this case Apple and Google.

Publishers for the most part have already begun preparing as if this process is irreversible. They are implementing new ways to replace revenues that are lost when customers no longer buy the whole ‘package’. This new repackaging can take many forms. Subscriptions to popular online journals, such as the New England Journal of Medicine, increasingly offer unique features, including email alerts, user communities, access to unique data sets, and multimedia content designed to make the value proposition of buying the “whole package” greater than the sum of the parts. Obviously, this can be an expensive proposition and requires technical expertise for adding and managing new features on the Website – not necessarily a core competency of scholarly publishers. Likewise, iTunes, Rhapsody and other online music sites also want to encourage larger transactions. As mentioned above, they are leveraging ‘related songs’ technologies and recommendation engines to encourage more discovery and therefore more purchases while still allowing the user to maintain their sense of control in determining what they want to buy, i.e the soft sell.

DeepDyve is jumping into the fray by offering publishers other ways to “repackage” their content that are very easy and inexpensive to implement. Similar to the features at the music sites, DeepDyve’s More Like This technology enables a user to discover related articles. The technology takes the contents from a single article and uses it to reach deep into a publisher’s archives to find additional articles that represent a more complete offering of the publisher’s work on the topic. Publishers benefit when more of the right content is presented to a prospective customer. More Like This can also be used across whole collections of journals, and often enables users to discover otherwise hidden relationships between subjects in different disciplines.

Because DeepDyve is not dependent on any meta-data or taxonomy, the implementation of this tool is a snap. More Like This functionality is now available to any Website as an API – by simply copying and pasting a simple programming script a publisher can now, in a sense, rebundle an ad hoc “journal” from any set of content that is customized for the user’s needs at that moment. Rather than fight the momentum of unbundling, it’s possible that tools such as More Like This and ‘related artist’ will introduce a new means of bundling – one that is aligned with the evolving needs of the user.

The Future of Blogs

Monday, March 2nd, 2009

Occasionally, we will be inviting noted experts in the field of search and information to guest-blog for us.  Today, I’m pleased to introduce our first such entry from Joseph J. Esposito who also serves as an advisor to DeepDyve.

I was very pleased when Bill Park, the DeepDyve CEO, asked me to write a guest blog for DeepDyve.  It comes at a good time for me, as I have been blogging for a while now on a number of sites, principally Pubfrontier, but also at O’Reilly’s Tools of Change, The Scholarly Kitchen, and Teleread — blogging all over the place, but there is a feeling of same old, same old setting in.  It really is time to reinvent the blog, and DeepDyve is the place to do it.

No, DeepDyve is not suddenly going to make me smarter or any easier to get along with, but DeepDyve could bring its technology to bear on blogging to give us something new–and, as the B-school types like to say, something “value-added.”

Here’s my problem with blogs: they all pretty much look alike, do the same things.  Blogs differ because writers differ, but the blog as a form, well, it’s a short piece with some links; a blog roll in the margin; an archive; and maybe a snapshot of the family dog.  You read, you click on a link, follow it to the target, and maybe come home again.  There is a lot of horsepower under the hood of computers these days, and blogs just aren’t taking advantage of it.

So here is what I would like DeepDyve to do:  I would like DeepDyve to embed its search bar into blogging software so that when someone writes a blog entry (like this one), it automatically generates a search using DeepDyve’s KeyPhrase technology.  So imagine that you are working in WordPress or Movable Type.  You draft your post, tweak it a bit, add a few snarky comments, and then click on “publish.”  But instead of the post going directly to the Web, it first passes through the DD search system.  The entire text of the blog is used as a DeepDyve search query.  The results that come back can then be posted alongside the entry itself.

Imagine a blogger working in the life sciences, for example.  Her post is about a recent news story about a connection between certain diets and dementia.  She clicks to publish and then sees her post next to a list of related articles and stories to further inform her audience.  The blog post, in other words, has been re-created as part of a network of information all interlinked by related concepts.  This DeepDyve blog “plug-in” doesn’t only dive deep; it also brings up a net of related resources.  Contrast this with an ordinary blog post, that relies on the painstaking work of the author to identify links.  Links should not be work; they should be automated.

This new kind of blog is something that pretty much only DeepDyve could build.  This is because unlike other search engines (or research engines, as Bill likes to say), DeepDyve allows you to use a query of any length.  Your blog post could be as long as a full-length article.  You couldn’t use such a long query with Google.  Try it and you will see.  Google puts an upper limit of 32 words on a query, and even queries that long either retrieve far too many pages to be useful.  But a DeepDyve blog would retrieve only the truly relevant material and put your post into the context of all other documents that resemble it.

Now this isn’t to say that DeepDyve cannot be a useful tool in blog writing today.  For example, I can achieve the same benefits of above by simply copying and pasting my entire blog entry into the DeepDyve search bar and running a query.  It will bring back all related articles and links that I can then incorporate into my blog as needed.  But if DeepDyve offered that seamless plug-in, well, that would make DeepDyve an essential and easy step in the blog-writing world.

So, Bill, let’s give it a shot.  It’s great that DeepDyve has pushed search technology to a new frontier, but let’s now have DeepDyve change the nature of communications about research as well.

Joseph J. Esposito is an independent consultant providing strategy assessment and interim management to the information industries. He has served as an executive at Simon & Schuster and Random House, as President of Merriam-Webster, and CEO of Encyclopaedia Britannica, where he was responsible for the launch of the first Internet service of its kind.