Thursday, June 13, 2013

Do-they-or-don’t-they? Ex Libris & EBSCO Information Services; Content-Neutrality & Content Silos. A living example of why librarians should pay attention when warned of the dangers of content silos.

Photo property of 
The Orbis Cascade Alliance last month did a very important service to the library community and the scholarly & academic community at large by issuing a clear and simple statement and request to both Ex Libris and to Ebsco Information Services (EIS):  “We are writing to express our great concern about your failure to make EBSCO academic library content seamlessly and fully available via Ex Libris discovery services....We urge EBSCO and Ex Libris to quickly resolve this issue. Failure to do so is a disservice to your customers and the faculty, students, and researchers we serve.”  They made this communication, and all the followup communications from all parties, available for all to see. 

This is not the first time this topic has come up. The International Coalition of Library Consortia (ICOLC) issued a statement back in June 2010 stating that they “encourage online providers and aggregators to allow their metadata to be included in emerging discovery layer services on a non-exclusive basis”.  The International Group of Ex Libris Users also issued a petition outlining the same request. And the whole industry came together, through the Open Discovery Initiative, with a clear aim of ”defining standards and/or best practices for the new generation of library discovery services that are based on indexed search”. 

(NOTE: Before proceeding, it can’t go without note that I have twice served as the President of Ex Libris North America.) 

Analyzing the documents, statements and behavior in both marketing materials and from the Orbis-Cascade website, I think it appears the main points of contention between these two companies revolve around the following issues:

How many Discovery systems does a library need to buy and use, in order to discover the content to which they subscribe? 

Most librarians, and most discovery product suppliers, would say one.  Ex Libris certainly does.  EIS’s answer appears to be it depends.  Well, do-they-or-don’t-they?  It really depends.  Let me explain my interpretation of these documents.

EIS markets EDS, their own version of a next-generation, single search discovery solution, in the same category as Primo (Ex Libris), Summon (Serials Solutions) and WorldCAT (OCLC) all products which include an article index along with the discovery interface. There is extensive literature about the benefits of these next-gen discovery solutions, and there is great demand in the marketplace for indexed-based search products. Libraries can find numerous benefits using these options and they can and should evaluate and choose the product that best meets their needs. Most libraries expect, and select, one discovery service for the majority of their content for all of their users.

Yet, EIS’s letter to Orbis-Cascade outlines that their strategy is not to allow library discovery services to index the EBSCO content for the benefit of paying customers. They only offer one method to search the EBSCO content – using an API to their own (note those two words, they’re very important!) discovery service, EDS, which is sold and marketed separately.

What does mean for your library? Well, it appears to say that EIS does not believe in one discovery service, unless it is EDS.  That libraries subscribing to EIS content and using a discovery service other than EDS, will be forced to pay for and use a second, additional discovery service, EDS, in order to access their EBSCO subscription content.  So,  you can either choose to buy EDS or you can be forced to buy EDS if you believe, as Orbis-Cascade and other libraries worldwide do, that another suppliers discovery product gives your organization better value for your money, and you want to use that product to discover your paid-for EBSCO subscription content.

Indexed-search or API-based search?  

EIS, we also learn, believes that indexed-based search is just a bad method to utilize. Their May 8th letter to Orbis-Cascade states that their: "view is that an API solution is superior to a solution that relies strictly on metadata." No, wait. Do-they-or-don’t they? It appears their statements contradict themselves. 

If they don’t believe in indexed-based search, they are clearly in opposition to the vast majority of the profession who believe it to be a core component of modern discovery services. What makes this even more peculiar is the fact that EIS markets EDS as an index-based discovery service – containing both EBSCO content as well as external, non-EBSCO content.  In fact, they just recently announced a partnership with Gale in which they say: “EBSCO Discovery Service creates a unified, customized index of an institution’s information resources, and an easy, yet powerful means of accessing all of that content from a single search box.”

However, I’m sure their marketing people are smart people, so maybe, this isn’t really marketing confusion, but rather a calculated strategy to force their competitors into a “controlled” method for accessing EBSCO content? EIS provides a hint to the answer in one of their letters: Ranking. If your main interest is the content providers’ needs – not the libraries and end-users – it is vital to control the ranking. So, it would appear they’ve found a way to continue to control the ranking of their content and thus avoid content ranking being controlled by the discovery system of the library’s choice. 

And yet there is another problem. EIS participates in the NISO Open Discovery Initiative, which is tasked to help the industry create best practices and recommendations in support of index-based discovery. It is therefore surprising to learn now that a key member of the group is actually clearly opposed to indexed-based search. Now, this is all right – I believe that any and all arguments against the method of indexed-based search should be discussed and reviewed by this group.  However, that should happen in an open and transparent fashion. But as far as I know or can find, this representation of a core objection to the method of indexed-based search was not shared within that group. Should we then assume that because of EIS’s participation in the ODI group, we should expect a change in their corporate strategy in support of index-based discovery? Or is EIS telling us in their letter that the ODI group’s work is good for others, but doesn’t apply to them? Do-they-or-don’t they? Does this make sense to anyone else? Because it sure doesn’t to me… 

The Library, Content Providers and End-Users

EIS is apparently also declaring that single search solutions are bad for content providers. They say that giving users the option to discover content across traditional content silos is against content providers‘, as well as EIS’s, best interests.   

Now let me state that I fully recognize the great importance and the role of native interfaces to the specific content environments – which often include specialized search tools and user interfaces that best match the content in that “silo”. However, it is not an all-or-nothing case, even though EIS is apparently trying to portray it that way. 

Giving users what we all know they want instead of forcing them to only search in traditional content silos is bad for the content provider, EIS says. This is a such a troubling argument, yet it pretty clearly shows where EIS’s thinking is at (and I don’t believe it’s where many librarians thinking is at!). Even more worrisome, I don’t think this is what the majority of content providers think. So EIS is really doing them a disservice as well, because they are out of alignment with those vendors and they are marginalizing the content provided by those vendors.  They’re placing it in a silo and limiting its access, use and impact.  As libraries begin analyzing content usage statistics more fully, and base decisions on renewals on those statistics, content vendors are going to find themselves penalized by the approach of EIS if their content is only available via using EDS. 

But EIS’s priority seems clear - to ensure that users are not “led to believe” that they can get a better user discovery experience outside the content silos via a single indexed-search discovery solution. EIS is well aware of the overwhelming evidence that users want and need single search across content silos, or else why did they build EDS?? So why won’t they make access to all of their content and content enhancements openly available to all discovery interfaces?  It appears they don’t want the competition.  Instead, they want to take the position that their discovery interface preserves the unique values of searching using their native silo searches.  The reason for that is because they limit access to many of those unique enhancements of the content, to their (EIS) search or discovery interfaces. Which is really a big problem for librarians.  As a result, your library is, in effect, locked into a content silo, in which you can only access those content features if you use the products made available by EIS for that content silo.  If you believe that another discovery product better meets the needs of your users, too bad.  If you want their content and enhancements, forget your users’ needs; the only needs that matter are EIS’s.       

So, the bottom line?  It appears EIS does and doesn’t really believe in single search discovery products.  IF your library’s choice is EDS, well, then you can have it.  If you want to do discovery solely with another product, well  then the response appears to be; “Sorry – it’s our way or the highway.”  

As for end-users, well, again, EIS appears to be declaring that single search solutions are bad for them. Yet we now have more than a decade’s worth of overwhelming data about users’ needs and expectations, including the need to have the option (yes - an option) to go to one place, where they will be able to discover content across domains, products and content vendors. 

This is also a critical need of libraries that  face huge challenges in trying to make their users (particularly students) aware of the content for which the library subscribes and pays. These single-search, across content domains discovery tools are a critical component in today’s library technology, but they come in addition to, not instead of, native content interfaces. In fact, many discovery tools ensure that content presented as part of the single result list contains a back-link to the native interface. This is clearly what libraries and their users want: different ways to discover and access content across numerous content-silos. One does not come instead of the other. Rather they complete the spectrum of options and increase the visibility and discoverability of content, especially content that libraries select and for which they pay. The fact that we now have thousands of libraries around the world using single search systems is great evidence of this fact. 

What is indicated by EIS is a radically different understanding of what single-search solutions are all about and the purpose they serve.   Yet, EIS’s priority remains very clear - they seem to think they can force libraries and users, who subscribe to EIS content, into utilizing only EIS’s technology for discovering that content. Sounds like a silo to me.


EIS is also making some very vague claims about rights to content (and/or the lack thereof).  If you examine the second Ex Libris response (dated May 14th,2013) in the Orbis-Cascade list, you will see where Ex Libris very clearly states they aren’t asking for those rights.  In fact Ex Libris doesn’t even need them. In the vast majority of the cases, Ex Libris can get to that content directly through the primary publishers, often with the full-text for indexing purposes.  But this is another instance of do-they-or-don’t they?!?!  Because EIS is now apparently saying that, while they formerly had the rights, now they don’t.  Yet, they provided Ex Libris with content for indexing up until 2010. (Or maybe this agreement ending had something to do with the fact that they introduced EDS to the market in this same time frame?) If your head isn’t spinning yet, well then consider this.  They (EIS) just recently announced a partnership with OCLC.  That seems to imply they do have the rights to make that content available to other parties. So, which version of EBSCO’s marketing is true? Because it seems unlikely they both can be…

However, please investigate and decide on all of this for yourself (and for your library).  Look at the Orbis-Cascade documents and read them and then read between the lines to find the very telling insights that are shown about EBSCO and Ex Libris corporate thinking. 

I have often warned fellow librarians about the danger of content silos.  This appears to me to be a prime example of a very troubling strategy of one content aggregator trying to severely limit the choices and options of librarians. By limiting your library’s options in the discovery of the content your library pays for, by forcing you to use their discovery product, by restricting your ability to apply content neutral, impartial and fair ranking across the content domains. 
It’s about removing the checks and balances in your library’s supply chain.  If we allow this to happen we should expect further restrictions and limitations in the future, all of which are likely to carry severe cost implications for libraries as we see vendors solidify and then take advantage of their position in the market.     

In order to deal with this as a profession, we need to define the guidelines under which we’ll buy products and services dealing with content, content enhancements, and discovery services. We have the annual conference of the American Library Association coming up in mere weeks.  I suggest we encourage these two vendors in particular, as Ex Libris has suggested in part of their correspondence, to have an open meeting and discussion, with invited representation from their mutual customer base, so that we, as librarians, can make sure that whatever is agreed to there actually serves the needs of our organizations, profession and users.  Do-they-or-don't-they? The evidence seems pretty clear but here's a great opportunity for them to clarify.