The questions here are as follows: Does EBSCOhost display a surfeit of "exclusive" titles? Are questionable business practices underway with the potential for great harm to this segment of the library market and online industry? Is EBSCOhost (or anybody else, for that matter) going about acquiring titles and then pressuring publishers to yank those titles off of other hosts? After all, if you paid kilobucks for a year's access to a title list and those you particularly need suddenly vanish, do you need to go elsewhere or purchase access to a second service? What about embargoes? What if the title still appears on the list, but you and your patrons don't realize that "there" means only articles from a year ago or earlier? Will you or — heaven help us! — your end users even notice? 

To check out the facts, this writer obtained the latest available lists of holdings for EBSCOhost, InfoTrac Web, and ProQuest Direct and compared them with each other. We also compared the lists against the database of Web periodical archives that go into Hermograph Press' book Net.Journal Directory and its online version, Net.Journal Finder. Even between Net.Journal print editions, we had noticed that something was up with EBSCOhost because its regularly received files had suddenly gotten larger. 

First, though, some definitions. The word "exclusive" has a connotation — "not available anywhere else" — i.e., the synonym of "unique," but that's confusing here. Let's try these:

Unique, for the purposes of this article, means Periodical X appears on one of the three academic market services, and only one. It may, though, also appear on a commercial search service like Lexis-Nexis, Dialog, or Dow Jones, where anybody with a checkbook or credit card can access it.

Exclusive means Periodical X only appears on one of the academic services and also not on any of the "commercial services." This is the more restrictive case and the librarian's ultimate worry, particularly since services marketed to academic settings usually have strong licensing clauses restricting distribution outside the institution, nor do they have any pay-per-view options.

So is EBSCOhost locking up titles "uniquely"? The numbers of titles all these services list is large, but we cut it down significantly by only considering full text (in any format — HTML, text, PDF, etc.). We did not include titles carrying only citations or abstracts. We also eliminated any titles that act more as newswires than as magazines or journals. Bye-bye ANTARA and PR Newswire.

After this pruning, we ended up with the following: 4,039 titles on EBSCOhost, 3,602 on ProQuest Direct, and 2784 on InfoTrac Web, using our definitions. How many titles do we find listed by all three services? 827 titles. Titles appear on two of the three academic services from 391 times to 632 times, depending on which pair of services one examines. 

Now for the big answer — drum roll please! — for how many titles appear uniquely on each service:

• ProQuest 1,742 out of 3,602

The two main protagonists arguing on the listservs, EBSCOhost and ProQuest, are running virtually neck and neck, with EBSCOhost taking over the top spot after many years as a near-cellar dweller in the standings.

Look at it another way. Fifty four percent of the titles on EBSCOhost are unique to it; that is to say, they only show up on this service and NOT the other two academic services. Sounds pretty nasty — until you look at the rest of the numbers.

• ProQuest 48 percent unique

This is hardly a monopoly situation. This isn't The Wall Street Journal only on Dow Jones Interactive. Competition seems to have survived.

What this resembles most closely is the online universe situation of about 10 years ago, pre-Internet. At that time, you needed unique proprietary software to access any particular service. If you wanted something on Dialog, you had to get an account and learn their software. Want something else on Lexis-Nexis? Write another check and receive another diskette package. Oy! It's only on Dow Jones News/Retrieval? Where's that checkbook?

The real lesson here remains that librarians who need access to everything in the academic universe stillhave to fork out money for all three accounts. If you hadn't realized that before, it's not because EBSCO suddenly began locking up academic ink. It's been that way for quite some time.

 

EBSCOnfusing EBSCOhost is the long-standing aggregator service for the academic market. It comes from the EBSCO Publishing division of EBSCO Inc. You pays your fee, you gets your mass of journals. You can get the whole package of thousands of titles or smaller subsets, depending on your needs and wallet. It has no transactional pricing — leave your American Express card home. EBSCO Online was begun by the EBSCO Information Services division of EBSCO Inc. as a tool to manage access to publishers' electronic archives for their journals. If you have a print subscription to the Journal of Biology, you can get access to the electronic version using EBSCO Online. EBSCO simply took its subscription agency model for print subscriptions and applied it to the burgeoning set of electronic archives, strictly a middle-man position. But then EBSCO took it one step further. They made arrangements with some publishers to allow access to EBSCO Online users on a pay-per-view basis. You whip out your credit card and download the article — but —you must have at least one regular EBSCO Online e-journal subscription in order to have access to EBSCO Online at all and to access the pay-per-views. You can not come in with your credit card from outside the system, ala Lexis-Nexis or Academic Press' IDEAL.  There is considerable sharing of titles between EBSCOhost and EBSCO Online, but one — EBSCOhost — actually has the data in its computers, the other — EBSCO Online — is just a sophisticated portal between you and the publishers.

Buy or Die?
So all the players have "uniques." But what about "exclusives"? In other words, can you or can you not escape the teapot-only barriers of subscription services and buy a teacup of articles from a commercial service when you need to? Suppose you run a small library and can only afford one service, or maybe two in a good year, but never all three. Can you get access to journals not found on your service of choice only when you need them by going elsewhere? 

We separated out the titles unique to each service and ran them against the listings in Net.Journal Directory/Finder for the three main commercial services, Lexis-Nexis, Dialog, and the Dow Jones Publications Library. On each one, you can get an industrial-strength account or wave a credit card at your monitor and buy pay-per-view. 

One should point out certain factors here. ProQuest Information and Learning has a long-standing relationship with the Dow Jones/Factiva group, in which Factiva serves as the corporate market partner. So a lot of ProQuest unique titles ought to show up on Dow Jones Interactive. Gale Group has long-standing relationships as a supplier to Lexis-Nexis and Dialog; in fact, Dialog and Gale Group are now subsidiaries of the same parent company, Thomson. 

Meanwhile, EBSCO has made a conscious decision to withdraw from other services. You won't find its periodical files on OCLC any more. Its Collectanea service is mere history. One might expect fewer titles shared between the commercial service listings and EBSCOhost. 

The table below compares how many of each service's Uniques (only available on one of the three) appeared on a commercial service. For example, 15 percent of EBSCOhost's unique titles also appeared in Lexis-Nexis, so they were classified as unique but not exclusive. The other 85 percent of the titles are, of course, exclusive — only on EBSCOhost, period.

At first glance, what we could have anticipated from the established relationships between aggregators and commercial services seems to be true. Those titles unique to InfoTrac or ProQuest can often be bought elsewhere (though clearly not every unique title). On EBSCOhost, though, you have at most one chance in eight to find any full-text periodical unique to their collection on any one commercial service. I also ran just EBSCOhost's unique titles against the combined total title lists of the three commercial services. Four hundred seventy two titles appeared on the commercial services, though obviously some titles show up on more than one service. This means that a total of around 25 percent of the unique titles can be bought pay per view on a commercial service, or conversely, 75 percent of the titles remain truly exclusive. One does not have to run title lists from ProQuest and InfoTrac to know that the unique titles from these aggregators will appear more frequently on the commercial services than EBSCO's. ProQuest and Gale sell titles anywhere they can; EBSCO does not. Whether the exclusivity policy at EBSCOhost in relation to distribution through commercial services stems from contract stipulations with publishers worried about digital cannibalization of print subscriptions, numbers alone cannot determine. 

One other note. Harvard Business Review may be exclusive to EBSCOhost in the academic market, but you can still find it on Dialog and Lexis-Nexis and Reuters Business Briefing — but not Dow Jones Interactive (though that could change if Factiva ever completes its long-promised merger of the two services). While whipping out the university credit card may not be an academic librarian's preferred option, it nevertheless leaves HBR a non-exclusive offering outside the academic market.
 
 

Long, Long, Long Lag Times
One final complaint was raised. ProQuest Info and Learning also claimed that EBSCO was embargoing periodicals for as much as 2 years. That means that although EBSCO may have a run of titles going back who cares how far, the most recent "as much as 2 years" would not be available. A long lag occurs between the latest issues out in print and the latest issues available online. For libraries going from paper to paper+digital or digital-only, this may or may not become an intolerable situation. It all depends on whether you have the stomach to tell your patrons you can't get anything more recent than a year or more ago.

I asked each representative of the three aggregator services if they embargo. All said they do. ProQuest and InfoTrac both claimed they have very few embargoed publications and none with lag times as long as EBSCO.

EBSCO supplied me with a listing of embargoed titles and their time lags. The listing varied slightly from the listing of full-text titles I had received 2 weeks earlier, which I have learned to expect from rapidly moving online services. I did not "clean" this file of 5,093 titles, which was a few hundred more than the original file sent to me. Nevertheless the conclusion shouldn't differ. There are only 932 titles with embargoes in the periodical (non-newspaper) listings. That's 18 percent of the titles. Of these 932, the embargo is one year for 588, or 63 percent of the embargoed listings and 11 percent of the entire periodical list. Most of the rest are 6-month embargoes and only 52 titles (1 percent) have embargoes that extend more than one year. 
 

Good Business? The Executives Speak
"It's restricting access," says Vince Price, vice-president of marketing at ProQuest Info and Learning. "The aggregation [of titles] gets fractured and this will raise the cost structure."

The arguments of Price and Alan Paschal, CEO of the Gale Group, run something like this:

EBSCO is making outlandish offers to publishers to move all their titles to EBSCOhost, paying three, four, 10, even 100 times as much as the usual fees. Often the money paid exceeds revenues publishers receive now from all sources. Many publishers simply can't pass up a windfall like this. It protects their bottom lines in shaky times. Meanwhile, there's an embargo on the new holdings so those publishers can maintain their print products, causing libraries to continue print subscriptions and not go wholly digital. And the contracts extend from 1-5 years. This also protects the publishers, though some publishers undoubtedly don't want to play this game, believing in the current model of spreading themselves around to have a variety of inbound cash streams.

Price and Paschal maintain they would rather compete not on exclusives, but on quality and indexing and value-added. For example, they point out that each of their systems is codified and indexed with many more fields and keywords than EBSCO's and supports more sophisticated searching. Price says EBSCO has four index points while his ProQuest products offer 19 index points.

Both executives of the competing aggregators believe that this policy of EBSCO's will cause 1) publishers to have higher expectations in the future for money they can get from any of the three players when existing contracts come to an end, 2) initiate a bidding war for exclusives that will ultimately raise the costs that libraries and end users have to pay, 3) that libraries choosing just one service because budgets are too tight will be unable to get access to all the titles they need for their patrons with the missing titles unavailable anywhere else. Price said, "There will be three systems at twice the price."

Though Price doesn't like such a cutthroat way of doing business, he says, "We're prepared to do this strategy. But it will make electronic collections more difficult and less stable. As the pricing war escalates, content will move around and prices will escalate." 

Finally, Price asks, "What does the customer really want us to do? Do they want us to go pursue exclusives? They may think so, but that's not what they really want."

Alan Paschal is a lot less concerned than Price. He even concedes EBSCO's plan is a good strategy for getting market share. Still, he states "We don't like them [exclusives]. [It's] not good for libraries and publishers. It forces libraries to subscribe to multiple vendors. [and] hurts publishers because smaller libraries can't afford the costs, so publishers can't get their brands out farther. It is like the escalating costs of major league baseball. The ticket prices will get out of hand."

He's not worried about competing against EBSCO, however, no matter which way the company wants to compete. "We can outspend anybody."

Paschal adds, "This is a short-term strategy. EBSCO can't continue to spend the money. It's a blip."

Paschal also thinks the strategy of embargoes isn't a good one. He claims InfoTrac has few embargoes and none as long as EBSCO. "These kinds of services are research tools; people don't read the journals online like they do offline." He doesn't believe putting the information online without an embargo should hurt print subscriptions.

Tim Collins, division general manager of EBSCO Publishing, responds that the arguments were grossly exaggerated, even comical. As to the contention that EBSCO's titles would be hard to access anywhere else, including pay-per-view systems, he says, "This was a conscious decision to focus on the library market." This is why EBSCO dropped Collectanea or EBSCO files on OCLC. Collins states:

We're making an investment in growth and to improve our product. But I've seen some statements that are gross exaggerations. It's not like we won a lottery. It's rather comical. It's not as extreme as I've seen out there. We have a number of titles that are unique to our service, but they aren't exclusive. It's not like we can tell them who they can work with. But they do have a long history of working with EBSCO Information Services.

There are a number of publishers that we have signed arrangements with that EBSCO Publishing [the parent division of EBSCOhost, separate from EBSCO Information Services that runs the subscription service and EBSCO Online] has not worked with in the past and these publishers are scholarly/academic people. But these publishers have never worked with an aggregator before. So we approached them to work with us in our databases on EBSCOhost and they've agreed to do that When you've never been in an aggregator database before and suddenly they appear in ours, they've become unique. 

Examples of these new EBSCO partners include the American Institute of Physics and Springer-Verlag.

What about the Harvard Business Review? Collins says, "HBR is a separate thing. They made a decision to work only with us in the academic market. We're obviously happy with that. It's not exclusive, it's just they chose to work with us and not with them [InfoTrac and ProQuest]."

And the complaints about raising the cost of access? "When you have a new product with a lot more titles, yes, it will be a lot more expensive. Of course a BMW 740 is more expensive than a Lexus, it's got more content. A product with twice the content, sure it will be more expensive.

As far as embargoes? "The database is not a replacement for a current subscription, this is the difference between a database and an online journal. Would the libraries want to have no online content or embargoed content?" Collins does allude to the fact that one could get the current content on EBSCO Online for those EBSCOhost titles also found there.

Collins concludes, "We're hearing on the listservs that EBSCO is trying to limit the access to information. I think we're increasing it. If titles have never been in an aggregator database before and now they are available in ours, that's a good thing, right?"
 

Conclusions and Predictions
Does this tempest in a $500M market teapot really boil down to something substantial? EBSCO came late to the aggregator field. It has always been the third player in this game. Now it has hit upon a strategy to build market share at the expense of the other two players. This corporate model of business may seem somewhat distasteful, especially to the academic marketplace, but it is legitimate, though it may come with some trade-offs. 

First of all, should ProQuest and InfoTrac decide to play the same game, Gale Group probably has the most capital to play with. ProQuest Info and Learning can dig deep into the past with all its microfilm and microfiche holdings which, old-tech or not, represents a major trove and a still-viable market. EBSCO's biggest information ace is its direct relationship with publishers and their print products. Past trends have shown print products shrinking in sales while digital products grow and that publishers have made more money by diversifying their income streams, not restricting themselves to exclusive arrangements. 

If EBSCO plans to buy market share by buying exclusivity with higher up-front fees, does it have the capital (even from E.B. Stephen's diverse company holdings, such as the real estate empire) to play this out for the long period of time it would take to compete against the other two players? Is it, as Paschal says, "a blip"? Can EBSCO build and maintain market share by paying higher-than-average fees over the long haul? Only time will answer this question.

At this point in time, however, the arguments about exclusivity are somewhat misleading. All three players have unique titles — among the three of them, in the academic market, and in roughly equal amounts. Perhaps EBSCO has suddenly upped its position in the threesome, but that's not necessarily a cause to cry foul. 

Nonetheless, evidence clearly demonstrates a rise in exclusiveness in EBSCOhost, when compared to the commercial field, that the other two aggregator companies do not have. Many of the titles EBSCOhost has unique to its library service cannot be found on the big three commercial services. This might be a real complaint...except I've found that many of the EBSCO exclusives are more academic in nature than one would expected for collections on Dow Jones or Dialog and Lexis-Nexis to a lesser degree. They aren't even on EBSCO's other service, EBSCO Online.

I ran the EBSCO list of uniques against the title list for EBSCO Online. As readers may know from my other recent article in the June 2001 issue of Searcher ("The Continuing Adventures of Scholarly Jones, Chapter Two: You'll Pay to View This, Doctor"), EBSCO Online, an access/management entry point for libraries to e-journal archives, has a moderate percentage of pay-per-view titles now (see the "EBSCOnfusing" sidebar on page 42). I ran the unique EBSCOhost titles against the list of titles for EBSCO Online. There isn't much overlap yet, but those titles common to both come from the folks at, for example, Adis International or Blackwell Publishers, Carfax or Sage or Lawrence Erlbaum, and many other typical STM (scientific/technical/medical) publishers. 

Conclusion: Many of the EBSCOhost uniques are those that one might expect ProQuest or InfoTrac to want in their services for the library market. If the deals EBSCO made to get the titles on their service involved exclusivity in the library market, one could see the advantage to EBSCO, though not necessarily to libraries or end user.

Only 350 matches occurred on that cross-list checking, not a very high percentage. The good news is that many of the matches did fall in the pay-per-view domain. If you have access to EBSCO Online and a credit card and need one of the 2,000+ unique EBSCOhost titles, you have a shot at getting them on EBSCO Online, though it will cost you generally as much as $23 an article. But, yes, you probably are locked out of them if you don't have EBSCOhost or a site package with the STM publisher's own Web archives.

In conclusion, EBSCOhost has found at least a short-term solution to building up its market by adding STM and other periodicals, evidently by making higher than average offers to publishers. This may not be the usual way of operating in the academic library market, but it's probably not the foul some have claimed. Librarians and end users depending on commercial pay-per-view services may be shortchanged by exclusivity, but the academic library service world already has to deal with masses of unique holdings within each major aggregator's offerings. It just may not have been so obvious.

Whether EBSCO has found a viable long-term policy will depend on how much money it wants to spend, whether Gale Group and ProQuest want to raise the stakes in this academic poker match, and whether EBSCO's policy sways more publishers to its side over the long haul. And that, in the end, will determine the real effects on libraries and end users.