My Job in 10 Years: Are A & I Services in a Death Spiral?

As I mentioned the other day, the most recent issue of ISTL is full of very fine articles.

The one that really caught my eye is the Viewpoints article Are A & I Services in a Death Spiral? by Valerie Tucci. It echoes a lot of the themes that I first wrote about way back in December 2006 -- that the traditional A&I services will have a lot of problems competing with services such as Google Scholar which are free to the user.

Here's some of what Tucci has to say:

Given all the changes what will the future bring for these services and how will it affect libraries, librarians, and users? I believe that the A&I services will follow the downward spiral of newspapers. However, some services will enlist the synergy of the death spiral in figure skating. With trust and communication the strengths of free and fee-based services will be combined. A partnership will develop in which fee-based services will adopt more of the artificial intelligence underlying free services while retaining the special features that make these services so valuable. For example, I predict various versions of the databases depending on the needs of the user with different interfaces (e.g., SciFinder and STN CAplus) and different cost structures. What I do not see is a growing demand for the fee-based products. I believe revenue will decline and publishers will have to find ways to replace this revenue. This said, I have to caution that an entirely new paradigm may evolve (e.g., Google Scholar charging for searching), or a disruptive technology will be developed which trumps all current search technologies.

I'm more or less in agreement with this sentiment. Expensive databases such as INSPEC or Compendex will be increasingly difficult to justify for strained library budgets. They're quirky and difficult sometimes, and you have to have good knowledge on how they work and are structured to get the most out of them. They're also really focused on top level research -- not the kinds of tools that undergrads will be using for their engineering design lab report.

The cost per search will be high, the groups that use it and find it important (librarians and some high-level faculty and grad students) will never use them enough to get that cost down much. And since they're databases that are served by a bunch of different interfaces, it can be hard to see where the real ground breaking innovation will come from. They're low-hanging fruit at budget time.

On the other hand, as far as I can tell, use for other databases such as Scopus and Web of Science (and SciFinder, to some extent) is still pretty healthy. They're run by large commercial organizations who aren't going to give up their market share and revenues easily and have the resources to back it up (yes, I said it, CAS is run like a commercial organization). They're relentlessly customer focused (if not perfect, of course), their focus on citations and other "vanity" products gets the attention of faculty.

They focus intensely and relentlessly on innovating their products (if not always successfully). And they market directly to faculty and research officers, driving demand from library users. It's a lot easier for us to spend big bucks if other on-campus groups are encouraging us with their support.

On the third hand, as Tucci states in her article, there are publisher and multipublisher article databases like IEEE Xplore or Scitopia or that also provide a way into content that will be good enough for a lot of users' needs. A potential new player is single library or consortial content aggregations. Ontario's Scholars Portal Journals project serves purchased full text journal content to all of Ontario's institutions of higher education, each institutions only seeing what they subscribe to. My institution's Scholars Portal Journals database has access to over fifteen million full text articles.

So, it's a race. A race between budget constraints, free services and "good enough" on one hand and innovation and marketing and value-added services on the other.

I think that ultimately Google Scholar and it's ilk will win the day and become the de facto academic finding tools, that they will add enough value to their free products to make subscribing to all those A&I services problematic at best. On the other hand, even in the 10 year time frame I wouldn't be too surprised to see one or two of the big products still standing, somehow having found a way to add real value to all the data that they have access to -- and maybe even some smaller super-niche products too. Publisher and consortial aggregations will still likely have a roll as hosts for content.

Of course, as I mentioned way back in 2006, Google or Microsoft could buy up all those A&I companies practically with their petty cash fund. And if Google can make us pay for Google Books Search, well, once they end up as the last academic search engine standing...

Interesting times.

6 responses so far

  • Sigh. I used to spend a lot of time in libraries consulting assorted databases back in the day. Most of the still-extant ones are accessible online now and a lot of the others seem like anachronisms these days. Such is progress, I guess.

  • John Dupuis says:

    Well, Romeo, some of those old databases are available for free online now (ie. pubmed) but most of the ones your library would have subscribed to aren't. And I'm not sure it's fair to call them anachronisms -- yet, at least. There are still a lot of researchers that get a lot of value out of them to write them off completely for the time being.

  • Jill O'Neill says:

    John, your expectation of a hybrid landscape of both free and for-fee services is pretty much what is already in existence. PubMed is free; Scopus and Compendex aren't.

    As someone who works closely with providers of both for-fee as well as free information resources, I see far more on-going product development and evolution in the field. Google Scholar is about brute retrieval whereas these other services are more about slice-and-dice precision applied to the retrieval. Tucci's opinion piece is just that -- an opinion piece that offers no substantive documentation to support her claim that A&I is in a death spiral. A&I will survive where it serves the need for a specific community of scholarship.

  • John Dupuis says:

    Jill, yes I do expect a hybrid landscape -- but one with many fewer players. The brute force method of GS just works for an awful lot of people, including an awful lot of faculty. I think you would be hard-pressed to find many faculty who don't use Google at least a little (and worth noting -- it's important to distinguish between what faculty say they do and what they actually do).

    I figure at least one of Scopus or Web of Science will survive, but unlikely both. Targeted, high value products like SciFinder have a good chance. Pubmed, of course, but it's free and would be much harder for google to unseat. Beyond that -- there are an awful lot of indexes out there that will be really challenged to add enough value to their offerings to make them better enough than the free alternatives to be worth paying for.

    How long before we see the transition, before libraries start to drop A&I databases en mass? At least several more years, maybe only starting to see real change in the 5-10 year time frame.

    These are tools we, as librarians, have a strong attachment to. We use them, we love to recommend them, we really see and understand the value they provide. The challenge will be in justifying their high price tags to our funders.

  • Passerby says:

    For all the talk of the demise of the abstracting and indexing for-fee services, the fact remains that outside of formal academic, agency and private library group access and individual subscriptions to these services, one doesn't have access to located content. Locating citations of relevant recent and older publications in an on-line search is only half of the object - securing free or low-cost access for those outside of subscriber groups remains as a substantial barrier to scholarly activity.

    As a scientist/engineer who grew to love and appreciate ready access to science abstracting/search services, I can tell you that free-access Google Scholar is a very poor replacement.

    Abstracting and indexing is only as good as the search term programming and citation indexing. This is not Google's primary business activity. The 'hit-rate' on any given search is about 60% of the available material because of inefficient cross-referencing of terms between journal sources.

    On publication access:

    With the exception of a few select Federal government agencies (and even then, outside access to journals may be restricted for security reasons if deemed to be 'outside of typical agency activity'), professional science and engineering Federal staff receive little more than lip-service with regard to journal abstracting and publication access. We are expected to remain 'technically proficient through scholarly activity' but are not allowed the means to do so, and that includes provision of funds to publish research other than internal project reports.

    If you elect to pursue required scholarly activity through private means (self-funded and on the side) as a career sabbatical, you really come to envy the relative ease of academic colleague access to A&I services and electronic and print material.

    If the US wishes to excel once again in science and engineering innovation, it might take a good look at at the cost of technical publication access for those outside of traditional sphere of research. Iceland offers access to select geological journals as free to its citizens and Canada also has free or reduced cost access to Canadian technical journals.

    The Open Access movement is still in it's infancy; however, it might get a real boost if the research products of publicly-funded scholarly activity were made available at very low annual access fee through the funding agencies.

  • red pepper says:

    Abstracting and indexing is only as good as the search term programming and citation indexing. This is not Google's primary business activity. The 'hit-rate' on any given search is about 60% of the available material because of inefficient cross-referencing of terms between journal sources.

Leave a Reply