In this article we will first explain how we came to our decision to focus on delivery instead of worrying about discovery. Then, we will describe what actions we undertook to prepare ourselves and our users for a new situation and we will show what obstacles we still have to overcome to facilitate delivery. Finally, we will elaborate on the reaction of our users to the first steps we made in this process.

The aim of this article is to encourage you to rethink and evaluate your own efforts on discovery and delivery. We think that many of the problems academic libraries are facing can only be solved when we as libraries, publishers and vendors collaborate closely together on delivery issues.

Like every other library, we had always offered our users a catalogue. In 2002, we built our own discovery tool, Omega, aimed at finding electronic journal articles. At that time, we were able to explain the division to our users: ‘Looking for print material? Search the catalogue'; ‘Looking for electronic journals articles? Search Omega’ (see Figure 1).

Figure 1 

Utrecht University Library website (2012)

Over the last few years, things changed rapidly. New commercial discovery tools such as Primo and Summon entered the library market and we lost our pioneering role.

Meanwhile, more and more users had been finding their way to our licensed journals through larger and stronger search engines like Google Scholar: freely available on the internet and containing massive amounts of scientific material. And our users also switched to databases we paid for, such as Web of Science and Scopus. Our statistics showed that the use of our library catalogue and Omega remained stable, whereas the use of our licensed journals was growing, indicating that our users increasingly approached articles through other ways than our Library discovery tools.

“… our users increasingly approached articles through other ways than our Library discovery tools.”

Figure 2 shows the search behaviour of Utrecht users over the years, relative to the situation in 2006. The number of searches in our catalogue and Omega (the two bottom lines, light brown and red) stays more or less the same over the six years, whereas over the same period, the number of searches in Google Scholar (the pink line) can be seen to be increasing enormously.

Figure 2 

User statistics Utrecht University Library (2006-2012)

From international studies, we have learned that students and scholars are moving away from library websites and online library catalogues. Most students start their search in a general search engine on the web and don't use the library website as a starting point. Scholars use specific research databases and they also use general search engines on the web.

Since scholarly information can increasingly be found via publicly accessible search engines (like Google Scholar) and via paid multidisciplinary and subject-specific search engines, and students and staff are increasingly taking that route, we came to some radical conclusions in 2012:

  • we could shut down our custom-made discovery tool Omega
  • there was no need to buy and implement another Utrecht discovery tool
  • we would focus predominantly on delivery.

“… most searches were started in Google Scholar.”

We decided to hold on to our Library catalogue for the time being because we felt there was not yet an adequate alternative.

Implementing these conclusions meant we had to rethink the role of the library in providing access to scientific information. As part of this process we did a user survey, interviewing researchers and students of our seven faculties. Most searches by far were searches for full-text articles and most searches were started in Google Scholar.

We also asked participants what they most needed from the Library to improve their information searching. The main points mentioned were:

  • information on which search engines were available to them
  • information on how to access full-text articles
  • tailored support when and where they need it
  • information on where to find e-books.

Before we could shut down our search engine Omega, we undertook the necessary steps to accommodate our users as much as possible. Firstly, we undertook a project to redesign our listing of available search engines, taking a Lean Startup approach with multiple prototypes that are offered to our users for testing and comments. This project is still ongoing.

Secondly, we more prominently positioned information on off-campus access on our Library website, and also developed a simple Javascript bookmarklet users can add to their internet browser. This enables them to log in with their Utrecht ID and password when they are on a page without access, without first having to go to the Library website.

Thirdly, we added LibGuides on the effective use of search engines and tools to our existing suite of LibGuides, and more prominently positioned our LibGuides on the Library website.

Finally, to improve the findability of e-books, we worked hard to include all available e-books in our catalogue.

Prior to switching off Omega in September 2013, we had to inform our users, explain our decisions and offer them alternative and better ways of finding what they needed. Because, despite the fact that other search engines had become increasingly popular over the years, we had to take into account that our user statistics still showed that approximately 1.2 million searches were carried out annually in Omega. To communicate the impending changes, we drew up an extensive communication plan.

At the beginning of 2013, we started communicating ‘better ways of finding’, using social media and other channels to try and reach students and faculty. We had chosen to make it a positive message, showing alternative and better search paths. As a visual reminder of the impending changes, we added a countdown on our website (see Figure 3), counting down the days until Omega would be switched off.

Figure 3 

Countdown on Utrecht University Library website (summer 2013)

Behind the scenes, we redesigned our Library website (see Figure 4). We removed the search boxes from our homepage and instead offered a new structure, highlighting the support we offer scientists and scholars in different stages of their workflow: literature searching, data and literature management and publishing. On the homepage, we offered easy access to some initial resources and outlined different ways to get support from the Library, both online and in person.

“Omega was switched off as planned … But almost nothing happened. There were no major complaints…”

On Sunday 1 September 2013, Omega was switched off as planned. For the first days, we were prepared to receive complaints, and we anticipated all kinds of problems. But almost nothing happened. There were no major complaints and we saw a lot of hits on the ‘searching for literature’ subpage of our website. One thing that did happen immediately on 2 September was an enormous increase in visits to Google Scholar via our proxy server. Google thought they were searched by a robot and showed our users a captcha to let them prove they were human!

Figure 4 

Utrecht University Library website (September 2013)

Looking back, a year later, we can safely say we didn't face any major problems. We also didn't receive any major complaints. Obviously, there were disappointed users, mostly humanities students and scholars, but we were able to set them on the right track again, pointing them to databases that are of specific use to the humanities.

When we look at our user statistics, we see a clear impact of the shutdown of Omega (see Figure 5). The use of Google Scholar to access full-text articles has increased sharply, and also the use of Scopus has shown a sustained increase.

Figure 5 

User statistics Utrecht University Library (2012-2014)

Phasing out Omega was a project for 2013; improving delivery is a continuous process. We want to support our users, who are increasingly out there, on the web, in the cloud.

Our efforts on improving delivery include:

  • adding holding information to discovery tools, for example upload 250,000 missing holdings to the Dutch national catalogue (and from there automatically to WorldCat)
  • offering our SFX knowledge base to Google Scholar so that Google knows to which journals the Utrecht users have access
  • offer our repository (with Utrecht dissertations, but also with digitized objects from our special collections) for harvesting through OAI-PMH
  • supply easy authentication for our off-campus users, promoting federative authentication through SURFconext, as it is called in the Netherlands.

“… we believe that the OPAC is dead.”

Since our focus has been on delivery, we have become more aware of how often things go dramatically wrong: links that don't work well, material that is not findable and users who don't know how to get access to our licensed materials.

We have now embedded our focus on delivery in our entire organization and work processes. We keep track of where our users are (by means of trendwatching, international studies, user surveys and also our own statistics). We co-operate with other libraries or groups of libraries to be more able to influence suppliers and vendors, publishers, to use international standards, to improve linking mechanisms and to offer smooth access to licensed users.

Our next challenge will be to phase out our catalogue as an end-user discovery tool, because we believe that the OPAC is dead. In the world we live in today, you should not encourage your users to start their search in a local library catalogue. The Utrecht holdings are part of the Dutch national catalogue and through that we are part of WorldCat as well. We are looking in this direction to see if we can give our users a sufficient alternative, an alternative that takes them quickly and smoothly to content, print or electronic, that we have on the shelves or we can give access to.