Abstract

This paper outlines the results of a review of the cataloguing operations at the Library at Curtin University of Technology and the three other publicly funded universities in Western Australia (known collectively as WAGUL). It identifies the cost savings to be achieved by re-engineering cataloguing and related operations, and separates them from the savings to be expected from outsourcing. The review itself concerned all four WAGUL libraries equally, however, this paper includes current operational experience from Curtin only.

Academic Library Cooperation in WA - Historical Overview (Williamson, V, 1997)

Western Australia's publicly funded academic libraries (those of Curtin University of Technology, Edith Cowan University, Murdoch University and The University of Western Australia) have a tradition of cooperation. This is exemplified in a number of innovative service arrangements, including a free reciprocal borrowing scheme, special arrangements for inter-library loans, and agreement on the use of JEDDS (Joint Electronic Document Delivery System).

This tradition of library cooperation was placed on a more formal footing in 1993, with the establishment of WAGUL (the Western Australian Group of University Librarians).

During the period March to May 1996 the University Librarians initiated a review of the functions of WAGUL. The review was designed to identify the extent to which WAGUL could provide beneficial outcomes for staff and students of the four universities. It was also intended to provide a tightly framed strategic plan for any continuing activity.

In particular, the plan was to provide the framework for directing the combined resources of the libraries towards accountable objectives. In pursuing this broad outcome, WAGUL identified three key result areas in which it needed to be successful in order to fulfil its role. One of these was to reduce overhead costs of services in the WAGUL member libraries.

Cataloguing Costs in University Libraries

The cost of cataloguing had been a concern for several of the WAGUL libraries for some time. The four libraries together spend over $2.5 million per year on cataloguing. Staffing for cataloguing and endprocessing represents from 13-15% of WAGUL libraries' salaries budgets.

The cost of cataloguing was also becoming an issue of considerable interest in the professional arena in Australia, with an emphasis on strategies such as re-engineering and outsourcing.

The Collaborative Cataloguing Project was initiated in August 1996 at which time WAGUL began work on a project plan. In November 1996, WAGUL called for expressions of interest from consultants interested in undertaking the study.

The New South Wales based UNILINC Limited was appointed as Project Consultant in January 1997 and the final Report and Recommendations were presented in April 1997.

Background to Study and Methodology

UNILINC Limited operates a cooperative cataloguing service on behalf of fourteen libraries. UNILINC staff had worked with member libraries since 1978 and hence had indepth knowledge of cataloguing and other operations across a broad cross section of libraries. This knowledge had been distilled into a detailed questionnaire of cataloguing and related functions developed for the purpose of measuring time spent on these activities. The plan was to use this instrument as a key methodology along with site visits where workflow in the four libraries could be analysed in detail.

As preparation for the study UNILINC received a background paper from WAGUL which outlined total salary and equipment costs for cataloguing as at mid 1996.

Two UNILINC consultants visited the four university libraries between 10 and 19 February 1997. One and a half days were spent at each. The consultants focused on workflow in the process of reviewing the cataloguing operations in each of the libraries with a view to making recommendations on the following options:

• outsourcing to an external agency or agencies;
• insourcing by one or more of the libraries;
• cooperatively managing the work; and
• suggested changes to existing individual operations that might effect savings in costs.

They also sought information to verify the baseline costs for:

• original cataloguing;
• copy cataloguing;
• database management including authority work; and
• processing or other related activities.

To assist in this verification of costs, each cataloguing staff member was requested to complete a detailed questionnaire whilst the consultants were on site.

In addition, each cataloguing manager was asked to provide more complete and up-to-date costing data than that provided in the background paper.

Verification of Current Costs (Wade, R and Henderson, G, 1997)

Cost per Title

The following analysis sets out to identify the costs per title catalogued and the costs of each of the major activities performed as part of cataloguing and related services. The cost data was taken from the WAGUL background paper and further verified with each of the cataloguing managers.

Table 1. Average cost per title

Salaries and oncosts cover the range of cataloguing and related activities including core cataloguing functions, endprocessing, training, management and liaison, as well as non-cataloguing activities performed by cataloguing staff such as reference desk work.

These average figures for cost per title in Table 1 mask the fact that it takes considerably less time to do some cataloguing functions than others. It generally takes less time to re-use a copy of a record online than it does to do original cataloguing. However, taking a copy of a record for a different edition than the book in hand (cloning or close copy) also takes more staff time than a simple re-use copy. Additional copies of a title already held and in the catalogue take less time than does the first copy. Some original cataloguing involves creation of a brief record and is less time consuming than full original cataloguing.

Weightings that reflected this variation in time had been developed from the commercial contract cataloguing work carried out by UNILINC.* These weightings were applied to the various types of cataloguing undertaken by the WAGUL libraries, with the resulting costs per title.

Table 2. Cost per title weighted

Time Devoted to Cataloguing and Related Activities

Staff were asked to nominate the number of hours per week devoted to a range of cataloguing and related activities, using the detailed questionnaire.

The responses were anonymous. Staff were asked to think about their average week. There was provision to nominate days per year for special projects or regular and periodic activities. Staff were asked to self-check total times to make sure they accounted for a typical full working week.

Eighty specific activities were itemised in the questionnaire including provision for non-cataloguing activities such as rostered duties elsewhere in the library. The cataloguing and related activities ranged from the initial searching, through a range of copy and original cataloguing activities, database maintenance, endprocessing, repairs, training, management activities etc.

The responses to these questionnaires provide an outline of the time devoted to the various functions as follows:

Table 3. Time devoted to activities performed by cataloguing and related staff

(1) Local system, ABN and others (2) Local system and ABN

Cost of Activities Performed by Cataloguing and Related Staff

From the questionnaire responses it was possible to cost the various activities performed by cataloguing and related staff as shown in the table below.

The amounts for each activity were derived from the hours per week devoted to each function expressed as a percentage of total hours itemised as set out in Table 3. These percentages were then applied against the total costs as identified in Table 1. For the salaries component, no attempt was made to cost the individual activities according to nationally set job classification levels or actual salaries paid at the various levels.

Table 4. Cost of activities performed by cataloguing and related staff

Cost of Cataloguing and Related Activities

From this data it was possible to provide a cost per weighted title which does not include non-cataloguing related activities such as being rostered on the desk, acquisitions work, web activities where they do not relate to the catalogue etc.

Table 5. Cost per weighted title (excluding non-cataloguing or related activities

Table 6. Savings from Suggested Changes (Curtin)

The data from the questionnaire provided clear insights into problem areas in each of the cataloguing operations. These problem areas were also brought to light in the interviews but not always. It took the questionnaire data to bring to the fore problems which may have been obscured by other more obvious problems.

Alternatives to Inhouse Cataloguing - Outsourcing

As part of the brief, the consultants were required to review current developments in the outsourcing of cataloguing in Australia and elsewhere.

Up until mid 1996, cataloguing outsourcing had not been a resource management strategy that had been embraced in any great degree in university libraries in Australia. This was in contrast to North America where there were many such initiatives.

The UNILINC member libraries had since the early 1980's outsourced their authority work and more specialist cataloguing to UNILINC but this was taken somewhat for granted and not emblazoned in the literature as a revolutionary approach to library management. The work is either done at UNILINC, onsite in the libraries using contractors supplied by UNILINC or by cataloguers (particularly new parents) working from home.

Reasons for the lack of interest in outsourcing of cataloguing in Australia up until the mid 1990's are numerous. Book vendors now active in the outsourcing area in the Australian marketplace believe that part of the reason lies in the requirement for the online addition of holdings to, and database maintenance of, the national cataloguing system, the Australian Bibliographic Network (ABN). Another factor that has delayed the adoption of cataloguing outsourcing at a time when the US library literature was alive with the topic, was the impending redevelopment of ABN. Many libraries put any decision-making about major changes to cataloguing operations on hold pending the outcome of the development of what was to be known as World 1. By early 1996, it was becoming apparent that this particular redevelopment would not be eventuating in the short term.

A third factor relates to how material is ordered in Australian libraries. One of the most publicised forms of outsourcing involves MARC data coming with the book where the book is supplied as part of a standard order. Standard orders and approval plans are widely used in North America but not in Australia.

A fourth factor, and perhaps the most overwhelming, is the use of customised classification in many academic libraries in Australia. This is possibly a result of specialist collections being relatively small and highly valued. Customised classification is thought to assist users in finding what little there is. Unfortunately, the moment copy cataloguing involves the review of classification numbers by the library, the benefits of buying copy cataloguing records begin to diminish.

Such is the pace of technological change in aspects of library operations, particularly cataloguing, the consultants focused on projects implemented in the past two years or about to be undertaken. Contact was made with the one total outsourcing project underway in an Australian university, that of Griffith University. Contact was also made with Arnold Hirshon, one of the original architects of the outsourcing operation well established at Dwight University in Ohio USA. The experience at the J. Hugh Jackson Library (Stanford University) as reported in the literature was useful because that library had implemented the type of outsourcing being planned at Griffith. To provide further international flavour, contact was made with Manitoba University in Canada which was well into the outsourcing process. The Manitoba experience was of particular relevance to Curtin as the two libraries used the same local library system, that of Data Research Associates.

From these contacts it is apparent that outsourcing was seen as having brought major cost savings to the libraries concerned, or is expected to bring such savings in the future.

However, none of the libraries contacted was able to quantify the cost savings derived from outsourcing as, in each case, the outsourcing coincided with a review or re-engineering process.

The difficulty of separating out the benefits derived from internal changes compared with outsourcing is problematic for any library looking at strategies for improvement.

Such are the changes in cataloguing in the last twenty years, particularly in the area of copy cataloguing, a library should be able to achieve well less than 20% original cataloguing. The WAGUL libraries only require 8-12% full original cataloguing ie. 88-92% is copy cataloguing, close copy or brief record input. As a result of these charges, any library that is incurring an average cataloguing cost per title of more than $15.00 should consider the possibility that it may have a problem. The question is how to tackle that problem. Does a library look to refining its existing procedures to make them as efficient as possible or does it look to outsourcing as the first and obvious solution. What are the cost-benefits of each of these options?

Cost-benefit analysis of alternatives to inhouse cataloguing (Wade, R and Henderson, G, 1997)

There are three options for cataloguing outsourcing. One is partial outsourcing whereby the copy cataloguing and endprocessing needs of a library are met by an outside agency, often a book vendor. The second is complete outsourcing. The third covers the outsourcing of specialist material or activities such as authority work.

Vendor Supplied MARC Records and Endprocessing (partial outsourcing)

An increasing number of book vendors are offering MARC records and endprocessing services with the supply of the order. Material received this way is often referred to as shelf-ready. The aim is for the books and other material to go straight to the shelf without any checking against the local system database.

From the experiences and expectations of those libraries contacted or followed up in the literature, it would appear that 60% of total cataloguing could be achieved by the supply of MARC records with copy data with the assumption that there would be no any checking of or amendment to the records. Endprocessing would also be completed for this material. In this scenario, original cataloguing is retained inhouse.

In the area of copy cataloguing it is estimated savings could be as high as 75% whilst 90% savings could be expected for additional copies. There would be a flow-on effect to searching as very little if any checking is done when books are received from the book vendor outsourcing service. Classification could be expected to be reduced for the shelf-ready material. Associated activities of management, training and liaison would reduce accordingly.

Table 7 sets out the savings to be expected from Curtin outsourcing copy cataloguing and associated endprocessing after making the recommended changes as already set out in Table 6.

The data used to specify current staff activity costs for each of the functions, derives from the Curtin data as set out in Table 4. Savings on equipment and cataloguing tools are included to the extent that these costs were included in the activity costs in Table 4.

The costs of vendor supplied services in Table 7 are taken from a publicly available price list. It is quite likely that WAGUL as a group could negotiate a reduction on the cost of records and endprocessing. It is also likely that WAGUL could negotiate discounts on the books purchased from the vendors. Such possible discounts or savings have not been taken into account in this analysis for Curtin.

It is also expected that there would be set up costs in any outsourcing initiative. As part of the costing, provision has been made for reorganising internal processes to ensure the best fit with changes ensuring from the outsourcing.

Similarly, in Curtin's case, changes would need to be made to the local system to enable copy information such as barcode, location and call number to be incorporated automatically in the MARC record.

Allowance has been made for management of the outsourcing contract/s. A considerable amount of time would be need to specify requirements and settle contractual matters. The contract/s once developed would require monitoring particularly with respect to performance guarantees covering delivery times, relevance of material, quality of cataloguing records etc.

Also all outsourcing projects reviewed in the study had incurred or anticipated incurring costs in correcting at least a small percentage of records loaded from the vendor. The need for corrections is often picked up through sample checking or through users reporting a problem in using the catalogue. It is assumed 5-7% of records received may require correction.

Table 7. Cost-benefit analysis of vendor supplied services after suggested changes (Curtin)

From this analysis, Curtin would save $1,298,516 over 5 years by implementing recommended changes and a further $503,380 by partial outsourcing. Curtin could expect its cost per weighted title to reduce after implementing the recommended changes and again after partial outsourcing from the current $21.90 per weighted title to $13.58 after implementing changes and $10.65 after outsourcing.

The pattern was the same for each of the libraries participating in the study. Overall the savings from changes to existing internal processes were estimated to account for 60-70% of total savings.

Complete Outsourcing to an External Agency

For the WAGUL libraries to take this option individually or as a group, it would be necessary to establish an agency within one of the libraries or have the library materials shipped by the various book suppliers and publishers to another location before they were received by the library. This would involve double handling.

To completely outsource all cataloguing could result over time in a loss of any indepth knowledge of MARC and other catalogue and database requirements, from the library. This would adversely affect the library's ability to work on issues relating to its local system, web catalogue issues and other aspects of information service.

Published papers are somewhat reticent about charges and post-outsourcing costs. Cataloguing outsourcing agencies too do not publish their full range of charges. Quotations are provided according to individual library requirements for classification, call numbers, ABN activity, subject analysis etc., and according to the mix of original and copy cataloguing and the type of material requiring original cataloguing.

Provided the WAGUL libraries effect the recommended changes for the individual operations and for WAGUL as a group, and take advantage of the expertise already available, it should be possible to carry out cataloguing for a figure of around close to $11.00 per weighted title.

It is unlikely that use of an outsourcing agency would significantly better this figure. As a result, complete outsourcing was not a recommended strategy for WAGUL libraries to adopt either individually or as a group.

Outsourcing Specialist Cataloguing

Original cataloguing

Specialist original cataloguing could be "outsourced" within the WAGUL group to take advantage of resident expertise. Material could be sent to the cataloguing library or the expertise could be contracted to work at the library which required the cataloguing. The aim of the exercise in the first instance should be to achieve cost and quality benefits similar to those already achieved by Library B.

It was recommended that $20 per title (not weighted) for original cataloguing be taken as an initial benchmark cost within the WAGUL group. There was room for considerable further improvement on this figure. This would come about as the overall operations become more efficient.

It was also recommended that quality to be achieved should at least include the checking of all access points against the local system catalogue, with new names, series and subject headings checked against the relevant external authority files.

Authority structure

Maintaining an authority structure is an activity which lends itself to outsourcing either within WAGUL or to an outside agency. Also, as authority work does not require the book in hand or access to many of the cataloguing tools, it could be carried out remotely. Provided there is online access to the Library of Congress and to each of the local systems, this work could be carried out anywhere.

There are agencies in Australia that already provide this service. Alternatively, it could be done inhouse by one WAGUL library or by an individual on contract to the WAGUL group. The benefit of using an agency is that they are set up to do the work and have procedures in place for monitoring quality and cost.

The recommended approach was to use an outside agency. It was estimated from UNILINC's experience that each year 15,000 new authority records could be input, augmented and checked and 50,000 headings across the four systems amended (using global change facilities where appropriate) to bring them into line with the authority records, for a total cost of less than $40,000. This would include the cost of staff-time, authority record service and other overhead costs.

The approach recommended in the study was different from that used by many libraries in North America which use fully computerised authority services. At a cost of 66? per title processed by WAGUL the cost of using such a service would be close to $50,000. This would not include originally created authority records as they are charged on a per quote basis.

The use of a computerised authority service was not recommended at this stage. The advantages of the approach as recommended, compared with the US approach are that the most relevant headings receive attention, Australian usage could be incorporated into the reference structure and authority records not available online would be created as part of the process.

Conclusion

At the end of the study, each of the libraries and WAGUL were in a position to evaluate a mix of local, cooperative and commercial strategies for reducing cataloguing costs whilst retaining quality.

They also had before them clear cost estimates of the impact of adopting these strategies.

The WAGUL Collaborative Cataloguing project offered a rare opportunity to review a number of libraries' cataloguing operations, at the same time, in considerable depth, using the same tools. The result was a naturally emerging and powerful benchmarking process. Its power lies in the fact that

• each library was analysed (and was seen to be analysed) in the same way and at the same time,
• the results from the questionnaire mirrored what internally gathered data there was and

This benchmarking may not have been an intended consequence of the study but it was a useful outcome.

The study has shown that it is possible to cost cataloguing operations and to do so in such a way as to allow comparisons. It has also shown that it is possible to estimate the likely cost outcomes after changes to internal processes and after outsourcing.

The results indicate that libraries should explore the efficiencies that they can gain before going down the path of outsourcing. Outsourcing of cataloguing adds something to the equation but not as much as may first be thought.

Confirming by Operational Experience

In March 1997 Curtin University of Technology announced a university-wide voluntary separation scheme. This was in response to funding cuts flowing from the Australian Government's decision to reduce funding to universities. The terms of the scheme were attractive. Responsibility for approving staff departures under the scheme was devolved to the appropriate Senior Administrator, in the library's case, the University Librarian.

By the time the consultants’ report was received by WAGUL, a total of 12 requests had been made by library staff to participate in the voluntary separation scheme. While requests covered most units across the library, by the far majority came from staff working in the Technical Services Division. Twelve positions were made redundant without reducing service because of the productivity gains expected to be achieved after the reorganisation of cataloguing as recommended in the consultants’ report.

Subsequently, the Technical Services operation at Curtin has been completely restructured, as recommended, with Acquisitions and Cataloguing being merged into one unit.

Curtin expects to have implemented all of its specific recommendations by the end of 1997. In 1998 a follow-up study using the same methodology is planned. This will serve to show how accurate the estimated cost savings in the study have been.

[Postcript: On 11-12 May 1998, the consultants returned to Curtin for a preliminary follow-up study. The new procedures had really only been fully in place since the beginning of 1998, so it was not appropriate to readminister the questionnaire except to ascertain the non-cataloguing time. The cost per title data was derived from the new though-put figures since January extrapolated for the full year. This was combined with the expected staffing levels for 1998 (less the non-cataloguing time) calculated as salaries and oncosts for the year, to develop the cost per weighted title.

From this data it can be expected that Curtin will have achieved a cataloguing cost per weighted title $10.11. This represents a 52% reduction in cost since 1996.

The consultants made six further recommendations for minor changes to procedures. It is expected that with the implementation of these further recommendations, the cost per weighted title would reduce even further. It is expected to be equal to or below that anticipated in the original report as pertaining after the implementation of the recommendations and after partial outsourcing.

Under these circumstances, the consultants recommended that Curtin consider very carefully the value of outsourcing. Should the cost of the record obtained via outsourcing exceed the 66 cents identified in the original report, there would seem to be little or nothing to be saved from outsourcing.]

Rona Wade
Vicki Williamson

BIBLIOGRAPHY

(1) L Parkington and G Talbot "Outsourcing cataloguing and physical processing: A Canadian experience at the University of Manitoba Libraries." As yet unpublished paper

(2) R Wade and G Henderson Report to WAGUL on the feasibility of a collaborative approach to providing cataloguing services for the WA academic library community. Sydney, UNILINC Limited, 1997 (unpublished)

(3) V Williamson "Cooperation and collaboration: the WA academic libraries collaborative cataloguing project." Paper presented at the CAUSE 1997 Conference, Melbourne, April

(4) K Wilson "Outsourcing copy cataloguing and physical processing: A review of Blackwell's outsourcing services for the J. Hugh Jackson Library at Stanford University." LTRS. vol 39 no 4 1995 pp359-383