Jared L. Howland

Brigham Young University’s patron-driven acquisitions: Does it stand the test of time?

Jared L. Howland, Rebecca Schroeder and Thomas C. Wright

Citation Information: Howland, Jared L., Schroeder, Rebecca, Wright, Thomas C., (2013). “Brigham Young University’s patron-driven acquisitions: Does it stand the test of time?” In Customer-based Collection Development: An Overview. ISBN: 978-1-85604-931-3

Introduction

For many years, we at Brigham Young University (BYU) have used patron-driven acquisition (PDA) strategies in our overall collection development plan. This approach includes faculty and student research requests and patron suggestions. We also rely on our users to drive the purchase of additional copies of owned materials by buying books with multiple holds. In the last several years, we added an interlibrary loan (ILL) purchase program for items too new to borrow and ebook patron-driven acquisitions to our existing patron initiated purchasing plans. We use these programs in an effort to better utilize the acquisitions budget while still providing our patrons access to needed materials.

Brigham Young University

BYU is primarily an undergraduate institution with selected PhD and Masters programs. The Harold B. Lee Library, which is the library for all of campus except the law school, has an operating materials budget of $10 million. As with most academic libraries, the majority of this budget is allocated towards continuing subscriptions to academic journals and databases. Over the years, the pressure to maintain these continuations has eaten away at our monograph budget and continues to be a major issue as we seek more effective ways of providing the best information resources for our users.

BYU started using patron-driven models based on two realities. One is the continuing pressure on material budgets. We have basically maintained a flat budget for the last six years and have made serious adjustments to our purchasing priorities. The second reality is the declining use of our print monograph collection, a trend mirrored around the country. As a percent of the total circulating collection (2.4 million), use has declined from 30% to 16% in the last decade.

In 2009 we completed a retrospective ten year analysis comparing our monograph acquisitions to their circulation activity. The findings of this analysis were helpful as we examined our purchasing patterns. A book purchased in 1999 had a 50% chance of circulating in ten years. Put another way, of all books purchased in 1999 (55,490) only 50% had circulated by 2009. Only 34% of books purchased in 2008 had circulated by the end of 2009. Even after allowing for ten years, half of our books are not used. Perhaps a better model for acquisitions was needed.

There is another dynamic that is changing the way we look at our collections. Collection development has been transformed by the introduction of the network and ubiquitous computing to the academic library world. Whereas librarians previously worked from set criteria like authoritative bibliographies to build isolated static collections, we now have opportunities to share collections, turn access on or off at will, and decide whether we should own or just provide access to content.

Although there are many ways to measure PDA’s effectiveness, we began to evaluate the success of our various models in 2011 using a cost-per-use (CPU) metric. In that study we measured the use, cost, and CPU of each of our PDA models and found that, overall, using PDA models was a much more cost-effective way of acquiring library resources than the traditional just-in-case models.1

So it is natural that, as we continue to expand these programs, we wanted to have a better sense of their long-term success. Were the initial uses of PDA materials just that? Or was there an indication that the user-driven model is successful in predicting future demand as compared to the more traditional form of just-in-case collection building? This chapter revisits the previous study and uses that data to extend the timeline and track these purchases over time to determine if the initial success of the various PDA models has been sustained.

BYU’s PDA Models

BYU employs five PDA models to acquire monographs: faculty expedited orders, ILL requests, suggest-a-book online requests, multiple holds trigger, and ebooks. A faculty expedited order is simply the timely acquisition of a faculty request without subject librarian mediation. Faculty can initiate a request via a link on the library webpage. The library seeks to order, process, and deliver this book within ten working days. Since these acquisitions support the curriculum and research needs of the university, the books are charged to the appropriate subject librarian’s materials budget.

Suggest-a-book requests are also initiated through a link on the library web page. It differs from a faculty expedited book order in that a subject librarian mediates it. The acquisitions department collects the suggestions and forwards them to the appropriate librarian. If the librarian decides to purchase, the books are charged to his or her collection budget. This is a patron-driven process but there is no rush order or patron notification upon receipt.

Too new to borrow ILL requests are another purchasing method that fits the PDA model. Often, libraries are reluctant to loan newly published monographs. ILLiad requests for items in that category, and that fit other predetermined criteria, are rush ordered and paid for through a budget designated for this purpose. At BYU, we limit these orders by language (English) and price ($50 threshold for student requested non-fiction; $80 for all staff and faculty requests).

Holds placed on a book are reliable indicators of demand for that book. When patrons place multiple holds on items, BYU will often purchase multiple copies. This practice is librarian mediated after a threshold of three holds per copy has been reached. If the librarian decides to proceed, additional copies are rush ordered and, upon receipt, checked out to the next patron in the holds queue. These books are funded by a designated budget created for this purpose. These tend to be popular literature titles.

The electronic ebook PDA model is by far our largest patron-driven endeavor. A pilot was initiated in 2009 and we have continued to expand the scope. PDA for ebooks was introduced as a volunteer program wherein selectors were allocated funds outside their traditional budgets. Participation has increased each year and in 2014, all purchases will be integrated back into individual budgets so that all PDA ebook purchases will be funded entirely through selectors’ traditional budgets. This model allows us to load thousands of bibliographic records into our catalog that meet the academic profile established with our book vendors. This effort is coordinated through the library’s acquisition unit, the book agent, and a third party ebook vendor. The library’s approval profile is used to govern what records are placed in the catalog. Once loaded into the catalog, a purchase is triggered when a predetermined number of pages are viewed in conjunction with time spent in the book per use. BYU is not charged until these limits are reached for any individual book. Users view all these titles as part of the library collection. They are not aware of the purchasing process and the process is largely seamless to the patrons.

Methodology

For this study, we compared the cost, circulation or use, and the cost-per-use of titles in each of the print and electronic PDA models. For print PDA, we examined the titles that were either purchased or selected during fall semester 2009. We used reports from our Integrated Library System (ILS) to generate the title list and the cost of the individual books. To measure the use of each item, we counted each check-out and in-house use. The statistics showed us the monthly uses allowing us to measure the use for a full 26 months from the time a book was available to our patrons. CPU was calculated by dividing the total price of the items by the total number of uses they had. The mean and median were calculated from the CPU results. These statistics were then compared to a sample of traditionally purchased titles that were acquired during the same period of time.

The title list used in our current study was slightly different than the list used in the previous study. A few print titles have been lost since the previous study and were therefore deleted from the current study. Several other titles that had erroneously been included in the original analysis were identified and removed from the new study. Table 1 includes statistics describing the various samples used for this study.

Table 1. Summary of random samples
Total purchased (n) Total cost ($) Mean (median) cost ($)
Print
Faculty expedited 38 1,568.19 41.27
(32.50)
Suggest-a-book 33 1,215.03 36.82
(24.56)
Interlibrary loan 55 1,577.11 28.67
(21.60)
Holds queue 35 531.43 15.18
(16.19)
Traditional 217 8,199.54 37.79
(27.23)
Electronic
Patron-driven 324 24,828.92 76.63
(74.95)
Traditional 163 17,033.55 104.50
(90.00)

For electronic PDA, we measured the use of titles that were triggered and purchased during our PDA pilot. Our partners for the pilot were YBP Library Services (YBP) and ebrary and ran during the months of December 2009 through July 2010. For our study, we gathered the COUNTER statistics and the ebook cost from the vendor’s reports and website. As with the print measurements, the CPU was calculated by dividing the total cost of the books by the total number of uses. These statistics were then compared to a sample of traditionally acquired ebooks selected during the same period of time.

In comparing our previous analysis with the current usage and CPU of ebooks, we found that there were minor inconsistencies in the way our vendor made adjustments to their usage statistics. Another challenge that surfaced was that ebrary’s COUNTER statistics are full-text section requests, known as COUNTER report #2, whereas EBL numbers are for full-text title requests, known as COUNTER report #1. Therefore the EBL statistics are slightly lower compared to the ebrary materials. Most of the titles used in this study were ebrary but we did have a few EBL titles in the traditionally acquired sample. There is no way for us to adjust those numbers so we decided to keep the titles despite the inconsistencies on the basis that there are so few of them that the difference in the final analysis would be negligible.

Because of the problems described above, we recalculated the usage from the previous study using the newer methodology so that accurate comparisons could be made. This changed the numbers from what was previously reported but the conclusions and findings remain unchanged.

Results

On a strictly CPU analysis, PDA models are the clear winner. Table 2 shows the circulation and CPU data of each of the five BYU print PDA models. The models are listed in order of least to most effective, with faculty expedited orders as the least effective model followed by suggest-a-book orders and interlibrary loan orders. The most cost efficient model is the holds queue order. Our original analysis with four months of data indicated that the least effective PDA model had a 51.7% lower CPU than the traditional model. After 26 months of use, the least effective PDA model still had a 34.7% lower CPU on average than the traditional model and received an average of 168.8% more uses. The most effective PDA model had, on average, a 93.1% lower CPU after four months and a 95.4% percent lower CPU after 26 months and received 871.9% more uses on average than the traditional model.

Table 2. Cost and circulation data for print acquisitions
Total circulations Mean (median) circulations Mean (mediana) cost-per-use ($)
4 months 26 months 4 months 26 months 4 months 26 months
Faculty expedited 59 206 1.6
(1.0)
5.4
(1.0)
26.58
(22.69)
7.61
(20.00)
Suggest-a-book 47 221 1.4
(1.0)
6.7
(3.0)
25.85
(12.50)
5.5
(7.45)
Interlibrary loan 135 430 2.5
(2.0)
7.8
(3.0)
11.68
(12.88)
3.67
(6.24)
Holds queue 140 978 4.0
(4.0)
27.9
(30.0)
3.80
(3.60)
0.54
(0.55)
Traditional 149 703 0.7
(0.0)
3.2
(1.0)
55.03
(19.42)
11.66
(13.00)

aFor titles that had use

These statistics also show that findings from the previous study are corroborated and the longer books are available to patrons, the CPU generally decreases. This holds true for both PDA and traditionally acquired books. The change over time of the average CPU of the various print PDA models can be seen in the slopegraph shown in Figure 2. The most striking portion of this graph is the traditional acquisition model. The average CPU of traditionally acquired materials drops at a significantly faster rate than the PDA materials. This dramatic slope is exaggerated by the low initial use of traditionally acquired materials. The increased usage of traditionally acquired materials over longer periods of time explains why the average CPU for the least effective PDA model went from 51.7% lower to only 34.7% lower than traditional models.

Figure 2. Slopegraph comparing mean cost-per-use of 4 months of data to 26 months of data for various print acquisition models.

Figure 2. Slopegraph comparing mean cost-per-use of 4 months of data to 26 months of data for various print acquisition models.

When comparing acquisition models, the comparison of the number of titles that received use to the number of titles without any use is an interesting exercise. Almost all PDA titles had use, and their use increased over time. The suggest-a-book model had a higher number of items that did not circulate which shows the library’s policy to not notify the patron of the purchase could be a problem and suggests a need for process improvement. Even with this oversight, the suggest-a-book model only yielded 15.2% of titles without use while 44.7% of traditionally acquired materials received no use. With the other PDA models, there was only 1 book that did not get a use. So not only are there more uses for PDA acquired materials, but a much higher percentage of these materials get used.

Table 3. Comparison of print circulation across purchasing models
Percent (%) of titles with no circulation (n)
4 months 26 months
Faculty expedited 5.3
(2)
2.6
(1)
Suggest-a-book 42.4
(14)
15.2
(5)
Interlibrary loan 1.8
(1)
0.0
(0)
Holds queue 5.7
(2)
0.0
(0)
Traditional 67.7
(147)
44.7
(97)

Electronic Patron-driven Acquisitions

Electronic purchases show similar trends to what we saw with print acquisition models. As shown in Table 4, PDA models were 94.4% cheaper on average and received 1,301.1% more uses on average than the traditional model. All PDA materials received at least 11 uses so comparing use or non-use of PDA titles is not as meaningful for electronic materials. However, there were a fair number of them that received no use after the month in which a purchase was triggered. To compare, 66.9% of traditionally acquired electronic content received no use at all while 29.9% of patron-driven acquired materials received no use after the month the purchase was triggered. Curiously, some titles received significant early use, and then no more uses during the period of this study. One title received 991 COUNTER uses in one month and then no uses for the following 25 months.

Table 4. Cost and use data for electronic acquisitions
Total usea Mean (median) usea Mean (medianb) cost-per-use ($)
4 months 26 months 4 months 26 months 4 months 26 months
Patron-driven 35,148 77,562 108.5
(54.0)
239.4
(86.5)
0.70
(1.09)
0.32
(0.71)
Traditional 165 2,999 1.0
(0.0)
18.4
(0.0)
103.23
(7.50)
5.68
(6.80)

aCOUNTER-compliant uses

bFor titles that had use

Table 5. Comparison of electronic uses across purchasing models
Percent (%) of titles with no usea, b (n)
4 months 26 months
Patron-driven 61.1
(198)
29.9
(97)
Traditional 89.0
(145)
66.9
(109)

aCOUNTER-compliant uses

bFor patron-driven, we looked at titles that received no use after the month the purchase was triggered

Discussion

Patron-driven acquisition models, when compared to traditional acquisition methods, lead the way in both use and cost-effectiveness. This seems to suggest that librarians should do more to solicit patron feedback and suggestions regarding how collections are built. The more patron-driven outlets are available for purchasing, the more relevant our collections will become. Of course, this must be carefully balanced with an academic library’s mission to archive and preserve information for future generations. Patron-driven cannot be the only purchasing mechanism available to a library or we risk losing our long-term research relevance.

An interesting fact uncovered by this study is that for all but the faculty expedited orders, the PDA acquired titles were all much cheaper, on average, than the traditionally purchased materials. This is likely due to several factors. For one, we place restraints on the type and cost of materials for the too-new-to-borrow interlibrary loan books. The remaining print acquisition models, however, have no such restraints. This tends to direct our PDA purchases to less expensive books than those librarians buy for their collections.

In an effort to better support patrons and their needs, we have recently undertaken a project to consolidate all of our various options for requesting materials. For example, in our catalog, patrons can place a hold on items that have been checked out for less than three weeks, place a recall on items that have been checked out for longer than three weeks, or request an item through interlibrary loan. Each of these is a separate function and requires extensive library knowledge to successfully navigate the system to request materials. Rather than make patrons learn the intricacies of how the library operates, we hope to replace these various options with one button: “Get this item.”

This presents a powerful opportunity for us to completely overhaul the way we purchase materials in our library. Many vendors are now offering a print PDA model. Combining this print PDA with the “Get this item” button, which would obscure the fact that we have not yet purchased the item for the patron, has the potential of overturning the way we purchase materials for the library. It could supplement, or completely replace, our traditional approval purchasing program. This new program would guarantee that every item requested by patrons is used at least once, unlike the current situation shown in Table 2.

Historically, only 50% of the circulating materials we purchase are used within 10 years. If they do not circulate in the first 10 years of being available, the odds of them ever circulating drops precipitously. Contrast this to the 100% usage of the ebook PDA model and you have a strong case for change. The ebook model also offers 24/7 access and short turnaround times for users, which is a key factor in the higher use levels of these purchases. This suggests that we have been inefficiently allocating some of our collection resources for a very long period of time. We spend approximately $1.7 annually on monographs meaning about $850,000 worth of materials are never used. That is money spent annually on resources that could perhaps be reallocated for more targeted purchases to meet the needs of our students and researchers.

Additionally, some librarians express the concern that PDA removes them from the selection process. This is simply not true. Or, at least, it is no more true than the traditional approval model already has removed them from the process. PDA purchases allow librarians to set up a profile of the types of materials they would like to make available to users. This profile can exactly mirror the existing approval profile used for current approval purchases. Librarians can manipulate, manage, and revise their profiles for PDA just as they always have. The sole difference being that libraries do not own any of the materials until a PDA purchase is triggered. Looking at it this way, there seems to be little reason for continuing with traditional approval programs as currently constituted. There is always some value in having materials available just-in-case, but it is to our economic benefit to keep that just-in-case inventory as marginal as possible.

Another issue that libraries must face with electronic PDA models is the period of time untriggered records are left in the catalog. At BYU we have opted to leave records in the catalog indefinitely. This actually resembles a virtual just-in-case collecting model discussed earlier. However, this one still links expenditures to demand. Of course, increasing the number of just-in-case records in the catalog certainly raises the potential for increased expenditures over longer periods of time.

The allocation of usage between PDA and traditionally acquired materials is also a point of interest. Traditionally acquired materials are less likely to circulate. However, even when circulated, on average they received a lot fewer circulations than PDA materials. A PDA item is not only more likely to circulate; it is more likely to circulate a lot.

Finally, all of the evaluated purchasing models are trending to a lower CPU and show few signs of asymptotically trailing off anytime soon (see Figure 2). This indicates that further, even longer-term study is warranted before anything definitive can be said about the economics surrounding PDA compared to traditionally acquired materials. Out of the gate, there is plenty of evidence to suggest that PDA should be used by libraries more than they are right now. Over the first few years, PDA models are more cost effective, but traditionally acquired materials make up ground over longer periods of time. How much ground is made up over longer periods of time should be studied more carefully thus helping us determine the appropriate mix of PDA and traditional acquisitions.

Conclusion

BYU has used several different patron-driven acquisition models for a number of years. With the advent of PDA for ebooks, we have engaged in studies to determine the cost benefit of these models. This study followed in the steps of an initial four month look at cost and use of the library’s PDA models. It found that the favorable CPU numbers of the first study, in comparison to traditional collecting models, continued over a 26 month period. PDA purchases were essentially guaranteed at least one use and tended to have more significant use over time, confirming the assertion that initial demand is a predictor of future use.

The CPU of traditionally purchased titles did decline over the 26 month period but did not approach the efficiencies of the other models as the majority of items had still not been used. A longer time frame is needed to determine the long-term CPU of the traditional model in relation to the PDA purchases. However, given our own internal studies that show a ten year span is needed to reach 50% circulation, it is not anticipated that the traditional group will reach the patron driven in cost efficiencies.

We propose an expanding profile for patron driven acquisitions as a better use of collection funds to support a more relevant collection. However, for the foreseeable future, we will need to find the proper balance between these and traditional models for a number of reasons. Until all publishing is digital and the concept of out of print is obsolete, we will continue to pursue this balance.

References

  1. Rebecca Schroeder, “When Patrons Call the Shots: Patron-driven Acquisition at Brigham Young University,” Collection Building, 31 no. 1 (2012): 11–14.