





































276  ATL A 2018 PROCEEDINGS

Managing a Large-Scale 
Weeding Project
When We All Pull Together
By Leslie A. Engelson, Murray State University

ABSTR ACT The main circulating collection at Murray State University 
had never been thoughtfully and intentionally weeded. Several factors 
aligned, raising the need for a thorough weed of the collection result-
ing in weeding over a quarter of the collection (over 100,000 volumes). 
Most of the volumes were pulled and processed over a 32-week period. 
This session will discuss the circumstances that made the time ripe for 
weeding, the criteria that informed the decision-making process, tools 
used in that process, workflow for accomplishing the physical process-
ing, cross-library participation in the project, and things to consider 
when implementing a project of this magnitude.

SITUATION

The circulating collection at Murray State University (MSU) consisted 
of approximately 300,000 volumes and includes the oversize collec-
tion. It was created primarily by faculty selecting resources, as well as 
the consolidation of department libraries into the University Libraries 
system. The various areas of the collection have never been system-
atically developed or weeded, and the Research & Instruction (R&I) 
librarians have maintained a hands-off approach. As you can imagine, 
this created a spotty and inconsistent collection.

An analysis of the collection in 2013 revealed that the average date 
of publication of the resources in the collection was 1973, with most 
books published in 1968. Ninety percent of the titles were published 
before 2003. The oldest area of the collection was the CNs (Inscrip-
tions, Epigraphy), with an average publication date of 1945, and the 
most current area of the collection was ZA (Information sciences), 
with an average publication date of 2007. Almost half of the collection 



Listen and Learn Sessions  277

has not circulated since 2000. Of those that have circulated, only 13% 
have circulated five or more times.

Students often complained (on social media as well as formal 
surveys) of the lack of space. No private or group study rooms were 
available, and they often had to sit on the floor because all the seat-
ing was in use. Circulation statistics had been decreasing from about 
15% in 2011/2012 to about 12% in 2016/2017, while the availability 
and use of electronic resources was increasing.

The R&I librarians were insecure about making weeding deci-
sions because of their lack of experience and traditionally hands-off 
approach. Because of my experience with collection management, as 
well as a recent presentation on collection analysis tools at an ATLA 
conference, I approached the dean about making a similar presenta-
tion to the R&I librarians. As part of the discussion, we determined 
some goals for the project.

PROJECT GOALS

Establishing a desired collection strength and preferred format for 
subject areas was one of the first goals. The R&I librarians worked dili-
gently to determine, based on curricular needs, the desired strengths 
of the areas of the collection they were responsible for. This gave them 
a better understanding of the target they were trying to reach as they 
made retention decisions.



278  ATL A 2018 PROCEEDINGS

Secondly, they reviewed the circulating collection for items which 
were no longer relevant due to a change in programs and curricular 
needs.

An important goal of the project was to disseminate the withdrawn 
materials in a responsible manner. Because MSU is a state institution, 
there are legislative requirements that restrict how materials can be 
disposed of, but we were concerned about ensuring that as many titles 
as possible could be given away or sold and not put in the landfill.

The fourth goal was to start reaching the desired collection 
strength through acquisitions and access-based models.

Finally, the last goal was to renovate the space to create study areas 
for individuals and groups.

Since the establishment of these goals, we have had a change in 
leadership. The first, second, and third goals have been met, but it is 
unlikely that formally initiating efforts to meet the fourth goal of deter-
mining current collection strengths with the intention of developing 
the areas will happen. The fifth goal has been partially met but was 
limited based on a lack of funding.

CONTRACTS

A result of the collection management presentation was a decision 
to contract with Sustainable Collection Services (SCS) for the use of 
their GreenGlass collection analysis tool and with Better World Books 
(BWB) to disseminate materials in a more sustainable way.

Contracts with both SCS and BWB were discussed with the Univer-
sity Procurement Services office at length, as they had to conform to 
state law. There were quite a few hurdles with this process, because 
this department doesn’t understand how libraries do business, and 
their interpretation of state law regarding contracts and how mate-
rials purchased by the University can be disposed of is very rigid. In 
the end, they approved the contract with GreenGlass without going 
out to bid, as well as giving withdrawn books to Better World Books 
to sell for us.

PROCESS

After sending SCS a file of 309,156 bibliographic records with 338,560 
associated item records, SCS built a database of those records along 



Listen and Learn Sessions  279

with data from other databases such as WorldCat, HathiTrust, and 
Choice Reviews. This allowed us to compare our holdings with those 
of other universities in Kentucky, as well as benchmark libraries we 
had identified. Additionally, we could determine if the titles were 
available in full text in open access databases such as HathiTrust, 
as well as look at reviews in Choice Reviews. The interface provided 
the ability to search by and report on date of acquisition, publication 
data, circulation count, last circulation day, and LC call number range. 

The data in GreenGlass was current as of the date it was extracted 
from our ILS. Because it is not connected to our ILS, it is only a snap-
shot in time. We still have access to our GreenGlass database, but it 
does not reflect recent acquisitions, current circulation data, or the 
fact that we no longer own many of the items that are represented in 
the database.

CRITERIA

While we were waiting for the database to be created, the R&I librar-
ians established what we called “no-brainer” criteria that I could use 
to start the weeding process. These criteria fit the parameters of the 
type of collection we were developing (a current circulating collec-
tion that is appropriate for the community served and curriculum 
supported). Once these were established, I was able to run reports in 
our ILS and start withdrawing the following materials:

 • Superseded editions

 • Foreign languages not currently taught

 • Textbooks published before 1995

 • Medical or clinical books older than 1985

 • Software books published prior to 2007

 • Workbooks, lab books, study guides intended to be written in

 •  Unmendable materials or those that are more expensive to 
mend than to replace

 • Duplicate copies published prior to 2010
 • Books that clearly do not support the curriculum

Criteria were also established that excluded materials from being 
weeded. These criteria included:



280  ATL A 2018 PROCEEDINGS

 • MSU authors

 • MSU theses

 • Books added in the last 5 years

 • Books circulated a significant number of times

 • Books related to the region
 • Last Kentucky library to hold

Additionally, the R&I librarians established criteria for the areas 
of the collection for which they were responsible. These criteria were 
used to run reports in GreenGlass from which titles were selected 
for withdrawal. From these reports, 113,858 titles were identified for 
withdrawal, which equated to 37% of the collection.

WORKFLOW

Query results in the SPS database were exported in Excel spread-
sheets and included the query parameters. The R&I librarians would 
then remove items from the spreadsheet that matched the excluded 
criteria, and then the spreadsheets were loaded into a folder on a 
shared drive.

Keeping the original spreadsheets intact in order to keep a record 
of the query used and its results, I compiled each list into a master 
list, removing information that was not relevant for my needs. The 
information that remained included:

 • Location code (CIR or FOL)

 • Call number

 • Enumeration or Chronological data (from item record)

 • Title
 • Barcode

This project was initiated in 2013. Throughout 2014-2015, the R&I 
librarians made deselection decisions. The project was then put on 
hold for a couple of years, as we had a change in deans. Additionally, 
the previous dean had promised the faculty that they could review the 
books before we withdrew them. However, the new dean struggled 
with the logistics of how that could happen, given the interdisciplin-
ary nature of college curriculum, and was able to get approval from the 



Listen and Learn Sessions  281

provost to withdraw the books without faculty review. The withdraw-
ing part of the project began on August 28, 2017, and was stopped on 
November 15, 2018.

A total of nine staff and sixteen students from across library 
departments participated in the actual physical withdrawing process. 
Some helped throughout the entire project, and some just helped 
once. Each person’s participation was voluntary, and they were able 
to come and go during the project as other responsibilities affected 
their availability.

Once I was given approval to begin withdrawing materials, I 
decided to start in the Zs and work our way forward. One reason for 
this was that the Q-Zs were in an area of the library that had high 
humidity, which caused outbreaks of mold. We had the mold reme-
diated twice, but the problem would continue to happen unless the 
books were removed from this area. Additionally, this area was on the 
side of the building that was selected for the major part of the renova-
tion. Thus, it seemed like this was the best area to tackle first.

To begin, I printed out the spreadsheet, bundling eight or nine 
pages together for about two hundred titles per bundle. In order to 
keep track of which titles had been printed and which had been with-
drawn, I highlighted the titles printed in blue in the spreadsheet. 
Once they were withdrawn, I changed the highlighting to grey. I kept 
a running list of each bundle by LC letter and page numbers in a sepa-
rate spreadsheet along with the name of each person responsible for 
that bundle and the number of books in each bundle. 

This helped me monitor when lists were completed, who was 
responsible for that call number area, and who to go to if I had any 
questions. It also helped me keep track of how many had been with-
drawn so I could report this at a bi-monthly department heads meeting. 

Each person was responsible for pulling and processing the titles 
on their list. They had their own cart, as well as a piece of paper with 
their name printed on it in a large font. If they had to stop in the middle 
of completing their bundle, they would leave the books on their cart 
with their name sheet on top and then would pick up where they left 
off when they were able to come back to the task. This minimized steps 
in the process getting accidently skipped because it was unclear what 
stage the books were at in the process.



282  ATL A 2018 PROCEEDINGS

The pulling process involved locating the book by call number 
and comparing the title and barcode to that listed on the sheet. If 
the title was found, it was marked with a checkmark (✓) next to the 
barcode. If the title was not found, it was marked with a “not” symbol 
(   ). Sometimes the call number would match but the barcode would 
not, or the barcode would match and the call number would not. This 
was usually due to a mistake in the record, and the book was the one 
listed for withdrawal. 

After the books were pulled, they were then searched for by ISBN 
in BWB’s database to determine if the title was one that they wanted. 
If so, it was added to a box for BWB. If not, it was placed in a bin for 
recycling or discard. All books were physically processed for with-
drawal by crossing through the barcode and spine label with a black 
permanent marker and stamping DISCARD on the top of the text block 
and title page. We do not want these books making their way back to 
the library, so, want to make it clear that they have been withdrawn.

In addition to providing the boxes, BWB provided free pick-up and 
shipping. We had to have at least six boxes for pickup, and they could 
be sent either through the mail or by freight. Freight shipping requires 
forty boxes on a pallet that is shrink-wrapped and a loading dock. It 
was determined that this was too much work for shipping/receiving to 
manage. When we accumulated at least six boxes, we communicated 
with BWB to let them know and received postage ,which we printed 
out and applied to the boxes. We then contacted shipping/receiving 
to let them know how many boxes were ready for pick-up.

Coordination with the maintenance department and shipping/
receiving was essential, as we had countless bins of books for main-
tenance to pick up and dispose of and often ten to sixteen boxes daily 
for shipping/receiving to pick up. Their staffing and schedule were 
impacted, so it was important to communicate with those depart-
ments about the project early on.

At one point we ran out of BWB boxes, and so had to find space to 
store the withdrawn books until additional boxes arrived.

When all the titles in a bundle were pulled, the bundle was returned 
to me. I would then batch withdraw the titles on the list from our ILS 
using the barcodes on the spreadsheet. Titles that were not found were 
searched in the ILS to see if we still had the item or if it was checked out 
or lost. A circulation staff member who was good at finding missing 



Listen and Learn Sessions  283

books did a second search for those that were supposed to be on the 
shelf and often found many of those missed in the first search. 

At the end of each month, I ran a report of titles withdrawn, which 
included the OCLC number. I then batch removed our holdings from 
WorldCat using the numbers from that report. 

What is a major project without a progress chart? With this proj-
ect, ours was a Progress Shelf consisting of previously discarded 
books. Each book represented 1,100 books to withdraw. For each 1,100 
books withdrawn, the person who turned in the 
spreadsheet that included the 1,100th book got 
to pull a book from this shelf. 

While we were waiting for the project to 
begin, the Cataloging student workers created 
book covers using their creativity and senses 
of humor to come up with clever book titles 
and authors. It was a fun part of the project 
to be able to pull a book from the Progress 
Shelf and see what the title would be. Some 
examples are Charlotte’s Website by E.Z. White, 
Baker’s Men by Pat E. Cake, and Downpour by 
Wayne Dwops.

I made sure that everyone who was involved 
in the project long-term got to pull a book from 
the Progress Shelf. This was a great visual reminder of where we were 
at in the project and how much farther we had to go.

END OF PROJECT

The library is a popular and busy place on campus. During peak times, 
students often have to sit on the floor because there is no available 
space at the tables and chairs. When this project was initiated in 
2013/2014, one of the goals was to withdraw a quarter of the collection 
in order to be able to renovate the entire area. The plan was to install 
group study rooms as well as a reading area in order to address the 
ongoing complaints about the lack of space that were voiced in surveys 
as well as on social media. 

However, the budget situation started to decline in 2015 and 
has steadily gotten worse. Possible funding for the renovation  

©
 L

es
lie

 A
. E

ng
el

so
n



284  ATL A 2018 PROCEEDINGS

disappeared. The dean was concerned that, if we finished withdraw-
ing all the materials in the master withdraw spreadsheet, we would 
have too many empty shelves. The decision was made to withdraw 
enough books so that we could shift the remaining books out of the 
area where humidity was a problem and move in tables and chairs 
and study carrels in order to create some study space. 

The picture shows the area with the high humidity before the proj-
ect. As you can see, it is filled with stacks upon stacks of books. 

This is how that same area looks now. It is a popular area for students, 
and we have had no complaints about the removal of the books.

OUTCOMES

As of June 5, 2018, we have received almost $5,000 from the sale of 
3,552 books of those that were sent to BWB. Additionally, 2,445 of 
the books sent were recycled. BWB provides environmental metrics 
that demonstrate the positive impact on the environment of selling 
or recycling these books rather than sending them to the dump. The 
environmental metrics indicate that we have saved 96 trees, almost 
13,000 lbs of greenhouse gases, and 113 cubic yards of landfill space. 

©
 L

es
lie

 A
. E

ng
el

so
n

©
 L

es
lie

 A
. E

ng
el

so
n



Listen and Learn Sessions  285

This is just the number of books sold or recycled. I did not keep track 
of the total number of books sent.

When the project ended, we had withdrawn all the books on the 
withdraw spreadsheet that were in the P-Z range. This totaled 50,438 
books and was 44% of the total project. 

However, a pleasant and unexpected outcome was the unification 
and comradery enjoyed by students and staff crossing departmen-
tal boundaries and working together on such a massive project. I got 
to know students from other departments who I had never spoken 
to before. Relationships with staff in other departments improved, 
which now facilitates communication with the cataloging department. 
The library came together for a mutual purpose, resulting in a sense 
of unification and mutuality. We were all truly sad when the project 
ended; not just because we didn’t get to complete it, but because we 
had such a positive experience with it.


