Investigating How Individuals Conceptually and
Physically Structure File Folders for Electronic
Bookmarks: The Example of the Financial Services
Industry
Lisa Gottlieb and Juris Dilevko
Faculty of Information Studies, 140 St. George St., University of Toronto, Toronto, Ontario, Canada M5S 3G6.
E-mail: dilevko@fis.utoronto.ca
Personal preferences in the development of categorical
folders for bookmarks are examined in terms of both the
choice and definition of folder domain and the overall
structure of the folder system. Study participants from
the financial industry were asked to organize the same
set of finance-related bookmarks from a given list, as
opposed to describing their organizational approaches
using their own personal bookmarks, so that the orga-
nizational systems could be compared across the sam-
ple. The selection of folder domain is influenced by con-
textual factors such as intended use and relevancy to
current projects. Similarly, the structure of the overall
folder system was determined in part by participants’
navigational preferences. While the majority of partici-
pants created folders that cover the topics of finance,
government, accounting, news, law, and tax, the actual
definition of these folders and the criteria for inclusion
vary across the sample. Furthermore, these criteria can-
not be readily discerned from the folder system itself.
Variation in folder domain and definition could adversely
affect the utility of bookmark management systems for
multiple users that involve some degree of standardiza-
tion. The same variation in interpretation of seemingly
identical folders suggests that systems with automatic
categorization would not provide users with enough
flexibility in how they could organize and access their
bookmarks.
1. Introduction
In The Myth of the Paperless Office, Sellen and Harper
(2002) cite the case of DanTech, the fictional name of an
actual Danish company that wanted to abolish paper from
its worksite. More specifically, DanTech’s goal was “to do
away completely with old organizational structures and
implement new ones” (Sellen & Harper, 2002, p. 34). In
addition to designing a new open-plan office layout and
grouping staff members into “small interdisciplinary project
teams” (p. 34), the company decided to shift all of their
paper files to an electronic environment. Not only were
existing documents scanned into a shared database, but
most of the originals were shredded as well. When it came
time to implement the new electronic document system,
management encountered a particular problem. Sellen and
Harper explain:
As it happened, it was the selection and design of the
electronic filing system that presented the greatest challenge
for the IT management team. When employees were asked
what kinds of data they would like access to, and how they
should be organized, everyone had different ideas...
For the documents that employees themselves produced,
the IT managers provided a very flexible system, allowing
people to organize their own filing systems using the con-
cepts of electronic shelves and binders. However, they
vastly overestimated people’s ability to do this. With no
structure at all, people found it very difficult to categorize
their documents and often got lost in their own filing sys-
tems. (pp. 36-37)
While employees had their own ideas of how to organize
the electronic filing system, formalizing and applying these
ideas proved difficult. The fact that this difficulty existed at
the individual level, where people “got lost in their own
filing systems,” suggests that categorizing documents in a
system for multiple users presents an even greater chal-
lenge.
The case studies presented in The Myth of the Paperless
Office, including the DanTech example, demonstrate the
relevance of examining how people use, organize, and cat-
egorize their personal information resources prior to devel-
oping a system for sharing these resources in an electronic
environment. While the transfer of paper-based documents
Received April 12, 2002; revised July 11, 2002; accepted July 11, 2002
© 2003 Wiley Periodicals, Inc.
JOURNAL OF THE AMERICAN SOCIETY FOR INFORMATION SCIENCE AND TECHNOLOGY, 54(2):124–139, 2003

to a central database usually is intended to decrease an
office’s reliance on paper, personal idiosyncrasies in how
people organize papers can affect the utility of the new
system. The information studies literature about the way
individuals organize their “personal information space”
(Kwasnik, 1991, p. 389) shows that the way paper-based
resources are arranged spatially reflects their relevance and
priority to current work projects. In other words, those
ubiquitous and seemingly chaotic piles of papers encoun-
tered in most offices actually embody an organizational
system that assists in the productive use of the information
those piles contain. Gladwell (2002) notes that “[pile]s are
living, breathing archives,” and that “even the most disor-
derly piles make perfect sense to the piler” (p. 93).
The fact that those piles found on office desks (and most
likely chairs, shelves, and floors as well) structure informa-
tion for individuals in a helpful manner presents somewhat
of a conundrum: If disorderly piles can be used and under-
stood as a type of organizational system, then why does the
transferring of these piles to an electronic environment
present such a challenge? In fact, organizing information in
electronic environments introduces two factors into the
equation. First, although the desk piles might be ordered
according to issues such as priority and relevance to current
projects, they are not formally named nor defined. In con-
trast, electronic systems require individuals to formally
identify categorical domains, usually by developing a series
of file folders. Second, spatial arrangement in an electronic
environment takes on a particular meaning, with systems
offering a myriad of folder combinations, for instance by
creating hierarchies or sub-folders.
This study addresses these issues by examining how a
group of individuals both conceptually and physically or-
ganize a given set of electronic bookmarks. Fifteen (15)
participants enrolled in a graduate accounting and business
management program were asked to create electronic fold-
ers for the same 60 URLs and to file the bookmarks for
these URLs within the folders. Customized questionnaires
were used to discover the rationale behind the participants’
organizational decisions. The resulting folder systems and
participants’ questionnaire responses were analysed to-
gether in order to determine the types of folders created,
how they functioned, and what categorical domains they
represented. The discussion of the data focuses on the
implications of the variations in participants’ approaches for
the design of bookmark management systems intended to
facilitate the sharing of bookmarks within a company or
organization.
2. Background and Literature Review
2.1. The Hidden Meaning of Piles of Paper
Both Case (1986, 1991) and Kwasnik (1991) have noted
the extent to which contextual factors such as spatial ar-
rangement and intended use affect how individuals organize
their personal information resources. In an exploratory
study, Case (1986) conducted interviews with 36 social
science and humanities professors at a major university,
asking them to describe how they organized the materials in
their offices. He observed that the offices housed an average
of 18.9 separate piles of materials in their offices, each of
which served “as an aid to . . . memory, as well as a physical
scheme of organization” (p. 101). In a subsequent set of
interviews with 20 history professors, Case (1991) found
that the historians utilized four conceptual storage levels for
their information resources: space, form, topic, and treat-
ment (i.e., how or for what purpose the materials were to be
used). Confirming his previous findings, Case notes that
“two physical factors—spatial configuration and document
form—were often considered before topic in determining
document storage locations in the office” (p. 665). Kwasnik
(1991) employed a similar methodology in her study of how
eight university professors organize their paper-based re-
sources. Based on participants’ descriptions of the various
piles that populate their offices, Kwasnik identified and
categorized the various factors that informed the professors’
organizing systems. In addition to document attributes such
as author, title, and topic, she noted that situation attributes
such as need/requirement, source, and use/purpose also
played a role in participants’ decisions. Indeed, the calcu-
lation of the frequency of the various factors identified
suggests not only the importance of contextual factors, as
opposed to document-based factors, but also the interrela-
tionship of the two in deciding how to organize materials.
The idea that individuals organize their personal re-
sources according to both textual and contextual factors, and
that these contextual factors are based on personal idiosyn-
crasies, also applies to personal resources in an electronic
environment. Mackenzie (2000) looked at the role of per-
sonal preferences in developing electronic folders for e-mail
messages. Survey responses were collected from 12 insur-
ance company employees about how they organize their
e-mail messages. In addition, the organizing habits of two
managers from the company were observed, followed by
semi-structured interviews. Mackenzie found that partici-
pants’ first priority in classifying e-mail is immediate need,
while the second is an employee’s future need of the infor-
mation contained in the messages. This highly contextual
basis for organizing materials was also encountered by
Barreau (1995), who examined how seven managers orga-
nized documents on personal information management sys-
tems (PIMs). After the managers identified 10 categories of
document types included in their PIM workspace, Barreau
noted that “[o]ne of the most frequently mentioned descrip-
tors of documents within the individual’s electronic envi-
ronment was frequency of use” (p. 333). Furthermore, in
instances where documents were collocated according to
topic, the notion of topic was often defined as “information
relevant to [an individual’s] projects” (p. 333).
2.2. When Piles Become Electronic Folders
Although individuals arrange their documents according
to personal factors such as intended use, there is nonetheless
JOURNAL OF THE AMERICAN SOCIETY FOR INFORMATION SCIENCE AND TECHNOLOGY—January 15, 2003 125