Content uploaded by Christophe Benavent
Author content
All content in this area was uploaded by Christophe Benavent on Nov 12, 2018
Content may be subject to copyright.
Determinants
of
e-consumer
productivity
in
product
retrieval
on
a
commercial
website:
An
experimental
approach
Mohamed
Slim
Ben
Mimoun
a,
*,
Marion
Garnier
a
,
Richard
Ladwein
b
,
Christophe
Benavent
c
a
SKEMA
Business
School-Univ
Lille
Nord
de
France,
Avenue
Willy
Brandt,
59777
Lille,
France
b
IAE
de
Lille,
CNRS
LEM
UMR
8179
104,
Avenue
du
Peuple
Belge,
59043
Lille
Cedex,
France
c
Universite
´Paris
Ouest,
CEROS
200
Avenue
de
la
Re
´publique,
92000
Nanterre,
France
1.
Introduction
When
consumers
visit
Internet
retailers
to
make
a
purchase,
many
of
them
fail
to
complete
the
transaction
and
prematurely
abandon
their
purchase
[1–3].
According
to
Ranganathan
and
Grandon
[4],
43%
of
attempted
online
purchases
fail
because
the
consumer
cannot
find
the
product;
even
if
consumers
do
not
need
to
move
in
space
to
shop
online,
they
still
can
become
lost
[5].
To
avoid
these
situations,
consumers
need
to
understand
the
organization
and
information
structures
of
a
shopping
website,
predict
the
links
to
follow,
and
read
and
understand
the
displayed
content
[6].
As
we
show
in
Fig.
1,
information
search
and
product
retrieval
processes
are
central
to
the
purchasing
process.
Problems
with
navigation
and
product
retrieval
on
websites
thus
are
significant
obstacles
to
online
purchasing
[7,8],
which
are
exacerbated
by
the
diverse
offers,
intangibility,
and
the
lack
of
a
salesperson
presence
that
characterize
online
selling
[9,10].
Because
these
challenges
make
it
more
challenging
to
find
viable
products,
the
online
shopping
process
grows
more
complicated,
which
considerably
reduces
consumer
productivity
[11].
1
We
define
consumer
productivity
as
the
consumer’s
efficiency
in
using
resources
to
find
information
and
effectiveness
in
retrieving
the
intended
and
needed
product.
If
we
can
identify
driving
forces
and
inhibitors
of
online
consumer
productivity,
online
retailers
would
gain
a
better
understanding
of
e-
shoppers’
behaviors
and
could
create
more
consumer-friendly
and
effective
websites.
Accordingly,
we
review
previous
research
from
various
fields
that
addresses
different
aspects
of
online
shopping
behavior
related
to
consumer
productivity
to
predict
which
factors
likely
affect
consumer
productivity
in
the
online
product
retrieval
process.
To
the
best
of
our
knowledge,
however,
no
studies
have
specifically
examined
the
determinants
of
e-consumer
productivi-
ty
for
product
retrieval
on
a
website.
We
further
posit
that
when
consumers
engage
in
online
shopping,
they
are
simultaneously
consumers
and
users
of
an
IT
artifact
[12].
Therefore,
designing
web-based
stores
should
involve
knowledge
from
various
fields,
such
as
marketing,
information
systems,
and
human–computer
Information
&
Management
51
(2014)
375–390
A
R
T
I
C
L
E
I
N
F
O
Article
history:
Received
4
February
2011
Received
in
revised
form
3
October
2013
Accepted
16
February
2014
Available
online
12
March
2014
Keywords:
E-consumer
productivity
Product
retrieval
Site
design
Online
experiment
Cognitive
absorption
Site
complexity
A
B
S
T
R
A
C
T
This
article
investigates
what
determines
e-consumer
productivity,
in
the
specific
case
of
product
retrieval,
on
a
commercial
website.
With
a
2
2
2
factorial
design
on
292
participants,
an
online
experiment
reveals
that
productivity
in
product
retrieval
(measured
in
terms
of
effectiveness,
efficiency,
and
time)
relates
to
website
design
(e.g.,
abstraction
level
of
labels,
animation),
user
characteristics
(e.g.,
Internet
experience,
product
category
familiarity,
cognitive
absorption),
and
situational
characteristics
(e.g.,
task
nature).
The
results
also
confirm
interactive
effects
among
the
type
of
strategy
used,
the
nature
of
the
task,
and
the
website
design.
These
findings
have
notable
implications
for
both
research
and
practice.
ß
2014
Elsevier
B.V.
All
rights
reserved.
*Corresponding
author.
Tel.:
+33
3
20
21
47
13.
E-mail
addresses:
m.slim_benmimoun@skema.edu
(M.S.
Ben
Mimoun),
marion.garnier@skema.edu
(M.
Garnier),
richard.ladwein@univ-lille1.fr
(R.
Ladwein),
christophe.benavent@u-paris10.fr
(C.
Benavent).
1
Ingene
[11]
considers
productivity
from
a
consumer
perspective
in
terms
of
the
ratio
of
outputs
divided
by
inputs.
Contents
lists
available
at
ScienceDirect
Information
&
Management
jo
u
rn
al
h
om
ep
ag
e:
ww
w.els
evier.c
o
m/lo
c
ate/im
http://dx.doi.org/10.1016/j.im.2014.02.003
0378-7206/ß
2014
Elsevier
B.V.
All
rights
reserved.
interactions
[13].
Specifically,
on
the
basis
of
studies
into
information
retrieval
on
the
web
[14–16],
productivity
[11,17,18]
website
atmospherics
[19,20],
and
website
complexity
[21–23],
we
hypothesize
that
e-consumer
productivity
for
product
retrieval
depends
on
situational,
user,
and
website
characteristics,
as
well
as
interactions
between
some
of
these
elements.
With
our
experimental
test
of
these
effects,
we
contribute
to
extant
literature
by
examining
antecedents
of
e-consumer
productivity
in
a
single
model
and
offer
implications
for
online
retailers
that
need
to
facilitate
consumers’
productivity
through
the
retailers’
design
of
self-service
devices
[24].
In
the
following
sections,
we
introduce
the
concept
of
e-
consumer
productivity
in
general,
as
well
as
in
the
context
of
online
search
and
product
retrieval.
We
then
describe
some
potential
effects
of
situational,
user,
and
site
characteristics
on
online
consumer
productivity.
We
test
our
six-hypothesis
concep-
tual
model
using
experimental
data
and
then
conclude
with
a
discussion
of
the
results,
the
study
limitations,
and
implications
for
research
and
practice.
2.
Consumer
productivity
in
the
context
of
product
retrieval
To
retrieve
a
desired
product
in
a
brick-and-mortar
context,
consumers
identify
the
space
and
time
in
which
the
product
is
available
by
selecting
a
commercial
interface
(i.e.,
store)
that
offers
the
product
and
decoding
the
specific
organization
of
that
commercial
interface.
Similar
to
bricks-and-mortar
shopping,
consumers
who
visit
e-commerce
sites
need
to
undertake
parallel
processes:
understanding
the
site’s
interface,
reflecting
on
information
needs,
and
progressing
in
the
problem-solving
process
that
led
to
the
site
visit
[6].
Failing
to
identify
the
category
within
a
global
assortment
or
having
problems
orienting
in
an
e-store
can
lead
the
consumer
to
abort
the
purchase
process
[25,26].
That
is,
the
web
may
free
consumers
from
time
and
space
constraints,
but
finding
a
needed
product
or
service
can
be
difficult
or
even
impossible
if
the
classification
system
on
the
web
interface
is
unclear
[27].
An
unusable
website
likely
inhibits
retrieval
of
important
information
[28]
and
reduces
e-consumer
productivity.
2.1.
Consumer
productivity
Various
concepts,
such
as
expertise
or
search
skills,
describe
consumer–user
interactions,
but
the
concept
of
e-consumer
productivity
relates
to
a
user’s
ability
to
retrieve
information
from
the
website
that
is
necessary
for
purchases.
It
includes,
but
is
not
limited
to,
the
product
description,
price,
image,
payment
type,
delivery
methods,
and
after-sales
support.
Information
search
is
often
the
first
step
of
the
coproduction
process
[29],
and
information
search
costs
affect
the
likelihood
of
purchase
[30,31].
An
online
consumer
is
simultaneously
a
customer
of
the
(virtual)
store,
a
user
of
self-service
elements
of
the
coproduction
process
[29],
and
a
user
of
an
information
system
[12].
We
thus
consider
two
approaches
to
study
e-consumer
productivity:
consumer
productivity
(from
the
marketing
literature)
or
online
search
performance
(an
IS-driven
approach).
We
also
note
that
the
definition
and
measure
of
productivity
is
complex
and
highly
context-dependent
[18,32–34],
such
that
there
is
no
single
meaning
attached
to
this
term;
the
inputs
and
outputs
used
to
measure
it
also
vary
with
its
application
[29].
The
marketing
literature
investigates
productivity
from
both
firms’
and
customers’
perspectives,
as
we
summarize
in
Table
1
by
describing
various
approaches
to
and
definitions
of
productivity.
As
the
works
in
Table
1
indicate,
there
appears
to
be
some
parallel
between
customer
efficiency
and
more
classical
concepts
of
manufacturing
or
employee
productivity,
where
time
is
the
main
customer
input.
Although
the
concepts
of
consumer
efficiency
[29]
and
productivity
[18]
are
linked,
they
are
not
identical.
Furthermore,
as
Sheth
and
Sisodia
[41]
suggest,
effectiveness
and
efficiency
appear
as
two
complementary,
multiplicative
dimensions
of
consumer
productivity.
Thus,
effi-
ciency
and
effectiveness
are
widely
recognized
as
relevant
concepts
for
assessing
productivity
in
general
and
consumer
productivity
in
particular.
Productivity
can
then
be
considered
as
Fig.
1.
Purchase
decision
processes
online
and
central
questions
related
to
product
retrieval.
M.S.
Ben
Mimoun
et
al.
/
Information
&
Management
51
(2014)
375–390
376
the
efficient
and
effective
transformation
of
available
inputs
to
desired
outputs
[24].
Anitsal
and
Schumann
[18]
further
argue
that
a
manufacturing
view
of
productivity
relates
more
to
efficiency
(i.e.,
the
input
side),
whereas
a
service
orientation
is
focused
on
effectiveness
(i.e.,
the
output
side).
We
therefore
define
these
elements
of
consumer
productivity
more
clearly.
Efficiency
is
the
minimization
of
inputs
required
to
reach
an
aim
[42,43],
as
defined
by
Vuorinen
et
al.
[42,
p.
379]:
‘‘the
degree
to
which
an
activity
generates
a
given
quantity
of
outputs
with
a
minimum
consumption
of
inputs,
or
generates
the
largest
possible
outputs
from
a
given
quantity
of
inputs’’
or
by
Ojasalo
[43,
p.
9]:
‘‘the
degree
to
which
the
system
utilizes
the
‘right’
resources’’.
Sink
[35]
formulates
it
as
follows:
Efficiency
¼Expected
resources
consumption
Actual
resources
consumption
Effectiveness,
in
contrast,
relates
to
maximizing
the
expected
effect
when
the
desired
outputs
are
accomplished
[35].
Therefore,
for
the
first
approach
related
to
e-consumer
productivity,
effectiveness
and
efficiency
are
representative
dimensions
from
a
consumer-oriented
perspective,
independent
of
the
terms
used
to
contextualize
productivity.
2.2.
Online
search
performance
The
second
approach
to
studying
e-consumer
productivity
regards
consumers
as
users
of
the
information
system.
Therefore,
the
literature
on
online
search
behavior
is
instructive.
Some
studies
use
two
generic
categories
of
assessment
variables:
search
outputs
and
search
outcomes
[44].
Others
authors
[15]
highlight
performance
(i.e.,
correctness
and
time),
such
as
when
Schaik
and
Ling
[16]
assess
task
performance
in
terms
of
speed
(e.g.,
the
time
spent
on
the
task)
and
efficiency
(e.g.,
number
of
pages
loaded)
or
when
Turetken
and
Sharda
[45]
consider
that
online
end-user
success
has
two
dimensions:
effectiveness
(i.e.,
the
desirability
of
the
achieved
outcomes)
and
efficiency
(i.e.,
how
well
the
user
exploits
available
inputs
in
producing
outcomes).
These
findings
align
with
our
previous
discussion
of
consumer
productivity
research,
which
also
used
efficiency
and
effectiveness
as
dimensions
to
assess
productivity
[18].
Furthermore,
website
usability
investigations
tend
to
present
effectiveness,
efficiency,
and
satisfaction
as
formal
indicators
of
site
usability,
such
that
strong
usability
decreases
inputs
(i.e.,
time,
effort,
and
emotional
energy)
and
thereby
improves
productivity
in
terms
of
efficiency
and
effectiveness
[24].
Villey-Migraine
[46]
defines
effectiveness
as
the
capacity
to
achieve
a
given
objective
and
defines
online
task
efficiency
as
the
ability
to
do
so
with
minimum
effort.
In
general,
less
effort
means
more
efficiency.
Those
definitions,
applied
to
a
search
task,
align
with
our
previously
considered
definitions
of
efficiency
and
effectiveness.
Still
other
research
regards
effectiveness
and
efficiency
as
insufficient
indicators
and
integrates
time
into
search
performance
measures
[31,47].
Limayem
et
al.
[48]
cite
time
saving
as
a
key
outcome
of
online
shopping
and
information
search;
Hostler
et
al.
[49]
use
time
to
determine
user
performance;
Xiao
and
Benbasat
[12]
highlight
the
importance
of
time
for
measuring
consumer
effort;
and
Lankton
et
al.
[50]
include
time
(viz.,
task
speed)
in
their
model
of
information-seeking
performance.
3.
A
model
of
e-consumer
productivity
Using
existing
definitions
of
consumer
productivity
and
considering
the
necessity
of
envisaging
productivity
regarding
its
specific
context,
we
define
e-consumer
productivity
in
product
retrieval
as
the
combination
of
the
efficiency
with
which
the
consumer
uses
resources
to
find
information
on
a
website
and
her
or
his
effectiveness
in
retrieving
the
needed
product.
Drawing
on
both
consumer
productivity
and
the
online
search
performance
literature
we
use
effectiveness,
efficiency,
and
time
(as
an
additional
variable
to
the
basic
definition)
as
relevant
dimensions
to
assess
consumer
productivity
in
a
product
retrieval
context.
Next,
we
consider
previous
research
into
online
consumer
behaviors
[12,19,20,31,51]
and
information
retrieval
[14,52]
to
support
our
examination
of
the
influences
of
three
categories
of
determinants
that
are
widely
recognized
as
mainly
and
mostly
affecting
online
consumer
behavior
and
information
search:
(1)
situational
characteristics,
(2)
user
characteristics,
and
(3)
site
design.
3.1.
Situational
characteristics
Consumer
situations
include
all
factors
specific
to
a
certain
time
and
place
of
observation
that
can
influence
consumers
once
they
enter
the
store
[53].
Situational
factors
have
a
significant
influence
on
online
purchasing
success,
too
[51],
such
as
the
interaction
between
the
task
nature
[52,54]
and
the
task
(or
behavioral)
strategy.
3.1.1.
Influence
of
task
nature
Four
distinct
theoretical
frameworks
appear
in
studies
of
tasks:
task
qua
task,
2
task
as
behavior
requirements,
task
as
behavior
descriptions,
and
task
as
ability
requirements
[55].
To
study
a
task,
it
is
necessary
to
separate
individual
and
task
effects
and
describe
tasks
independently
of
the
people
who
perform
them
[55],
which
indicate
the
applicability
of
the
task
qua
task
and
behavior
requirements
frameworks.
Using
a
combination
of
these
frame-
works
indicates
that
any
task
contains
three
essential
components
[55]:
products
(i.e.,
one
output),
acts,
and
information
cues
(i.e.,
two
inputs).
Furthermore,
for
online
tasks,
various
typologies
appear:
-
goal-directed
versus
experiential
[54,56,57];
-
specific
search
task,
non-specific
search
task,
or
general
browsing
(no
specific
aim)
[58];
and
-
closed,
with
specific
purposes,
or
open,
with
more
general
purposes
[59].
However,
in
a
controlled
experiment,
it
is
very
difficult
to
measure
the
outcomes
of
loosely
defined
objectives,
such
as
in
experiential
tasks
or
general
browsing
[60].
Indeed,
in
the
case
of
general
browsing,
the
individual
has
no
defined
purpose
and
no
idea
of
information
to
search
for,
which
prevents
the
study
of
productivity
or
performance
in
such
a
context.
Because
links
and
analogies
exist
among
different
task
typologies,
however,
most
search
tasks
can
be
studied
empirically
according
to
the
distinction
between
open
and
closed
tasks
[14,52].
For
this
reason,
a
majority
of
experiments
in
the
area
use
Marchionini’s
open
(with
a
large
but
defined
purpose)
and
closed
(with
a
precise
defined
purpose)
tasks
[52].
We
similarly
adopt
this
distinction
in
our
research.
3.1.2.
Influence
of
interaction
between
task
nature
and
task
strategy
A
task
strategy
instead
refers
to
‘‘the
process
that
individuals
use
to
accomplish
a
task’’
[61,
p.
7].
Differences
in
the
task
nature
require
users
to
adopt
unique
mechanisms
to
interact
with
the
online
environment
[54,56,62],
such
as
three
distinct
information-
seeking
strategies
[63]:
-
a
search-dominant
strategy,
in
which
the
consumer
immediately
clicks
on
the
search
button
of
the
browser,
such
that
his
or
her
site
behavior
is
dominated
by
the
use
of
a
search
engine;
2
In
the
task
qua
task
framework,
‘‘the
task
is
described
as
a
class
of
phenomena
that
are
totally
independent
of
individual
phenomena’’
[55].
M.S.
Ben
Mimoun
et
al.
/
Information
&
Management
51
(2014)
375–390
377
-
a
link-dominant
strategy,
in
which
consumers
almost
exclusively
use
hyperlinks
on
the
site;
and
-
a
mixed
strategy,
combining
both
strategies
in
incremental
or
circular
ways
[61].
This
interaction
between
strategy
and
nature
is
fundamental
to
our
study.
For
example,
if
goal-directed
users
face
complexity,
they
may
employ
a
search-dominant
strategy
relying
on
search
functions
available
on
the
site
[54],
whereas
users
executing
open
tasks
may
rely
on
links-dominant
strategies
to
reduce
complexity.
The
type
of
strategy
used
for
a
particular
task
nature
influences
the
performance
of
a
closed
search
[64];
in
addition,
for
an
open
task,
the
use
of
search
engines
has
a
negative
impact
on
search
effectiveness
(i.e.,
the
capacity
to
reach
a
given
objective)
[65],
and
a
site
with
no
search
engine
capability
hampers
closed
task
execution.
Recent
exploratory
research
also
confirms
that,
for
a
closed
search
task,
a
search-dominant
strategy
leads
to
better
efficiency
than
a
links-based
strategy
[67].
Thus,
search
efficiency
depends
strongly
on
how
well
the
adopted
(information-seeking)
strategy
fits
the
specific
task
[66].
More
recently,
the
exploratory
results
of
Authors
[67]
indicate
that,
for
a
closed
search
task,
the
adoption
of
a
dominant
search
strategy
(based
on
the
use
of
search
functions)
leads
to
better
efficiency
than
the
use
of
a
links-based
strategy.
In
other
words,
the
strategy
used
affects
both
effectiveness
and
efficiency
levels
[64,67]
and,
thus,
consumer
productivity.
However,
the
effect
still
depends
on
the
nature
of
the
executed
task
[54,66].
That
is,
search-
dominant
strategies
better
fit
closed
tasks
than
do
links-dominant
strategies,
whereas
the
opposite
is
true
for
open
tasks
[65].
We
therefore
hypothesize
the
following:
H1.
The
impact
of
the
adopted
strategy
type
on
e-consumer
productivity
depends
on
the
task
nature.
H1a.
When
a
closed
task
is
performed,
the
use
of
a
‘‘search-
dominant
strategy’’
leads
to
greater
productivity,
and
the
use
of
a
‘‘links-dominant
strategy’’
leads
to
weaker
productivity.
H1b.
When
an
open
task
is
performed,
the
use
of
a
‘‘links-domi-
nant
strategy’’
leads
to
a
better
productivity,
and
the
use
of
a
‘‘search-dominant
strategy’’
leads
to
a
weaker
productivity.
3.2.
Individual
characteristics
Research
in
psychology
and
sociology,
as
well
as
in
management
and
marketing,
has
revealed
that
a
person’s
characteristics
affect
his
or
her
shopping
behaviors.
However,
considering
the
vast
multitude
and
diversity
of
individual
variables
that
might
affect
consumer
behavior,
for
clarity,
we
include
only
variables
that
benefit
from
some
consensus
in
the
prior
literature,
independent
of
the
adopted
approach.
Thus,
we
investigate
the
influences
of
Internet
experience,
product
category
familiarity,
and
cognitive
absorption.
3.2.1.
Internet
experience
Experience
is
an
important
contingent
factor
that
affects
user
performance
[55,68];
Internet
experience
is
one
of
the
most
widely
cited
determinants
of
online
user
behavior
[44].
Similarly,
network
familiarity
is
a
main
influence
on
Internet
consumer
behavior
[69].
We
therefore
propose
the
following:
H2.
Internet
experience
positively
affects
e-consumer
productivity.
Regarding
the
specific
effect
of
Internet
experience
on
information
search,
Khan
and
Locatis
[70]
examine
the
search
performance
of
novices
and
experts
and
find
that
experts
can
better
prioritize
search
tasks.
Thus,
experience
positively
affects
search
performance
[71],
likely
because
it
helps
people
understand
the
interrelationships
among
the
elements
of
the
task
stimulus
and
distinguish
relevant
from
irrelevant
information
[72].
Finally,
experience
with
the
Internet
makes
information
searches
take
less
time
and
require
less
effort
[73].
Therefore,
we
predict
the
following:
H2a.
More
Internet
experience
leads
to
better
efficiency
in
exe-
cuting
tasks
on
a
commercial
website.
H2b.
More
Internet
experience
leads
to
less
time
spent
executing
tasks
on
a
commercial
website.
Because
prior
research
has
indicated
that
Internet
experience
affects
effort
and
time
requirements
but
not
necessarily
the
search
result,
we
do
not
predict
any
specific
effect
on
effectiveness.
3.2.2.
Familiarity
In
addition
to
Internet
experience,
information-seeking
perfor-
mance
can
be
improved
by
familiarity
with
the
focal
topic
area
in
a
specific
web
search
(i.e.,
domain
familiarity)
[74,75].
For
online
shopping,
domain
familiarity
likely
involves
product
category
familiarity.
In
general,
people
with
greater
product
familiarity
should
be
able
to
acquire
more
relevant
information
on
a
site
because
they
know
where
to
find
suitable
information,
which
makes
them
more
effective
in
their
searches
[76].
We
can
then
formulate
the
following
hypotheses:
H3.
Product
category
familiarity
has
a
positive
impact
on
e-con-
sumer
productivity.
H3a.
A
higher
level
of
product
category
familiarity
leads
to
better
effectiveness
in
executing
tasks
on
a
commercial
website.
Furthermore,
in
the
precise
situation
of
an
online
search
task,
people
who
are
familiar
with
a
task
domain
tend
to
limit
their
attention
to
task-related
information,
which
minimizes
their
efforts
and
improves
their
efficiency
[54];
furthermore,
consumers
with
a
higher
level
of
category
knowledge
are
more
efficient
in
searching
online
information
[77].
Therefore,
we
also
hypothesize
the
following:
H3b.
A
higher
level
of
product
category
familiarity
leads
to
better
efficiency
in
executing
tasks
on
a
commercial
website.
Finally,
Mazursky
and
Vinitzky
[78]
stipulate
that
product
category
familiarity
decreases
online
shopping
duration.
Thus,
we
formulate
the
following
hypothesis:
H3c.
A
higher
level
of
product
category
familiarity
leads
to
less
time
spent
executing
tasks
on
a
commercial
website.
3.2.3.
Cognitive
absorption
Researchers
have
noted
the
importance
of
intrinsic
motivations
for
understanding
online
consumer
behavior
[79–81],
including
the
sense
of
flow
[56,57,82–84].
In
a
flow
state,
‘‘people
are
so
involved
in
an
activity
that
nothing
else
seems
to
matter’’
[82].
Although
flow
is
considered
valuable
construct,
Koufaris
[85]
identifies
flow
as
being
too
broad
and
ill-defined,
considering
the
many
ways
available
to
operationalize
and
test
flow.
In
response
to
these
concerns
and
to
examine
and
incorporate
holistic
online
experiences,
Agarwal
and
Karahanna
[86,
p.
665]
propose
a
cognitive
absorption
construct,
defined
as
‘‘a
state
of
deep
involvement
with
software,’’
which
affects
consumer
behavior
through
five
dimensions:
temporal
dissociation,
focused
immer-
sion,
heightened
enjoyment,
control
and
curiosity.
Specifically,
while
experiencing
temporal
dissociation,
a
user
loses
track
of
time
and
perceives
that
there
is
ample
time
left
to
complete
a
task.
M.S.
Ben
Mimoun
et
al.
/
Information
&
Management
51
(2014)
375–390
378
Heightened
enjoyment
refers
to
‘‘capturing
the
pleasurable
aspects
of
the
interaction’’
[86,
p.
673].
Focused
immersion
suggests
that
all
of
his
or
her
attentional
resources
are
focused
on
the
particular
task,
which
reduces
the
cognitive
burden
associated
with
task
performance.
Amplified
curiosity
suggests
that
the
act
of
inter-
acting
with
the
site
induces
excitement,
which
in
turn
should
serve
to
reduce
the
perceived
cognitive
burden
associated
with
the
task.
A
sense
of
being
in
charge
and
exercising
control
over
the
site
interaction
also
reduces
the
perceived
difficulty
of
the
task
[86–
88].
This
sense
of
perceived
control
is
also
similar
to
the
emotional
response
of
dominance
in
environmental
psychology
[89]
or
a
feeling
of
being
‘‘unrestricted
or
free
to
act
in
a
variety
of
ways’’
[90,
p.
19].
In
summary,
the
intrinsically
motivating
state
of
cognitive
absorption
lowers
the
perceived
cognitive
burden
associated
with
a
task
and
makes
the
user
more
willing
to
succeed
in
it,
which
should
improve
product
retrieval
effectiveness
on
a
website.
Thus,
we
propose
the
following:
H4a.
A
higher
level
of
cognitive
absorption
leads
to
better
effec-
tiveness
in
executing
tasks
on
a
commercial
website.
In
addition,
such
cognitive
absorption
also
may
increase
the
duration
of
the
task
execution,
through
amplified
curiosity
and
temporal
dissociation
dimensions:
H4b.
A
higher
level
of
cognitive
absorption
leads
to
more
time
spent
executing
tasks
on
a
commercial
website.
However,
we
do
not
expect
any
specific
effect
of
cognitive
absorption
on
efficiency
on
the
basis
of
the
literature.
3.3.
Website
characteristics
Finally,
site
design
elements
have
important
influences
on
information
searches
because
they
can
facilitate
or
hinder
access
to
important
information
[15,45,91–93].
Palmer
[94],
Madeja
and
Schoder
[95],
and
Ranganathan
and
Grandon
[4]
study
the
relationships
between
website
design
elements
and
website
performance
and
reveal
that
website
success
is
closely
associated
with
website
design.
3.3.1.
Understanding
website
design
and
website
complexity
Website
navigation
tools
can
create
or
prevent
pleasant
consumer
experiences
[19,96,97];
just
like
their
offline
counter-
parts,
online
stores
can
create
a
shopping
environment
that
affects
consumer
behavior.
The
environmental
characteristics
of
virtual
stores
consist
of
two
general
categories:
high
and
low
task-
relevant
cues.
The
former
include
‘‘all
the
site
descriptors
(verbal
or
pictorial)
that
appear
on
the
screen
which
facilitate
and
enable
the
consumer’s
shopping
goal
attainment’’
[97,
p.
179],
whereas
the
latter
are
‘‘all
site
information
that
are
relatively
inconse-
quential
to
completion
of
the
shopping
task’’
[97,
p.
180].
Product
descriptions,
prices,
terms
of
sale,
delivery
policies,
pictures,
product
reviews,
and
navigation
aids
are
highly
task
relevant
insofar
as
their
purpose
is
to
help
customers
achieve
their
shopping
goals
(i.e.,
such
information
has
utilitarian
aspects).
In
contrast,
background
patterns,
colors,
music,
and
animations
are
examples
of
low
task-relevant
cues
because
they
are
irrelevant
for
task
completion
but
may
have
a
positive
effect
on
the
hedonic
and
experiential
aspect
of
shopping
activity
[19,96,97].
Building
on
this
online
environment
model
[96],
Richard
[20]
posits
that
the
poor
organization
of
many
websites
is
due
to
a
profusion
of
hyperlinks
and
excessive
animation,
both
of
which
increase
the
site’s
complexity
[23].
Site
complexity
appears
to
offer
a
fundamental
characteristic
that
depends
on
colors,
the
number
of
links,
the
number
of
graphics,
the
home
page’s
length,
and
the
presence
of
animation
[21,22].
Most
research
on
complexity
is
based
on
Berlyne’s
[98]
idea
of
stimulus
complexity
[23],
yet
the
complexity
of
a
stimulus
remains
difficult
to
define
[22].
Even
Berlyne
[98,
p.
38]
describes
complexity
as
‘‘the
most
impalpable
of
four
elusive
concepts’’
(the
others
being
novelty,
uncertainty,
and
conflict)
and
defines
it
as
‘‘the
amount
of
variety
or
diversity
in
a
stimulus
pattern’’
[98,
p.
38].
The
complexity
of
a
stimulus
pattern
depends
on
the
number
of
distinguishable
elements,
dissimilarity
between
elements,
and
degree
to
which
several
elements
can
be
responded
to
as
a
unit
[22,98].
Furthermore,
Berlyne
[98]
identifies
two
dimensions
of
complexity:
structural
and
interactive.
Structural
complexity
reflects
the
range
of
structural
elements
and
irregularity
in
their
arrangement,
corresponding
to
distinct
information
cues
that
must
be
perceived
and
processed
to
perform
a
task
[23].
Interactive
complexity
arises
because
people
frequently
must
adapt
to
changes
in
the
cause-and-effect
chain
during
task
execution
[55].
Similarly,
Gupta
et
al.
[23]
propose
a
model
of
website
complexity
distinguishing
between
the
interactive
complexity
of
a
website
and
its
structural
complexity.
Interactive
complexity
refers
to
as
‘‘the
degree
to
which
users
find
the
hyperlinks
at
a
web
site
ambiguous
and
the
expectations
based
on
the
hyperlink
format
incongruous
with
the
ensuing
web
page.’’
[23,
p.
44];
moreover,
interactive
complexity
depends
on
elements
such
as
the
capacity
of
hyperlinks
to
allow
individuals
to
form
expectations,
the
uncertainty
between
pieces
of
information
presented
on
the
website
and
the
presence
of
banners
or
pop-up
windows
that
hinder
site
navigation.
To
describe
a
website’s
structural
complexity,
they
identify
two
subdimensions:
the
range
of
structural
elements
and
the
dissimilarity
of
these
elements.
They
depend
on
the
length
of
text,
the
number
of
animations,
the
number
of
hyperlinks
and
the
number
of
web
pages
to
assess
such
complexity
Likewise,
Geissler
et
al.
[22]
further
indicate
that
web
site
complexity
results
from
four
factors:
the
number
of
hyperlinks,
number
of
graphs,
page
length,
and
presence
of
animations.
Accordingly,
we
consider
two
design
factors:
the
level
of
label
(hyperlink)
abstraction
and
the
presence
of
animation.
The
level
of
hyperlink
abstraction
offers
a
typical
example
of
a
high
task-relevant
cue,
whereas
the
presence
of
animation
is
a
good
example
of
a
low
task-relevant
cue
[19,96].
Furthermore,
both
the
level
of
hyperlink
abstraction
and
the
presence
of
animation
affect
structural
and
interactive
complexity
[22,23].
In
particular,
the
level
of
hyperlink
abstraction
reflects
the
range
of
different
structural
elements,
and
the
presence
of
animation
is
connected
to
the
dissimilarity
of
structural
elements
[23].
Considering
interactive
complexity,
a
greater
level
of
label
abstraction
shapes
the
probabilistic
nature
of
the
links’
results,
whereas
the
presence
of
animation
increases
the
navigation
duration.
Concerning
structural
complexity,
level
of
hyperlinks
abstraction
is
related
to
the
range
of
different
structural
elements,
whereas
presence
of
animation
is
connected
to
the
dissimilarity
of
structural
elements
[23].
We
present,
in
what
follows,
their
potential
impacts
on
e-consumer
productivity.
3.3.2.
Level
of
label
abstraction
A
usable
online
catalogue
offers
a
product
classification
that
is
clear
to
customers
and
helps
them
locate
products
without
having
to
move
forward
or
backward
in
the
hierarchy
[7,28].
However,
hyperlink
abstraction
and
discrepancies
in
navigation
expectations
may
increase
website
complexity
[15,23]
because
hyperlinks
indicate
what
users
should
expect
and
allow
them
to
orient
themselves
and
predict
the
content
underlying
each
hyperlink
[93].
Hyperlinks
exhibit
two
main
levels
of
abstraction
[99]:
generic
and
concrete.
Generic
labels
describe
page
content
in
an
abstract
way,
whereas
concrete
ones
specify
the
page
content.
These
less
abstract,
concrete
labels
help
reduce
site
complexity
and
improve
information
retrieval
and
site
navigation
effectiveness
[99].
We
can
then
hypothesize
the
following:
M.S.
Ben
Mimoun
et
al.
/
Information
&
Management
51
(2014)
375–390
379
Table 1
Productivity in marketing literature.
Term Context Orientation Authors Definition Assessment of productivity Inputs Outputs
Productivity Economics
manufacturing
Firm-oriented Cox [32]
Sink [35]
Gro
¨nroos
and Ojasalo [36]
Productivity is the relationship
between the outputs generated (gods
and services) from a system and the
inputs provided (labor, capital, energy,
materials and data)
Productivity is the relationship of the
amount produced by a given system
during a given period of time, and the
quantity of resources consumed to
create or produce those outputs over
the same period of time [35].
The effective transformation of input
resources into outputs, the quality of
which is unchanged [36].
The manufacturing orientation is
related to production efficiency [36].
Production efficiency
Effectiveness
Labor
Capital
Energy
Materials
Data
Goods
Services
Value
Productivity Services Firm-oriented Gro
¨nroos and
Ojasalo [36]
Parasuraman [37]
Ibid.
The service orientation is related to
effectiveness [36].
Internal, efficiency
External efficiency
(=effectiveness)
Capacity efficiency
Labor
Material
Capital
Energy
Materials
Data
Service volumes
and economics
Service quality
Productivity Services Customer-oriented Parasuraman [37]
Anitsal [38]
The ratio of the service outputs
experienced by a consumer to the
inputs provided by that consumer as a
participant in service production [37].
Customer efficiency
Customer effectiveness
Time
Effort
Service performance
Satisfaction
Shopping productivity Retailing Customer-oriented Ingene [39],[11] Customer productivity is value-added
by the customer in the coproducer role.
Time
Money
Cognitive effort
Emotional effort
Products purchased
Information acquired
Pleasure in shopping
Client productivity Business-
to-business
Dual orientation
(firm and customer)
Martin et al. [40] High ‘‘client productivity’’ can be
regarded as timely, quality and value-
added inputs made to consulting
projects for transformation of such into
achievement of preset common
objectives.
Effectiveness Data
Decisions
Achievement
Customer efficiency
a
Services
Service
coproduction
Customer-oriented Xue and Harker [29]
Xue et al. [17]
Customer A is evaluated as more
efficient than consumer B if Consumer
A consumes fewer inputs to produce at
least the same amount of certain
outputs as Consumer B, or if Customer A
produces more outputs using at most
the same amount of certain inputs as
Customer B.
Customer efficiency can be pre
´cised
into three types of efficiency:
transaction, value, and quality
efficiency
Efficiency Time
Effort
Number of transactions
accomplished
Value per unit of cost
Financial assets
Intellectual assets
Customer
productivity
Self-service Customer-oriented Parasuraman [37]
Johnston and Jones [33]
Anitsal [38]
Anisal and Schumann [18]
Customer productivity reflects
efficiency and effectiveness
(excellence) of a customer [38].
Customer Efficiency
Customer Effectiveness
Time
Effort
Money
Emotional energy
Service
performance
Satisfaction
a
Xue et al. [17] use ‘‘efficiency’’ and ‘‘productivity’’ synonymously.
M.S.
Ben
Mimoun
et
al.
/
Information
&
Management
51
(2014)
375–390
380
H5.
An
increase
in
the
level
of
label
(hyperlink)
abstraction
nega-
tively
affects
e-consumer
productivity.
H5a.
An
increase
in
the
level
of
label
(hyperlink)
abstraction
negatively
affects
e-consumer
effectiveness.
Similarly,
Nadkarni
and
Gupta
[54]
and
Eroglu
et
al.
[97]
argue
that
clear,
unambiguous
hyperlinks
facilitate
efficient
scanning
of
goal-relevant
information
and
help
shoppers
move
through
a
site
quickly
and
efficiently.
Ambiguity
instead
is
a
source
of
distraction
for
website
users
that
hinders
their
information
search
[100]
while
also
creating
confusion
that
reduces
their
information
retrieval
capacity
and
search
efficiency
[65].
A
better
fit
between
search
terminology
and
hyperlinks
instead
increases
search
efficiency
[70].
Therefore,
we
postulate
the
following.
H5b.
An
increase
in
the
level
of
label
(hyperlink)
abstraction
negatively
affects
e-consumer
efficiency.
H5c.
An
increase
in
the
level
of
label
(hyperlink)
abstraction
positively
affects
the
duration
of
task
performance.
That
is,
an
increase
in
hyperlink
abstraction
levels
leads
to
the
loss
of
time
and
reduced
information
search
effectiveness
and
efficiency,
which
in
turn
reduces
e-consumer
productivity
[54,65,97,98].
3.3.3.
Animation
When
a
task
is
difficult,
the
presence
of
animation
on
the
website
has
a
negative
effect
on
user
performance
and
blocks
access
to
task-
relevant
information
[101].
We
can
first
hypothesize
that
H6.
The
presence
of
animation
negatively
impacts
e-consumer
productivity.
Schaik
and
Ling
[16]
investigate
the
effect
of
irrelevant
onscreen
material
in
web-based
systems
leads
to
greater
perceived
distraction
when
it
is
animated
rather
than
static
[16],
and
such
a
distraction
is
harmful
for
task
execution
[97,100].
Some
empirical
evidence
further
suggests
that
animation
distracts
users
from
important
details
or
subtle
information,
which
can
negatively
affect
information-
retrieving
efficiency
and
increase
task
performance
duration
[102,103].
This
discussion
leads
to
the
following
hypotheses.
H6a.
The
presence
of
animation
negatively
impacts
e-consumer
efficiency.
H6b.
The
presence
of
animation
positively
affects
task
perfor-
mance
duration.
However,
we
predict
no
effect
of
animation
on
the
ability
to
find
or
effectiveness
in
finding
a
product.
Because
prior
research
reveals
such
a
plethora
of
potential
influences
on
e-consumer
productivity,
integrating
all
of
them
into
a
single
testable
model
likely
would
be
unwieldy.
To
achieve
a
more
parsimonious
model,
we
examine
the
direct
and
interaction
effects
of
seven
key
variables:
task
nature
and
task
strategy
type
(situational);
presence
of
animation
and
level
of
hyperlink
abstraction
(site);
and
Internet
experience,
product
category
knowledge,
and
cognitive
absorption
(individual).
These
variables
appear
as
the
most
important
determinants
of
e-consumer
productivity
in
both
information
systems
and
marketing
research.
We
summarize
our
hypotheses
in
the
integrative
model
in
Fig.
2.
4.
Methodology
4.1.
Experimental
design
We
conducted
a
laboratory
experiment
with
a
2
(task
nature)
2
(presence
of
animation)
2
(hyperlink
abstraction)
factorial
design
to
test
the
proposed
model
and
hypotheses.
Three
factors
were
experimentally
manipulated:
the
nature
of
the
task,
presence
of
animation
and
level
of
labels
abstraction.
For
this
experiment,
we
developed
two
task
presentations
and
four
websites
and
randomly
assigned
the
350
participants
to
them.
3
A
series
of
pretests
informed
this
experimental
design.
We
initially
had
planned
to
improve
the
external
validity
of
our
study
by
using
an
existing
commercial
website
and
conducting
a
field
experiment,
but
in
an
initial
pretest,
we
discovered
several
sources
of
bias
(e.g.,
site
architecture,
Internet
connection
speed,
screen
size)
that
seriously
affected
internal
validity.
To
avoid
those
biases,
we
created
an
experimental
website
with
the
same
width
(number
of
subcategories)
and
depth
(number
of
products)
for
each
product
category.
We
conducted
the
experiment
using
a
LAN
connection
in
a
laboratory
equipped
with
the
same
computers,
browser,
configuration,
microprocessor,
and
screen
size.
Participants
were
randomly
assigned
to
one
experimental
condition.
Fig.
2.
Research
model:
determinants
of
product
retrieval
e-consumer
productivity.
3
We
chose
a
between-subjects
rather
than
within-subject
design
for
two
major
reasons.
First,
we
have
multiple
factors
and
a
large
sample
with
more
than
25
respondents
per
cell.
Second,
we
wanted
to
avoid
adding
to
the
length
of
the
experimental
procedure
(already
30
minutes
on
average).
M.S.
Ben
Mimoun
et
al.
/
Information
&
Management
51
(2014)
375–390
381