Content uploaded by Monika Kukar-Kinney
Author content
All content in this area was uploaded by Monika Kukar-Kinney
Content may be subject to copyright.
Electronic copy available at: http://ssrn.com/abstract=1985860
Mission Aborted: Why do Consumers Abandon their Online Shopping Carts?
Monika Kukar-Kinney
Angeline G. Close*
*Contact Author
Electronic copy available at: http://ssrn.com/abstract=1985860
1
Mission Aborted: Why do Consumers Abandon their Online Shopping Carts?
ABSTRACT
Despite placing items in virtual shopping carts, online shoppers frequently abandon their
online carts. We identify the cognitive and behavioral reasons behind the rampant cart
abandonment. Common wisdom blames breakdowns on the e-purchase stage, such as privacy
concerns and frustration with a lengthy checkout process. However, based on two studies
employing diverse consumer samples, we show that online shoppers’ considerations at the other
stages play a much larger role in cart abandonment. In particular, many customers use carts
simply as organizational and shopping research tools as well as for entertainment, which may
induce them to ultimately buy online or offline. Our framework extends theories related to buyer
behavior and reveals new inhibitors to purchasing in the digital age. As such inhibitors disengage
the customer and may result in a temporary or permanent customer loss, our research urges
scholars and online vendors to explain and assess consumer motivations for cart abandonment.
2
To more fully understand buyer behavior in the digital era, it is crucial to also examine
consumer “non-buying” behavior. Non-buying behavior is especially apparent while shopping
online, with many shoppers frequently not completing their purchases despite placing items in
their virtual shopping carts. Industry studies find that 88 percent of online shoppers have
abandoned their filled electronic cart in the past (Forrester Research 2005). To understand why
such rampant abandonment occurs, it is vital to examine consumers’ perceptions of virtual carts
and their motivations for placing items in the carts in the first place. The way that consumers use
their virtual carts is thought to differ both from the way they use shopping carts in a bricks-and-
mortar store setting as well as from the ways that managers and e-tailers intend. Identifying the
driving forces behind the consumer virtual cart use and the inhibitors to purchasing items in the
cart will better prepare e-tailers for creating more consumer friendly sites. Further, application of
this knowledge may lead to amplified conversion rates from online shopping to online buying.
While shopping cart abandonment is widespread, surprisingly few academic studies in
the marketing literature focus on this issue. Thus, we seek to fill the gap with theory-contributing
research on this phenomenon. Specifically, the objectives and contributions of the present
research are: 1) identifying drivers of electronic cart abandonment (i.e., a form of online non-
buyer behavior), and 2) explaining why such abandonment occurs. We provide support for our
framework via two empirical studies based on contrasting consumer samples. In turn, we discuss
areas for further scholarly research and contribute suggestions for increasing conversion rates.
Defining Shopping Cart Abandonment
Abandonment means “to give up, discontinue, withdraw from”, or “to leave, or desert”
(Random House Dictionary 2007). Hurwicz (1999) refers to electronic cart abandonment as
“when apparent planned purchases are never completed online”. Hurwicz’s definition implies
3
purchase intention. Intention is a cognitive state that reflects a buyer’s plan to buy in a specified
time period (Howard and Sheth 1969). While placing an item in a virtual cart is often a signal of
the consumer’s interest in the product, we cannot assume that consumers do so with an intention
of buying it during that shopping session. Thus, we define electronic cart abandonment as the
situation in which consumers place item(s) in their online shopping cart without making a
purchase during that online shopping session. For abandonment to occur, the shopper must have
placed one or more items in their cart before abandoning the cart in its entirety.
THEORY
Applying a Conceptual Framework to Online Shopping Behavior
Purchase Inhibitors. Throughout the purchase process, consumers may encounter or
experience a range of inhibitory situations, which may trigger them to abort the purchase, and -
in the online shopping context - abandon the shopping cart. Consistent with Howard and Sheth’s
Theory of Buyer Behavior (1969), there are five broad categories of inhibitory situations: 1) high
price/overall cost, 2) lack of availability, 3) time pressure, 4) the consumer’s financial status, and
5) social influences. Here, we apply a variety of these inhibitors to an online context (e.g.,
concern with the total cost, slowness of the website, entertainment aspect, etc.). In addition, we
introduce new inhibitors to include issues that are heightened in the context of online shopping,
such as privacy and security inhibitors, thus extending the existing framework. In Figure 1, we
identify various inhibitors applicable to online shopping process.
---- Insert Figure 1 about here ---
Just as with bricks-and-mortar shopping, consumers are thought to go through stages of
an online shopping process. Li and Chatterje (2006) propose a four-stage model of online
shopping, updating the Howard and Sheth (1969) theory of buyer behavior. We examine the
4
applicability of this framework and clarify some key areas where our proposed framework
differs. Broadly, consumers’ online purchase process consists of four stages: 1) information
search, 2) consideration, 3) evaluation, and 4) purchase decision (Li and Chatterjee 2006).
E-Search. Search is the process by which consumers select a particular element of a
stimulus in order to clarify the cognitions related to a brand or product as well as to satisfy
motives (Howard and Sheth 1969). For instance, a shopper may be concerned with the product’s
availability and has a motive to first gather information about the product and whether it is
available. 1 In the proposed information e-search stage, an online shopper browses through pages
of one or more websites. After gathering information from multiple online and offline sources,
online shoppers may enter the e-consideration stage.
E-Consideration. While Li and Chatterjee (2006) conceptualize the consideration stage as
placement of items into the virtual cart, we recognize that an e-shopper does not have to place an
item into the cart to be considering it. For example, some e-shoppers may instead place items of
interest in the separate wish lists. Alternatively, shoppers may bookmark the product webpage.
Thus, consumers’ consideration sets may be broader than what is implied by the items contained
in their cart. Hence, we consider the e-consideration stage as a consumer placing an item into a
cart or wish list, or book-marking it.
E-Evaluation. Online shoppers are thought to evaluate products in their consideration set.
Li and Chatterjee (2006) label the decision to view the shopping cart and/or start checkout as the
evaluation stage. However, because not all products under consideration may at this point
actually be in the cart, we cannot claim that beginning the online checkout is the evaluation
stage. In the bricks-and-mortar context, taking a final glance at the items in a grocery cart at the
1 If an item of interest is not available, the online shopper is often not able to put it into their shopping cart.
While product availability is an inhibitor that may impact placement of items in the cart in the first place, it
is not expected to affect cart abandonment—the focus of the research here.
5
register is well-beyond the evaluation stage. Thus, we classify the evaluation stage as the e-
shoppers analyzing the items in their evoked consideration set based on their unique purchase
criteria. At this stage, consumers compare and contrast their choice criteria, focusing only on
those attributes that are salient in their motives (Howard and Sheth 1969).
E-Purchase Decision. We define the e-purchase decision stage as the consumer’s
behavioral commitment to buy (i.e., pay for) the online item(s) or a decision against buying them
during a specific online transaction. For example, when e-shoppers use the “buy it now” feature
on eBay, the item automatically goes into their e-cart. The consumer then receives an email with
payment options (e.g., Paypal, their proprietary online payment system, cashier’s check, etc.).
Thus, we do not assume that all online purchases must be paid for online. When consumers begin
to enter their personal or financial information online, they demonstrate a more involved
commitment and purchase intent.
E-shoppers may go through the four stages of the purchase process out of sequence (Li
and Chatterjee 2006) for various reasons. For one, a consumer may not need additional product
information, skipping directly to the purchase decision. Second, a consumer may change one’s
mind and revert to information search. Third, a consumer may rethink the purchase and stop at
any point. The purchase process may also be picked up in an offline, bricks-and-mortar store.
HYPOTHESIS DEVELOPMENT
Based on the theoretical framework set above, we develop hypotheses about the factors
impacting the extent to which consumers abandon their shopping carts across different stages of
their online shopping process. Figure 2 shows how the proposed determinants of the shopping
cart abandonment fit within the theoretical framework, and Figure 3 displays the hypothesized
conceptual model.
6
---- Insert Figures 2 and 3 about here ----
Factors in E-Search Stage: Entertainment and Shopping Organization
Consumers may enter the e-search stage with diverse purposes or motives. First, they
may search because they intend to purchase during that online session. Second, they may check
products on various websites as a part of a purposeful ongoing search (Bloch, Sherrell, and
Ridgway 1986), but without plans to purchase the product immediately. Forsythe and Shi (2003)
suggest that consumers use the Web to search for information more than they do to make
purchases. In the ongoing search case, consumers may use the cart to help them organize items
of interest for a potential future purchase or to narrow down their selections about which they
want additional information. In a Forrester Research survey, 41 percent of participants placed
items in the online cart for research purposes (Magill 2005). Shopping carts also allow customers
to easily return to the item, as well as to evaluate a narrowed-down set of options. Third,
Mathwick, Malhotra and Rigdon (2001) show that Internet users often use the Web as a means of
entertainment or escapism. Thus, some consumers may place items in a shopping cart for purely
hedonic reasons. These experiential shoppers (Wolfinbarger and Gilly 2001) view shopping as a
fun and experiential activity rather than a means to obtain a product or service (Bellenger and
Korgaonkar 1980; Brown, Pope and Voges 2003; Holbrook and Hirschman 1982).
In all three situations, consumers may place items in the online cart; however, the
purposes for selecting items are thought to vary depending on their intentions. The abandonment
will more likely happen, the more likely the items were placed in the cart for purposes other than
immediate purchase, such as for entertainment (H1) or shopping organizational purpose (H2).
H1: The more that consumers use the online shopping cart as a form of entertainment
(i.e., out of boredom, for fun), the more likely they are to abandon it.
7
H2: The more that consumers use the online shopping cart as a shopping organizational
tool, the more likely they are to abandon it.
A relationship may also exist between the two motivations for cart use. As discussed above,
consumers may shop online with experiential in addition to goal-oriented motives (Wolfinbarger
and Gilly 2001). Wolfinbarger and Gilly (2001) show that online shopping brings increased
senses of freedom and control as compared to traditional shopping. The more likely that
consumers seek entertainment or to alleviate boredom by shopping online, the more likely they
may be to use the cart as a wish list or other shopping organizational purpose, as past research
would suggest that organizational endeavors provide online shoppers with a more involving
recreational activity than mere page browsing. Therefore,
H3: The more that consumers shop online in search of entertainment, the more likely they
are to use the cart for shopping organizational purposes (i.e., wish lists, marking items of
interest, narrowing the consideration set).
Factors in E-Consideration Stage: Shopping Organization and Waiting for a Lower Price
The contents of an online shopping cart provide an indication of the consumer’s
consideration set. However, as discussed earlier, the presence of items in the cart is not a perfect
indicator of the consideration set, as other potential choices may not have been added to the
shopping cart, or the items were placed into the cart solely for purposes other than immediate
purchase. In fact, the items may be placed into the cart for ease of further consideration, so that
consumers do not need to move among pages to compare products and prices. We propose that
the more that consumers use the cart as an organizational tool, the more likely that consumers
will consider attributes of a specific item, such as price, and whether a lower item price can be
obtained at a different time or place before committing to the purchase. Hence,
H4: The more that consumers use the e-shopping cart as a shopping organizational tool,
the more likely they are to wait for a lower price before buying.
8
While H4 tests the role of the price of a specific item, online shoppers may also be especially
sensitive to the aggregate total of all items in the cart, or to other evaluative factors that raise the
overall cost—including shipping and handling costs, tax if applicable, or other fees, which we
discuss next in the context of e-evaluation stage.
Factors in E-Evaluation Stage: Online Price and Sale Seeking
Price of the item and the associated shipping and handling fees (especially when
considered as excessive) have been shown as a purchase inhibitor in previous research (e.g., Xia
and Monroe 2004; Li and Chatterjee 2006). Many Internet users expect e-tailers to offer lower
prices on products (Maxwell and Maxwell 2001); yet, the overall cost of the final order may
discourage shoppers from purchasing (Lueker 2003; Magill 2005). Unlike traditional channels,
shipping and handling fees often appear at the end of the online transaction and add to the total
cost of the order. Seeing the total cost, consumers may decide not to purchase the items in their
cart until they find a lower price (H5). Rowley (2002) suggests that shoppers undergo a cognitive
process in which they collect information prior to making a purchase decision. These consumers
may add items to their cart as a means of comparing prices to find one that meets or is below
their reference price, which is a function of previous and current price information (Monroe
2003). If customers expect that an item will go on sale or anticipate that a lower price is available
elsewhere, they may abandon their shopping cart (H6). Therefore, we propose:
H5: The more that online shoppers are concerned about the total cost of the order (i.e.,
cost of goods in cart, shipping charges, sales taxes, other fees), the more likely they are to
wait for a lower price.
H6: The more that online shoppers tend to wait for a lower price, the more likely they are
to abandon their current e-shopping cart.
Factors in E-Purchase/Check-Out Process Stage: Privacy Concerns, the Length of the
Checkout Process, and Frustration with Loading Time
9
Scholars have identified privacy and security concerns as reasons why consumers avoid
the Web (Laroche et al. 2005, Xie, Teo and Wan 2006). In fact, security of personal and financial
information is a top concern of online shoppers (Miyazaki and Fernandez 2001). When a website
does not meet their privacy and security expectations, concern may became prevalent especially
during the checkout process, which requires the consumers to enter personal and financial
information. This concern may lead consumers to abort the online purchase process. Therefore,
we propose that consumers’ privacy and security concerns about a website are associated with an
increased amount of electronic shopping cart abandonment.
H7: The more that online shoppers are concerned about the privacy and security at a
website, the more likely they are to abandon their e-shopping cart.
Just as online shoppers expect security, they also expect convenience (Rohm and
Swaminathan 2004). In fact, many consumers shop online because physical shopping is
inconvenient (Seiders, Berry and Gresham 2000; Wolfinbarger and Gilly 2001). However, just as
traditional shopping checkout can be frustrating, frustration can develop during an online
checkout process. The requirements needed to complete the purchase, such as logging-in and
entering a password (Magill 2005) or creating a new account may cause consumers to terminate
the purchase process. Hence, we expect that the higher the level of consumers’ frustration with
the length of the purchase process, the greater the shopping cart abandonment.
H8: The more frustrated online shoppers are with the length of the purchase process, the
more likely they are to abandon their e-shopping cart.
Another source of frustration may stem from the excessive time a website may take to
load. If the pages take long to load, consumers may become impatient about waiting through
each step of the purchase process and may consequently seek other channels to buy the product
(Koloszyc 1999). Due to an increasing availability of high-speed Internet access, consumers are
10
becoming accustomed to fast loading times, and hence may become quickly frustrated when
encountering slow webpages. Thus, the greater the frustration with the loading time, the more
frustrated the consumers will be with having to go through all the steps of the purchase process
(H9). Moreover, if a consumer has already entered personal and financial information and it takes
long for the submitted information to load, one may become concerned about the security of the
information in transit. Therefore, a long webpage loading time may also translate into a higher
level of consumer concern about the security of their information (H10).
H9: The more frustrated online shoppers are with the website loading time, the more
likely they are to be frustrated with the length of the purchase process necessary to
complete the online transaction.
H10: The more frustrated online shoppers are with the website loading time, the more
likely they are to have a greater concern about their privacy and security while shopping
at the site.
EMPIRICAL RESEARCH
Study 1: Methods and Measures
In Study 1, we used a paper and pencil survey to test the proposed hypotheses. The
survey was pre-tested and administered to 183 undergraduate students (59 percent males) at a
private east-coast university. Students are savvy Internet users who frequently buy online and are
thus judged to be a suitable population. Eighty-seven percent of participants reported visiting
online stores at least once a month, while 55 percent reported visiting them at least once a week.
Once they visited an online store, the participants reported to actually buy from the visited store
on average twenty percent of the time. To satisfy the screening criteria, the respondents had to
have shopped online at least once during the preceding six-month period. After applying this
screener, the final sample consisted of 168 respondents.
11
We measured the dependent variable of interest, the extent of online shopping cart
abandonment, with the question: “How frequently do you abandon your online shopping cart
after having placed something in it during the same Internet session?”. The participants
responded by stating the percentage of time (0–100 percent). The average abandonment rate was
33.48 percent (SD = 30.43). This finding is consistent with Oliver and Shor (2003), who found
an e-cart abandonment rate of 32 percent. We adopted measures for independent variables or
developed them for this study and provide details on measurement in Table 1.
---- Insert Table 1 about here ----
Study 1: Analysis and Results
To determine the measurement properties of the scales, a confirmatory factor model
including all theoretical constructs was assessed by maximum likelihood estimation in AMOS
(see Anderson and Gerbing 1988). The standardized item loadings ranged from .64 to .95 and
therefore displayed sufficient item validity and reliability. All inter-construct correlations were
significantly lower than one, satisfying the test of discriminant validity. The constructs exhibited
sufficient reliabilities ranging from .75 to .86. We show the construct reliabilities in Table 1, and
inter-construct correlations in Table 2. All goodness-of-fit indices (χ2(111)=140, p=.03;
incremental fit index [IFI]=.98, Tucker-Lewis index [TLI]=.97, confirmatory fit index
[CFI]=.98, root mean square error of approximation [RMSEA]=.039) met or exceeded the
recommended cut-off criteria (Hu and Benter 1999), and therefore the model fits the data well.
To test the conceptual model, we employed latent variable structural equation modeling
(LVSEM) with maximum likelihood estimation in AMOS (see Figure 3). LVSEM was chosen
because it helps control for measurement error, can improve ways to measure reliability and
validity, and can help evaluate more complex inter-relationships simultaneously (MacKenzie
12
2001). While the overall fit of the model was significant (χ2(126)=176, p<.01), additional
goodness-of-fit indices (IFI =.96, TLI=.95, CFI=.96, RMSEA=.049) all were above acceptable
levels indicating that the model fits the data well (Bagozzi and Yi 1988). The model explains 28
percent of variance in the extent of shopping cart abandonment. Summary results for the tested
model and the standardized structural path parameter estimates are presented in Figure 3 and
Table 3 and are discussed next2.
---- Insert Tables 2 and 3 about here ----
The results show that the more the participants place items in the cart for entertainment purposes,
the greater is their extent of cart abandonment (β=.22, t=2.78, p<.01). In turn, the greater the use
of the cart for entertainment purposes, the more likely respondents also use the cart for
organizational purposes (β=.49, t=5.84, p<.01). Thus, H1 and H3 are supported. We predicted that
the greater the consumer’s extent of using the e-cart for organizational purposes, the more likely
the consumer will abandon the shopping cart (H2) and the more likely he or she will wait for a
sale or a lower price before purchasing the item(s) in the cart (H4.). Our results support these
hypotheses (H2: β=.24, t=2.49, p<.01; H4: β=.40, t=4.34, p<.01). The findings further show that
the greater the concern about the cost of the order, the more likely consumers will wait for a
lower price (H5: β=.41, t=4.68, p<.01), providing support for H5. Furthermore, the greater the
tendency to wait for a lower price, the greater the extent of the abandonment (H6: β=.24, t=2.81,
p<.01). The next set of hypotheses predicts that both the extent to which the consumer is
concerned about privacy and security at the website (H7) and the length of the purchase process
(H8) will lead to a greater extent of abandonment. As expected, the data support H7 (β =.13,
t=1.77, p<.05) and H8 (β=.14, t=2.00, p < .05). Last, we hypothesized that the more consumers
2 Given that the direction of the hypothesized relationships was predicted in advance, one-sided t-tests were used to
test the hypotheses.
13
get frustrated with the time it takes for pages to load, the more frustrated they will be with the
number of steps that they must go through before completing an online purchase (H9: β=.36,
t=4.62, p<.01), and the more concerned they will be about their privacy and security at the
website (H10: β=.19, t=2.17, p<.05). The findings provide support for both of these hypotheses.
Study 1: Evaluation of Alternative Models and Discussion
While all of our hypotheses were confirmed in Study 1, we also ran several alternative
competitive models based on potentially competing arguments to determine whether the
proposed model best fits the data. Three competitive models were estimated and are discussed
next. In the first model, the constructs of waiting for a lower price and concern with the cost of
the order were combined because they both contain a price/cost aspect (χ2(129)=285; p<.001;
IFI=.87, TLI=.84, CFI=.86, RMSEA=.085). In the second model, frustration with the loading
time and frustration with the length of the purchase process were combined, since they both
indicate frustration with the shopping process (χ2(144)=265; p<.001; IFI=.90, TLI=.88, CFI=.90,
RMSEA=.071). In the third model, three additional paths were added to the proposed model
(Figure 3); each path originated in frustration with loading time, and they ended in: wait for sale,
organizational purposes, and shopping cart abandonment (χ2(123)=173; p<.001; IFI=96,
TLI=.95, CFI=.96, RMSEA=.049). These paths were added because one could argue that slow
loading time should impact each of the four stages of the online shopping process and hence, all
endogenous variables. As seen above, the first two models show a significantly worse fit than the
proposed model. The third model has a similar fit with regards to most goodness-of-fit indices;
however, the change in the chi-square per change in degrees of freedom is not significant. Hence,
the fit of the third competitive model is not any better than the fit of the original model.
Consequently, the original conceptual model is preferred as it is more parsimonious.
14
While Study 1 advances theoretical and empirical knowledge of the inhibitors across the
stages of online buyer behavior, thus shedding light on online non-buyer behavior, it does
possess some limitations. First, the nature of the student sample may limit the generalizability of
findings to other consumer populations. Second, the online shopping cart abandonment was
measured with a single item continuous scale, thus we could not control for the measurement
error associated with this variable. Study 2 was conducted to address these two limitations.
Study 2: Methods and Measures
Study 2 replicates Study 1 findings with a more heterogeneous consumer sample, an
online method, and additional measures of cart abandonment. We conducted an online survey
using a snowball sample of 247 adults (44.1 percent males). The respondents were recruited by
undergraduate students in a large Southwestern University, who were each asked to invite up to
five adults to join the study in exchange for extra credit. Sixteen percent of the sample was 20
years or younger, 47 percent was 21-30 years, 20 percent was 30 to 40 years, and 17 percent was
older than 40 years. Thus, the sample was substantially more heterogeneous than the sample
employed in Study 1. Twenty-nine respondents did not satisfy the screening criteria of having
shopped online within the last six months, and were consequently removed from analysis,
resulting in the final sample size of 218. 80.2 percent of respondents reported visiting online
stores at least once a month, while 36.8 percent visited them at least once a week. Once they
visited an online store, participants reported purchasing from it on average 36.7 percent of the
time. Thus, this older sample visits the online stores slightly less frequently, but noticeably more
frequently buys after visiting compared with the younger sample in Study 1.
In Study 2, we measured cart abandonment with four items using a seven-point frequency
scale, anchored at 1=never and 7=always (see Table 1). The scale reliability was .90, and the
15
individual factor loadings were all above .75. The reported e-cart abandonment rate, measured
by the single item used in Study 1, was 34.19 percent (SD = 30.99), consistent with the rate of
33.48 percent obtained in Study 1. Measures of other variables were identical to Study 1.
Study 2: Analysis and Results
As in Study 1, the confirmatory factor model displayed a good fit with the Study 2 data
(χ2 (164)=212, p=.007; IFI=.98, TLI=.97, CFI=.98, RMSEA=.037). The only difference was
that electronic shopping cart abandonment was now represented as a latent variable, measured
with the four items. Scale and item reliabilities were all above the recommended levels and are
shown in Table 1, while Table 2 displays the inter-construct correlations.
Next, we tested the structural model (see Figure 3). The overall fit was significant (χ2
(179)=301; p<.001); however, other goodness-of-fit measures indicated a good fit (IFI=.95,
TLI=.94, CFI=.95, RMSEA=.056) (Bagozzi and Yi 1988; Hu and Bentler 1999). Using the latent
cart abandonment variable, the model explains 28.3 percent of variance in the extent of
consumers’ shopping cart abandonment (as compared to 24 percent if the observed, i.e., single-
item cart abandonment variable were used). Summary results for the tested model and the
standardized structural path parameter estimates are presented in Table 3 and are discussed next.
The results of Study 2 replicate the initial findings with a more diverse sample. The
findings show that the more the respondents place items in the cart for entertainment, the greater
is their extent of cart abandonment (β=.15, t=2.23, p<.05), supporting H1. We next predicted that
the greater the extent of using an e-cart for organizational purposes, the more likely the consumer
will abandon the cart (H2). Indeed, data support this hypothesis (H2: β=.31, t=3.94, p<.01).
Further, the greater the use of the cart for entertainment, the more likely respondents also use the
cart for organizational purposes (β=.35, t=4.78, p<.01), in support of H3. We also expected that
16
the greater the use of the cart for organizational purposes, the more likely the consumer will wait
for a sale or a better price before purchasing the item(s) in the cart (H4). Our results support these
expectations (H4: β =.15, t=2.39, p<.05). The findings further show that the greater the online
shoppers’ concern about the cost of the order, the more likely they will decide to wait for a sale
before buying the item(s) in the cart (H5: β=.66, t=9.05, p<.01), supporting H5. Moreover, the
greater the tendency to wait for a lower price, the greater the extent of cart abandonment (H6: β
=.15, t=2.01, p<.05). Next, we predicted that both the extent to which consumers are concerned
about privacy and security at the website (H7) and the length of the purchase process (H8) will
lead to a greater extent of shopping cart abandonment. As expected, the data support H8 (β=.25,
t=3.75, p<.01); however, the findings are not as predicted for H7 (β = -.17, t = -2.30, p<.05). The
results show that the greater the concern about the privacy and security of the site, the lower the
abandonment rate. H7 is therefore rejected. The final set of hypotheses proposed that the more
consumers get frustrated with the time it takes for pages to load, the more frustrated they will be
with the number of steps they have to go through before completing an online purchase (H9:
β=.52, t=8.17, p<.01), and the more concerned they will be about their privacy and security at
that website (H10: β=.40, t=5.10, p<.01). Both hypotheses are supported.
Using the same single item measure of shopping cart abandonment as in Study 1, the
results are very similar: the model fit is similar and the direction and significance of the results
for hypotheses tests above are consistent.
Study 2: Discussion
The Study 2 findings are consistent with Study 1 with respect to nine of the ten
hypotheses. H7 is supported in Study 1, suggesting that the greater the consumer’s privacy and
security concerns at a website, the more likely the consumer will abandon the cart. However,
17
Study 2 findings indicate the opposite, i.e., a significant negative relationship between
consumers’ concern about the privacy and security and the extent of their shopping cart
abandonment. A possible reason for this finding is that the older sample in the second study was
found to be more risk averse, as it on average displayed a higher level of privacy and security
concerns (Mean privacy/ Study 2 = 5.96) than the younger sample employed in Study 1 (Mean privacy/
Study 1 = 5.09). Thus, given the high level of privacy and security concerns, if such concerns do
arise with regards to the specific website, these consumers should be less likely to shop there and
hence less likely to place items in the website’s shopping cart to start with. Because of the
importance of privacy and security to these consumers, they are likely to place items into carts
only at the websites they trust, and consequently, should be less likely to abandon the carts due
to security concerns. Hence, a negative relationship between the extent of privacy concerns and
shopping cart abandonment in Study 2.
SYNTHESIZED DISCUSSION
Theoretical Contributions and Implications for Consumer-Based Retailing and E-tailing
The present research explicates a theoretical model of the determinants of consumer
online shopping cart abandonment and their inter-relationships, thus building a theory of non-
buyer behavior in the context of online shopping. We positioned the variables impacting
shopping cart abandonment within a theoretical framework of the four stages of buyer behavior.
Using this framework, the present research extends past knowledge by identifying consumers’
tendencies to place items in the shopping cart for reasons other than immediate purchase (i.e., for
research and organization and for entertainment value of the process) as important drivers of
abandonment. As a second area extending the foundation set by Howard and Sheth (1969), we
identify inhibitors that Internet shopping brings out to the buying process (see Figure 1). With
18
our studies, we extend the framework to include new issues that may be heightened in the
context of online shopping, such as privacy and security inhibitors.
In addition to contributing knowledge to the theoretical domains set by Howard and
Sheth (1969), we also calculated total effects of all variables on electronic shopping cart
abandonment in order to determine which variables are (after accounting for both their direct and
indirect effects) most important influencers of abandonment. As Table 4 shows, using the cart for
entertainment purposes and as an organizational tool are clearly the most important, with their
total standardized effects ranging from .33 to .34 for using the cart for organizational purposes
and between .27 and .38 for using it as a form of entertainment across the two studies.
Conventional wisdom suggests that electronic cart abandonment is a “bad thing” because
it lowers shopping transaction conversion rates or it may imply a non-consumer friendly site
(Hoffman and Novak 2005). Scholars have also used it as a measure of consumer dissatisfaction
(Oliver and Shor 2003). However, our studies show that consumers often leave items in their
virtual shopping cart for reasons other than dissatisfaction with the product, the e-tailer or the
purchase process. Thus, we conclude that shopping cart abandonment can also be positive for
consumers and retailers. For example, placing an item in an e-cart for organizational purpose
may serve as a measure of awareness, interest, desire, or future purchase intent. Researchers have
made the assumption that when a shopper chooses not to purchase the items in their online cart
immediately, the items represent a lost sale (New Media Age 2002). However, online cart
abandonment does not necessarily mean that the consumer will never make the purchase; rather,
consumers may plan to delay their purchases or purchase decisions. Analysis of additional
questions collected in Study 2 shows a strong positive correlation between the consumer placing
items in the cart for organizational purposes and one’s intent to purchase the item later (ρ = .61, p
19
< .01) as well as with one’s intent to later decide whether to make the purchase (ρ = .66, p < .01).
We find that online shoppers are accustomed to using their cart to assist them in the e-
consideration and e-evaluation stages by providing an organized place for their desired items, a
place to “store” items, a wish list, and as a tool to track prices for a possible later purchase.
The abandoned carts also provide e-tailers with vast information on their customers’
consideration sets. Items in a cart are indicative of a consumer's interest and provide an
opportunity for e-tailers to gather psychographic information. Moreover, many e-tailers have a
brick-and-mortar retail counterpart. Past studies have not included the crucial notion that e-
shoppers who leave their e-cart may have intentions to purchase that item from the company’s
retail store. Thus, the company may make the sale after all, and purchases may be a result of
online browsing and the organizational cart use. Analysis of additional data in Study 2 shows a
positive correlation between using the cart as a shopping organization tool and one’s intent to
purchase the item(s) from a land-based store (ρ = .19, p < .01). Thus, from a multi-channel
management perspective, cart abandonment should also not be viewed as a lost sale.
Interestingly, a key finding is that consumers often use e-carts as e-wish lists, despite how
e-tailers intend their shoppers to use these tools. An e-wish list is a separate feature intended to
serve as a pre-shopping list, a tool to organize the consideration set, a list of desired items, or a
list to share with others (e.g., for their birthday). While e-wish lists are for desired goods that are
not necessarily intended purchases at that time, e-tailers provide the e-carts to assist in gathering
goods for immediate purchase. Despite many sites offering both tools, we find that many online
shoppers ignore the provided wish-list tool, instead using the e-cart for both purposes.
We also find that some online shoppers set out to place items in their cart as a form of
entertainment. E-shoppers may get the thrill of enacting the shopping rituals, and satisfying
20
impulses to shop without the potentially negative consequences of buyers’ remorse and impulse
buying. Placing sought-after (perhaps unobtainable) items in a virtual cart may provide
consumers a chance to achieve feelings of willpower, control, and satisfaction without having to
pay for the items. In this respect, abandonment may be a saving grace for those consumers who
seek the thrill of shopping, yet have too limited resources to purchase the selected items. Even
though the final result for these consumers may not be a purchase, these consumers should be
still likely to spread positive word of mouth about the e-tailer and their experience at its website.
Another important driver of shopping cart abandonment was consumer’s tendency to wait
for a lower price (total standardized effects of .24 and .15 in Studies 1 and 2, respectively), while
concern about the total cost of the order was less pronounced (total standardized effect of .10
across both studies). This finding further highlights that items left in the cart may not necessarily
represent a lost sale, but rather an opportunity to make the sale in the future by sending a
promotional offer to the consumer, providing lower or free shipping on the item(s), or sending a
reminder email about the items in the shopping cart when the price has been lowered. Rewards
such as discounts or coupons (Xie, Teo and Wan 2006) may also encourage those consumers
who have a tendency to wait or search for a lower price elsewhere to complete the purchase3.
Further, across studies, frustration with the length of the purchase process was also found
to increase the frequency of abandonment (total standardized effects of .14 and .25 across the
two studies). Thus, retailers need to strive to make online purchases as hassle-free as possible.
Not requiring consumers to create a username and password when shopping for the first time and
providing one-click buying options for returning customers (e.g., Amazon.com) may help
3 The e-tailers do need to be aware that for those consumers who do not have the promotional coupon code, but are
given an option to enter it, the cart abandonment may actually increase (Oliver and Shor 2003). A personalized offer
or coupon redemption option appearing only once the customer has been identified by the website may be a solution
to this problem.
21
consumers through the purchase process as quickly as possible and may reduce concerns about
the privacy of stored personal information and the security of information in transit.
E-tailers also should address consumers’ concerns with regards to privacy and security
offered by the website. In the first study, privacy and security concerns led to an increased
frequency with which consumers abandoned their carts (standardized total effect =.13). We
speculated that in the second study, the average level of concern was high enough to even
prevent consumers from placing items in the cart (standardized total effect on abandonment = -
.17). To encourage consumers to place items in their carts, as well as to decrease the level of
consequent cart abandonment, websites could incorporate customer testimonials (Laroche et al.
2005). In addition, e-tailers should provide evidence that transactions are secure, such as privacy
and security seals, in order to increase consumers’ trust.
Finally, Magill (2005) suggests that the number of lost sales from loading time is
decreasing due to conversion from dial-up to broadband. Supporting this proposition, frustration
with the webpage loading time was the least important of all predictor variables in the model
(total standardized effect on shopping cart abandonment of .07 and .06 across the two studies).
For many participants, slow loading time was not an important determinant of abandonment,
likely due to their accessibility to a high-speed connection.
---Insert Table 4 about here ----
Limitations and Directions for Future Research
While the present research offers important contributions to both theory and practice, we
acknowledge some limitations. We employed samples from two diverse populations based in the
United States. We encourage other scholars to further test the model using different populations.
For example, Kuhlmeier et al. (2005) found that consumers from different countries have
22
varying levels of perceived risk. Therefore, investigating factors associated with cart
abandonment in a multinational context could help determine how to satisfy online shoppers not
only in the U.S., but across the world as well.
Second, we focused on survey data, which is self-reported by online shoppers. Although
the percentage of abandonment obtained in this research was extremely consistent with both
prior research and across the two studies using diverse samples (33 and 34 percent in Studies 1
and 2, as opposed to 32 percent reported by Oliver and Shor 2003), click-stream data would
allow for a supplementary accurate measurement of shopping cart abandonment. Although such
data is often proprietary and site-specific, it would be valuable to examine it in future research.
Combining click-stream data with a customer survey based on the present research findings
would further enable scholars to explain reasons for a specific instance of cart abandonment.
Many other fruitful opportunities for scholarly research exist to expand these findings.
For instance, Laroche et al. (2005) concluded that perceived risk differs from one category of
products/services to another. Consequently, a study in which participants indicate cart
abandonment for specific types of products or services would also be valuable. While we
included exogenous variables such as time pressure (i.e., frustration with the loading time and
the length of purchase process), price/cost, and social influences (i.e., entertainment), we
encourage scholars to further extend our findings by incorporating other exogenous variables
such as culture, class, and personality traits. Intriguing areas of future research are also studying
motivations for consumer shopping cart use and associated utilities consumers derive from
placing items in the cart as well as the relationship between the use of the virtual cart and offline
future purchase behavior from a multi-channel management perspective.
23
Figure 1: An Extension of Inhibitors to the Online Shopping Process
Known Inhibitors
High Price
• Price of item(s) too high
• Shipping costs too high
• Handling fees too high
Lack of Availability
• Of the product (e.g., sold out)
• To online access
• To the e-tail site
• Of shipping to the geographic area
(e.g., no international shipping)
Privacy & Security Issues
• With the Internet in general
• With specific e-tail sites
• Privacy of specific purchases
• Privacy of personal
information
• Security of financial information
Time Pressure
• Product is needed at time of
purchase
• Delivery too slow
• The online purchase process too
slow
• Webpage loading time too slow
Technology Glitches & Issues
• The Internet service provider,
computer, or printer does not work
• The website does not work (e.g., down
for maintenance)
• The payment system does not work
• The online sale or promotion code
does not work
Shopper’s Financial Status
• No access to accepted payment
methods (e.g., Paypal, e-checks)
• Limited availability of funds
Social Influences
• Online shopping not available
(e.g., for a gift on a registry)
• Family/friends influence not to buy
online
• Lack of entertainment/boredom
Emerging Inhibitors
24
E-Search Stage:
Entertainment tool
Organizational tool
E-Consideration Stage:
Organizational tool
Wait for sale
E-Evaluation Stage:
Total cost
Wait for sale
E-Purchase Stage:
Webpage loading time
Length of purchase process
Security of info
Figure 2: E-Purchase Process and the Determinants of Electronic Shopping
Cart Abandonment
Online
shopping cart
abandonment
25
Figure 3: Key Determinants of Consumer Shopping Cart Abandonment:
The Conceptual Model and Results (Study 1 [Study 2])
Total cost
Loading
time
Frustration
with purchase
process
Security of
info
Entertainment
Organizational
tool
Shopping
cart
abandonment
Wait for
sale
.36*** (.52***)
.19** (.40***)
.41***(.66***)
.22*** (.15**)
.24*** (.31***)
.40***(.16***)
.24*** (.15**)
.49*** (.35***)
.13** (-.17**)
.14** (.25***)
26
Table 1: Measures
Construct items and scale reliability
Item reliability
(Study 1)
Item reliability
(Study 2)
Shopping cart abandonment - Study 1
How frequently do you abandon your online shopping cart after having
placed something in it during the same Internet session (i.e., you do not
purchase the item(s) in your cart)? (open-ended; answers in %)
Shopping cart abandonment - Study 2 (α = .90) b
How often do you leave items in your online shopping cart without
buying them?
How often do you place an item in the online shopping cart, but do not
buy it during the same Internet session?
How often do you close the webpage, or log off the Internet before you
buy the item(s) in your online shopping cart?
How often do you abandon your online shopping cart?
N/A
.93
.79
.77
.87
Concern about privacy/security (α = .75 [.72]) a
I am concerned that someone will steal my identity.
I am concerned that the retailer will share my information with third
parties.
Internet privacy is important to me.
.73
.65
.81
.70
.85
.61
Frustration with length of the purchase process (α = .86 [.91]) a
I get frustrated with the time it takes to complete the Internet purchase.
I get frustrated with the amount of information I need to provide before
an Internet purchase can be completed.
I get annoyed with the number of steps I have to go through before my
purchase is complete.
.64
.90
.95
.80
.88
.96
Frustration with Webpage loading time (α = .75 [.78]) a
I feel annoyed when graphics delay the time it takes Web pages to load.
I get irritated when the Web pages are loading for more than a few
seconds.
N/A
N/A
Tendency to wait for a better/sale price(α = .82 [.80]) b
I decide to wait that the item will come on sale before buying it.
I decide that I may be able to find better sales at another online store.
I decide that I may be able to find better sales at a land-based store.
.80
.73
.78
.82
.81
.64
Concern about the cost of the order(α = .83 [.89]) b
I decide not to buy when I see the shipping charges for my order.
I decide not to buy when I see the amount of sales tax added.
I decide not to buy when I see the total amount at the checkout.
.76
.82
.78
.84
.82
.92
Using the cart for organizational purposes (α = .79 [.79]) b
I use the shopping cart as a “wish list” for myself.
I place items I am interested in the shopping cart.
I place items in the shopping cart so I can more easily evaluate a
narrowed-down set of options.
.75
.69
.79
.68
.70
.86
Entertainment value (α = .81 [.90]) b
I select and place items in the shopping cart for fun.
I select and place items in the shopping cart when I am bored.
N/A
N/A
Note: a These items were measured on a scale 1=strongly disagree, 7=strongly agree.
b These items were measured on a scale 1=never, 7=always.
27
Table 2: Construct Inter-Correlations
Shopping cart
abandonment
Privacy/
security
concerns
Length of
purchase
process
Loading
time
Wait for
sale
Cost of
order
Organization
tool
Entertainment
Shopping cart
abandonment
1
.02
.34***
.13*
.37***
.34***
.48***
.33***
Privacy/security
concerns
.19**
1
.47***
.40***
.14*
.18**
.24***
-.02
Length of
purchase process
.10
.09
1
.52***
.37***
.27***
.39***
.12*
Webpage
loading time
-.01
.19**
.36***
1
.19***
.13**
.23***
.05
Wait for sale
.44***
.14*
.03
.05
1
.70***
.43***
.33***
Cost of order
.24***
.18**
.17**
-.01
.51***
1
.47***
.33***
Organizational
purpose
.46***
.17**
-.21**
-.03
.51***
.36***
1
.34***
Entertainment
.38***
-.07
-.07
-.03
.27***
.17**
.49***
1
Note:* p < .10, ** p < .05; *** p < .01. Correlations obtained in Study 1 are reported below the diagonal, and
correlations from Study 2 are listed above the diagonal.
Table 3: Testing the Proposed Model Relationships
Study 1
N= 168
Study 2
N = 218
Hypothesis
Sign
Path from ! to
Standardized
estimate
t-value
Standardized
estimate
t-value
H1: +
Entertainment ! E-cart abandonment
.22***
2.78
.15**
2.23
H2: +
Organizational purpose ! E-cart abandonment
.24***
2.49
.31***
3.94
H3: +
Entertainment ! Organizational purpose
.49***
5.84
.35***
4.78
H4: +
Organizational purpose ! Wait for sale
.40***
4.34
.15**
2.39
H5: +
Concern about costs ! Wait for sale
.41***
4.68
.66***
9.05
H6: +
Wait for sale ! E-cart abandonment
.24***
2.81
.15**
2.01
H7: +
Privacy/security concerns ! E-cart abandonment
.13**
1.77
-.17**
-2.30
H8: +
Length of purchase process! E-cart abandonment
.14**
2.00
.25***
3.75
H9: +
Loading time ! Length of purchase process
.36***
4.62
.52***
8.17
H10: +
Loading time ! Privacy/security concerns
.19**
2.17
.40***
5.10
Goodness-of-Fit Statistics
Chi-square (d.f.)
176 (126)
301 (179)
IFI
.96
.95
TLI
.95
.94
CFI
.96
.95
RMSEA
.049
.056
**p < .05, ***p < .01.
Table 4: Total Standardized Effects on Shopping Cart Abandonment
Total
standardized
effects
Privacy/
security
concerns
Length of
purchase
process
Loading
Time
Wait for
sale
Cost of
order
Organization
tool
Entertainment
Study 1
.13
.14
.07
.24
.10
.34
.38
Study2
-.17
.25
.06
.15
.10
.33
.27
28
REFERENCES
Abandonment. Dictionary.com. Dictionary.com Unabridged (v 1.1). Random House, Inc.
http://dictionary.reference.com/browse/abandonment (accessed: February 13, 2007).
Anderson, James C. and David W. Gerbing (1988). Structural Equation Modeling in Practice: A
Review and Recommended Two-Step Approach. Psychological Bulletin, 103(4), 411-23.
Bagozzi, Richard P. and Youjae Yi (1988), On the Evaluation of Structural Equation Models.
Journal of the Academy of Marketing Science, 16(1), 74-94.
Bellenger, Danny N. and Pradeep K. Korgaonkar (1980). Profiling the Recreational Shopper.
Journal of Retailing, 56(3), 77-92.
Bloch, Peter H., Daniel L. Sherrell, and Nancy M. Ridgway (1986). Consumer Search: An
Extended Framework. Journal of Consumer Research, 31(1), 119-126.
Brown, Mark, Nigel Pope, and Kevin Voges (2003), Buying or Browsing? An Exploration of
Online Shopping Orientations and Online Purchase Intention, European Journal of
Marketing, 37(11/12), 1666-1684.
Burke, Raymond R. (2002), Technology and the Customer Interface: What Consumers Want in
the Physical and Virtual Store, Journal of the Academy of Marketing Science 30(4), 411–
432.
K. P. Chiang and R.R. Dholakia (2003), Factors Driving Consumer Intention to Shop Online,
Journal of Consumer Psychology 13 (1), 77–183.
Forrester Research (2005), “Rethinking The Significance Of Cart Abandonment”
Carrie A. Johnson, Charles P. Wilson, Sean Meyer, ed.
Forsythe, Sandra M., Shi Bo (2003). Consumer Patronage and Risk Perceptions in Internet
Shopping. Journal of Business Research, 56(11), 867-875.
Hoffman, Donna L. and Thomas P. Novak (2005). Beyond the Basics: Research-Based Rules for
Internet Retailing Advantage. eLab Press: Vanderbilt, TN.
Holbrook, Morris B., Hirschman Elizabeth C. (1982). The Experiential Aspects of Consumption:
Consumer Fantasies, Feelings, and Fun. Journal of Consumer Research, 9(2), 132-140.
Howard, John A. and Jagdish N. Sheth (1969). The Theory of Buyer Behavior. New York: John
Wiley.
Hu, Li-tze and Peter M. Bentler (1999). Cut-off Criteria for Fit Indexes in Covariance Structure
Analysis: Conventional Criteria Versus New Alternatives. Structural Equation
Modeling, 6(1), 1-55.
29
Hurwicz, Michael (1999). Closing the Web Deal: Answering Customer Queries. Data
Communications, 28(12), 21-22.
Koloszyc, Ginger (1999). Abandoned ‘Shopping Carts’ Pose Major Challenge for Internet
Retailers. Stores, 81(7), 41-44.
Kuhlmeier, David and Gary Knight (2005). Antecedents to Internet-Based Purchasing: A
Multinational Study. International Marketing Review. 22(4), 460-474.
Laroche, Michel, Yang Zhilin, Gordon H.G. McDougall, and Bergeron Jasmin (2005). Internet
versus Bricks and Mortar Retailers: An Investigation into Intangibility and its
Consequences. Journal of Retailing, 81(4), 251-267.
Lueker, Tom. Abandonment Surveys Help Boost Sales. Marketing News 2003; 37(24), 16-21.
Li, Shibo and Patrali Chatterjee (2006). Shopping Cart Abandonment at Retail Websites - A
Multi-Stage Model of Online Shopping Behavior, Working Paper.
MacKenzie, Scott B. (2001). Opportunities for Improving Consumer Research through Latent
Variable Structural Equation Modeling. Journal of Consumer Research, 28 (June), 159-
166.
Magill, Ken. Building a Better Shopping Cart. Multichannel Merchant 2005; 1(8), 18-
19.
Mathwick, Charla, Malhotra Naresh, and Edward Rigdon (2001). Experiential Value:
Conceptualization, Measurement and Application in the Catalog and Internet Shopping
Environment. Journal of Retailing, 77(1), 39-56.
Maxwell, Sarah and Nicholas Maxwell (2001). Channel Reference Prices: The Potentially
Damaging Effects of Napster. Proceedings of the 2001 Fordham University Behavioral
Pricing Conference: 32.
Miyazaki, Anthony D. and Ana Fernandez (2001). Consumer Perceptions of Privacy and
Security Risks for Online Shopping. Journal of Consumer Affairs, 35(1), 27-44.
New Media Age (2002). Online Store Layouts Confuse Shoppers. Dec.19, 14.
Oliver, Richard L. and Mikhael Shor (2003). Digital Redemption of Coupons: Satisfying and
Dissatisfying Effects of Promotion Codes. Journal of Product and Brand Management,
12(2), 121–134.
Rohm, Andrew J. and Vanitha Swaminathan (2004). A Typology of Online Shoppers Based on
Shopping Motivations. Journal of Business Research, 57(7), 748-57.
30
Rowley, J. (2002). Window Shopping and Browsing Opportunities in Cyberspace. Journal of
Consumer Behaviour, 1(4), 369-78.
Seiders, Katheleen, Leonard L. Berry, and Larry G. Gresham (2000), Attention, Retailers! How
Convenient is Your Convenience Strategy?, Sloan Management Review, 41(3), 79-89.
Wolfinbarger, Mary and Mary C. Gilly (2001). Shopping Online for Freedom, Control and Fun.
California Management Review, 43(2), 34-55.
Xia, Lan and Kent B. Monroe (2004), Price Partitioning on the Internet, Journal of Interactive
Marketing, 18(4), 63-73.
Xie En, Teo Hokc-Hai, Wan Wen (2006). Volunteering Personal Information on the Internet:
Effects of Reputation, Privacy Notices, and Rewards on Online Consumer Behavior.
Marketing Letters, 17(1), 61-74.