Content uploaded by Kurt Matzler
Author content
All content in this area was uploaded by Kurt Matzler
Content may be subject to copyright.
Consequences of Customer Confusion in Online Hotel
Booking
Kurt Matzlera
Martin Waigunyb
a Institute for Business Administration and Economics
Department for Marketing and International Management
University of Klagenfurt, Austria
kurt.matzler@uni-klu.ac.at
b eBusiness-Institute (biztec)
University of Klagenfurt, Austria
martin.waiguny@biztec.org
Abstract
Consumer Confusion is a relatively new concept in consumer behavior research. Increasingly,
consumers get swamped by a large variety of opportunities to choose, similar products and
information, and the increasing complexity of products and services. While previous studies
mostly explored the markets for convenience- and shopping-goods in the stationary retail
industry, this paper investigates the consequences of customer confusion in online booking of
hotels. In the first part of the paper, the concept of consumer confusion and its outcomes are
discussed. In the empirical part of the paper, based on a survey on 228 tourists, the causal
relationships between customer confusion and its consequences are tested using structural
equation modeling.
Keywords: consumer confusion; online buying behavior; choice behavior; online booking
1 Introduction
In previous years, products and services have proliferated at an enormous rate, and
customers have more options available than ever. The emergence of the Internet, as a
new distribution channel, even supported the extension of assortments and product
lines. Especially online-retailers tend to offer a wide variety of products to cover all
customers’ needs. The Internet also leads to a tremendous augmentation of
information access.
The increasing number of products available and the increasing amount of decision-
relevant information leads to a phenomenon called consumer confusion if information
is too similar, too complex, too ambiguous, and too much (e.g. Mitchell et al., 1997;
Mitchell et al., 1999; Rudolph et al., 2003; Turnbull et al., 2000; Walsh et al., 2002).
These trends can also be observed in the tourism industry. Besides an enormous
amount of information on hotels available from online-reservation-systems and
online-travel-agencies, most hotels have established their own homepages, too. From
a customer’s point of view it becomes nearly impossible to overlook the market as the
number of online offers increases steadily. Customer confusion seems to be relevant
also in the tourism industry. In this study, therefore, the authors investigate the
antecedences and consequences of customer confusion in the tourism industry with a
specific focus on online hotel booking.
2 The Concept of Consumer Confusion
2.1 Dimensions of consumer confusion
Consumer confusion is a relatively new concept in consumer research. Turnbull, Leek
and Ying (2000) define consumer confusion as “consumer failure to develop a correct
interpretation of various facets of a product/service, during the information processing
procedure. As a result this creates misunderstanding or misinterpretation of the
market” (p. 145). This situation where consumers are swamped with information and
lose orientation can be provoked by too similar, too complex, too ambiguous, and too
much of products/services and information about them, e. g. advertisements, product
descriptions etc. (Rudolph et al., 2003). Literature suggests that there exist three
dimension of consumer confusion (Mitchell et al., 1999; Mitchell et al., 2004; Walsh,
2002; Wiedmann et al., 2001):
• similarity confusion,
• overload confusion and
• unclarity confusion.
Similarity confusion is defined as “a lack of understanding and potential alteration of
a consumer’s choice or an incorrect brand evaluation caused by the perceived
physical similarity of products or services” (Mitchell et al., 2004). Similarity
confusion can be provoked by brand similarity (Walsh et al., 2002) when competitors
imitate the brand or when quality or product attributes of different alternatives are
identical. Furthermore, similarity confusion is also a result of similarity in
advertisements and commercial messages, i.e. in information provided (Kent et al.,
1994; Poiesz et al., 1989).
Overload confusion relates to the fact that consumers are confronted with an overly
information rich environment. This information overload inhibits the customer to
process the information and to fully understand, and be confident in, the purchase
situation (Mitchell et al., 2004). Information overload is caused by the increase of
alternatives and an increase of decision-relevant information on these alternatives.
Unclarity confusion finally, occurs when customers are “forced to re-evaluate and
revise current beliefs or assumptions about product or purchasing environment”
(Mitchell et al., 2004). Unclarity confusion, for example, emerges when customers
receive new, valid or false information that is contradictory or does not coincide with
present knowledge. Here, it is quality of information not the quantity of information
that leads to customer confusion (Wiedmann et al., 2001). Unclarity confusion is
typically the result of ambiguous, unclear or contradictory information (Mitchell et
al., 2004; Turnbull et al., 2000) about the products or too complex products (Cohen,
1999; Rudolph et al., 2003).
2.2 Consequences of consumer confusion
Customers respond to confusion, whether it is conscious or unconscious, with several
confusion reduction strategies. It is to note that customers who are aware they are
confused are feeling a higher risk within the purchase decision (Turnbull et al., 2000).
Hence, confusion reduction strategies and also risk reduction strategies are employed
(Turnbull et al., 2000; Wiedmann et al., 2001).
Mitchell & Papavassiliou (Mitchell et al., 1997) investigated confusion reductions
strategies for the watch market in the UK and Drummond (2004) for the market of
higher education. Eight forms of reduction strategies were found: 1) do nothing and
ignore confusion, 2) abandon the purchase 3) postpone the purchase, 4) clarify the
buying goals, 5) seek additional information, 6) narrow down the choice set by
important criteria, 7) share the decision, 8) delegate the decision. Other studies found
that consumer confusion influences post purchase behavior such as loyalty,
satisfaction, and trust (Walsh et al., 2002).
3 Consumer Confusion in online booking
In the last years, the Internet has become a major communication tool. In January
2003, the Internet universe was estimated to be more than 399 million people with
home Internet access. The average Internet user spends more than thirty minutes
during one surfing session, views over 40 pages per session and views almost 900
pages per month (Nielsen, 2003). Travel is one of the most popular products sold over
the Internet (O'Connor, 2001). For example, Yesawich (2000) reports that six out of
ten adults are interested in using the Internet to purchase travel services, Rayman-
Bacchus and Molina (2001) found that about half of those with Internet access
regularly browse the web for tourist information. Marcussen (2003) reports a
remarkable growth in the European travel market during 2002, with further growth
expected in 2003. Online travel sales increased 55% during 2001 to 2002, reaching
EUR 7.35 billion (i.e. 3.5% of the overall travel market). In the year 2003, a further
increase to about EUR 9.7 billion (or 4.4% of the market) was expected. Further
projections on the development of the overall online travel market size for Western
Europe for the year 2006 range from EUR 16.4 billion to 38.7 (Marcussen, 2003).
Hotels accounted for 13.5 % of the online market in 2002, ranking second after air
travel (60.0%).
It can be expected that all three dimension of consumer confusion can be found in
online booking. Especially within reservations systems (e.g. www.tiscover.com,
www.hrs.de, etc.) sub pages with hotel information look rather similar and for the
internet user it should be difficult to distinguish between hotels with the contents
presented. Product and information overload can also be expected. For example, a
search for hotels in the Austrian destination Tyrol on Tiscover (www.tiscover.com)
brings up over 100 different hotel offers, the German Hotel Reservation Service
(www.hrs.com) has over 130.000 hotels in its databases. A Google
(www.google.com) search for hotels in the Tyrol brings up over 100.000 different
links. So choice overload could be expected at a high level. Furthermore, hotels are
rather complex products and if information is presented ambiguously consumers’
confusion should increase.
In this study, therefore, the consequences of consumer confusion in online booking
are measured. First, it is tested whether the confusion dimensions as suggested in the
literature, can be found in online booking. Second, it is tested how consumer
confusion is related to four reductions strategies that are highly relevant in online-
business and tourism management. Mitchell & Papavassiliou (Mitchell et al., 1999)
argue that some of the reduction strategies mentioned above are highly interrelated:
“confused consumers can often involve another person (i.e. spouse, family member,
friend) in the purchasing decision or even delegate the task to them completely)” (p.
329). Therefore, in our study we did not distinguish between decision sharing and
delegating. “Do nothing” obviously can not be considered as a reduction strategy
because a person will do nothing only if the confusion is felt below the level what is
tolerable (Mitchell et al., 1997). “Narrow down the choice set” basically means that
customers rely on familiar brands (Rudolph et al., 2003). Therefore, in our study we
relable this reduction strategy to “rely on familiar brands”. Postponing the purchase in
our view is not an independent reduction strategy but is a side effect of other
strategies such as information seeking, delegating etc. In our study, we focus on
behavioural outcomes of customer confusion. “Clarify the buying goal” is a cognitive
reaction to confusion and leads either to abandon the purchase, search additional
information, delegate the decision or narrow down the choice set. Thus we investigate
four reduction strategies and propose the following hypothesis:
H1: Similarity, overload and unclarity confusion are related to additional information
seeking
H2: Similarity, overload and unclarity confusion are related to abandoning the
purchase
H3: Similarity, overload and unclarity confusion are related to delegating the decision
H4: Similarity, overload and unclarity confusion are related to relying on familiar
brands.
4 Study
4.1 Methodology
The following study builds on preceding studies by Walsh (2002) and Sproles and
Kendall (1986), who researched consumer confusion respectively overload confusion
with a questionnaire. The present survey contains item batteries for each dimension of
consumer confusion taken from these studies and were adapted to the context of
online booking: 5 items for similarity confusion, 4 items for overload confusion and 3
items for unclarity confusion. The reduction strategies “Seek additional information”,
“Abandon the purchase”, “Delegate the decision” and “Rely on familiar brands”
where each measured with two items taken from Mitchell and Papavassiliu (Mitchell
et al., 1997) using a five-point Likert scale (from “strongly agree” to “strongly
disagree”, see Appendix). In order to reduce the length of the questionnaire, only
those items have been taken that have been considered relevant. The scales than have
been purified calculating factor reliabilities and average variance extracted
eliminating those items that reduced composite reliability and average variance
extracted. Reliability is a measure of the internal consistency of the construct
indicators, depicting the degree to which they “indicate” the common latent construct
(Hair et al., 1998). Calculating the composite reliability (CR) of the constructs and the
average variance extracted (AVE) (Fornell et al., 1981) indicators that decrease
reliability can be identified. Thus, indicators with low factor loadings and high
measurement errors decrease reliability and such indicators have been removed.
Table 1. Demographics
Characteristic Frequency %
Male 119 52.19 Gender
Female 109 47.81
18-30 years 120 52.63
31-40 years 54 23.68
41-50 years 39 17.11
51-60 years 13 5.70
Age
61- 70 years 2 0.88
primary 4 1.75
middle 13 5.70
apprenticeship 47 20.61
high school 113 49.56
professional 5 2.19
Education
graduate 46 20.18
less than 5 hours 65 28.51
5 - 10 hours 70 30.70
10 - 20 hours 54 23.68
Average Internet usage per
week
more than 20 hours 39 17.11
Online-purchase experience yes 186 81.58
no 42 18.42
1 booking 73 33.33
2 - 5 bookings 105 46.46
5 - 10 bookings 23 13.07
Average amount of hotels
booked per year
more than 10 bookings 18 8.61
no 80 35.56
ca. 25 % 55 24.44
ca. 50 % 56 24.89
ca. 75 % 21 9.33
Percentage of online booked
hotels
all 13 5.78
The questionnaires have been handed out to tourists on Innsbruck airport while they
were waiting for their departure flight. Collecting data at the airport was considered
advantageous as guests who have recently booked a hotel were easily accessed.
Furthermore, tourists waiting for their flight should be motivated to participate in the
study as they have enough time to fill in a questionnaire.
Subjects were asked whether they had experience in online information seeking for
travel services. Only tourist who already had experience in seeking hotel information
on the internet were included in the study. 228 fully completed and usable
questionnaires were collected. Table 1 lists the demographics of this study.
4.2 Results
4.2.1 Model fit
The causal relationships between the constructs have been tested using structural
equation modelling with AMOS 5.0.
First, the overall fit of the observed data to the model was tested. The first test yielded
a chi-square value of 131.459 (df = 90, p = .003; χ²/d.f. = 1.461), implying that the
model was not adequate. However, literature suggests that chi-square is only
recommended with moderate samples (Hu et al., 1999), e.g. 100 to 200 (Tabachnik et
al., 1996), as with larger sample sizes, trivial differences become significant.
A more appropriate fit index is the root mean square error of approximation
(RMSEA). It evaluates approximate rather than exact fit of the model and it attempts
to correct for the tendency of the Chi-square statistic to reject any model with a large
sample size. Recently the RMSEA has been recognized as one of the most
informative criteria in covariance structure (Byrne, 2001) with values less than .08
being acceptable. In our model, the value of .045 therefore clearly indicates a superior
model fit and even meets stricter recommendations, whereby the RMSEA should be
<.05 (Hu et al., 1999). Although the goodness-of-fit index (GFI) tends to be strongly
influenced by sample size (Fan et al., 1999), with a value of .934 our model complies
with the required values of > 0.9. Our measurement model showing a CFI value of
.957 well exceeds the lower bound of .90 and therefore can be considered as an
indicator for good model fit. Also the Tucker-Lewis index (TLI), which is less
susceptible to non-normality of data (West et al., 1995) and sample size (Marsh et al.,
1988), yields a corroborating value for good model fit of .943. The adjusted
goodness-of-fit-index is .900. Summarizing, the hypothesized model can be regarded
as to fit the sample data very well.
4.2.2 Reliability and validity
In the next step, reliability and validity of the measures were tested calculating the
composite reliability (CR) of the constructs and the average variance extracted (AVE)
(Fornell et al., 1981), resulting in the following values: Similarity confusion (CR=.64,
AVE=.47), Overload confusion (CR=.80, AVE=.57), Unclarity confusion (CR=.74,
AVE=.49), Seek additional information (CR=.56, AVE=.39), Abandon the purchase
(CR=.54, AVE=.37), Delegate the decision (CR=.65, AVE=.51), Rely on familiar
brands (CR=.54, AVE=.45). Composite reliability, being above or close to the
generally used threshold of .6 (Homburg et al., 1996), is satisfying. The average
variance extracted, however, is in five cases above or close to value of .5, in two cases
it is below .4 suggesting that the reliability of the measures of additional information
seek and abandoning the purchase is not satisfying.
4.2.3 Regression Paths
Figure 2 displays the results of the analysis. Five of the 12 paths show a significant
relationship between the constructs. Similarity confusion, overload confusion and
unclarity confusion explain 37% of the variance of additional information seek, 32%
of abandoning the purchase, 10% of decision delegation, and 5% of the variance of
relying on familiar brands.
Similarity confusion has a negative impact on additional information seek (β= -.84,
p< .01), no significant impact on abandoning the purchase, a negative impact on the
delegation of the decision (β= -.33, p< .05) and no significant impact on relying on
known brands. Overload confusion has no significant impact on the four reduction
strategies in this study. Unclarity confusion has a negative impact additional
information seeking (β= -.18, p< .05), abandoning the purchase (β= -.33, p< .05) but
leads to delegation of the decision decision (β= .40, p< .05).
Figure 1. Consequences of consumer confusion
Similarity Confusion
Overload Confusion
Unclarity Confusion
Seek
additional information
R² = .37
Rely on
familiar brands
R² = .05
Delegate
the decision
R² = .10
Abandon
the purchase
R² = .32
-.84***
-.18n.s.
-.33**
-.23n.s.
-.49n.s.
-.33n.s..
.09n.s.
-.37n.s.
-.18**
-.33**
.40**
.47n.s.
Notes: χ² (d.f.) = 131.459 (90), p = .003, CFI = .957, GFI = .934, AGFI = .900, IFI = .958, TLI= .943,
RMSEA = .045; **** p = .000, *** p = < .01, ** p < .05, * p < .10, n.s. = not significant
4.3 Discussion
The confirmatory factor analysis clearly showed that consumer confusion can also be
found in online hotel booking. Regarding the outcomes of consumer confusion, the
results of this study are highly surprising. In contrast to previous studies, we did not
find a relationship between the three dimensions of customer confusion and relying
on familiar brands. This might be explained by the fact that only a relatively small
number of hotels have strong brands as a member of a hotel chain or that brand
awareness and brand consciousness in the sample were low.
Second similarity confusion and overload confusion are not related significantly to
abandoning the purchase and the relationship between unclarity confusion and
abandoning the purchase is negative. This also contradicts previous findings.
This may have a number of reasons. Firstly, previous studies did not investigate
online purchases. By their nature, online purchase decisions might be different. A
customer who once has decided to book online might not be willing to abandon this
decision. The similarity of offers might be considered as a signal of equal quality of
the products or services and therefore the customer might not be inclined to leave the
decision. Information overload might be considered as a normal situation in online-
shopping and online bookers might be used to it. If consumers perceive ambiguous or
contradicting information they might feel that the decision not to book doesn’t solve
the problem and therefore are willing to book. These interpretations, however, are of a
speculative nature and need to be addressed in future studies. Secondly, moderating
variables in this study have not been controlled. Online bookers might be used to
higher purchase risk due to the nature of online purchase situations, consumer
involvement or situational variables such as time pressure, etc. might also have
influenced this result.
Surprisingly, similarity confusion negatively influences additional information search,
and delegation of decision. Again, similarity confusion could be interpreted by the
consumer as a signal of equal quality and brand similarity. Therefore, the offers could
be perceived as interchangeable. If this is true, customers might perceive a low
purchase risk and additional information seek or a delegation of the decision to others
might not be considered as necessary. In this case, customers could use other
reduction strategies such as to choose the offer with the best price or to decide on a
very limited number of key attributes. However, these are speculations and further
studies are clearely needed.
Overload confusion does not significantly influence additional information seek,
decision delegation or to relying on familiar brands. If customers receive ambiguous
or contradictory information they tend to delegate the decision but do not engage in
additional information seeking or do abandon the purchase.
Overall, this study has yielded some interesting findings. However, further research is
clearly needed. First, a more comprehensive approach is needed that includes
reductions strategies such as choosing the lowest priced offer, postponing the
purchase etc. Second, the effect of moderating variables such as consumers’
involvement, time pressure, purchase risk, etc. should be studied. Third, further
studies should explore theoretically and empirically the differences between online
and offline buying situations. Finally, a convenience sample has been used in this
study and it was not possible to control for any bias. Therefore, the generalizability of
the results might be limited.
Acknowledgements
The authors thank the Airport of Innsbruck for support in data collection, the
Förderverein des Institutes für Wirtschaftswissenschaften of the University of
Klagenfurt and the companies Uniquare Financial Solutions, Kelag, Frantschach Pulp
& Paper for the financial support for this study.
References
Byrne BM. 2001. Structural equation modelling with AMOS. Lawrence Erlbaum Associates:
Mahwah, New Jersey
Cohen M. 1999. Insights into consumer confusion. Consumer Policy Review 9(6): 210-213
Drummond G. 2004. Consumer confusion: Reduction strategies in higher education. The
International Journal of Educational Management 18(5): 317-323
Fan X, Thompson B, Wang W. 1999. Effects of sample size, estimation methods, and model
specification on structural equation modelling. Structural Equation Modelling: A
Multidisciplinary Journal 6(1): 56-83
Fornell C, Larcker DF. 1981. Evaluating structural equation models with unobservable
variables and measurement error. Journal of Marketing Research 18(February): 39-50
Hair JF, Anderson RE, Tatham RL, Black WC. 1998. Multivariate Data Analysis (5 ed.).
Prentice Hall: Upper Saddle River, New Jersey
Homburg C, Gierung A. 1996. Konzeptualisierung und Operationalisierung komplexer
Konstrukte. Ein Leitfaden für die Marketingforschung. Marketing ZFP(1): 5-24
Hu L-T, Bentler PM. 1999. Cutoff criteria for fit indices in covariance structure analysis:
Conventional criteria versus new alternatives. Structural Equation Modelling: A
Multidisciplinary Journal 6(1): 1-55
Kent RJ, Allen CT. 1994. Competitive interference effects in consumer memory for
advertising: The role of brand familiarity. Journal of Marketing 58(July): 97-105
Marcussen CH. 2003. Trends in European Internet Distribution - of Travel and Tourism
Services, Vol. 2003:
Marsh HW, Balla JR. 1988. Goodness-of-fit indexes in confirmatory factor analysis: The effect
of sample size. Psychological Bulletin 103(3): 391-410
Mitchell V-W, Papavassiliou V. 1997. Exploring consumer confusion in the watch market.
Marketing Intelligence and Planning 14(4): 164-172
Mitchell V-W, Papavassiliou V. 1999. Marketing causes and implications of consumer
confusion. Journal of Product and Band Management 8(4): 319-339
Mitchell V-W, Walsh G, Yamin M. 2004. Reviewing and redefining the concept of consumer
confusion. Mansucript Manchester School of Management: Manchester
Nielsen. 2003. Nielsen//NetRatings, Vol. 2003. Nielsen
O'Connor P. 2001. Room rates on the internet - is the web really cheaper? Journal of Service
Research 1(1): 57-72
Poiesz TBC, Verhallen TMM. 1989. Brand confusion in advertising. International Journal of
Advertising 8(3): 231-244
Rayman-Bacchus L, Molina A. 2001. Internet-Based Tourism Services: Business Issues and
Trends. Futures 33(7): 589-605
Rudolph T, Schweizer M. 2003. Kunden wieder zu Käufern machen. Harvard Business
Manager(Februar): 23-33
Sproles GB, Kendall E. 1986. A methodology for profiling consumers' decision-making styles.
Journal of Consumer Affairs 20(2): 267-279
Tabachnik BG, Fidell LS. 1996. Using Multivariate Statistics. Harper Collins: New York
Turnbull PW, Leek S, Ying G. 2000. Customer confusion: The mobile phone market. Journal
of Marketing Management 16: 143-163
Walsh G. 2002. Konsumentenverwirrtheit als Marketingherausforderung. Gabler Verlag:
Wiesbaden
Walsh G, Hennig-Thurau T. 2002. Wenn Konsumenten verwirrt sind. Empirische Analyse der
Wirkungen eines vernachlässigten Konstrukts. Marketing ZFP 24(2): 95-109
West SG, Finch JF, Curran PJ. 1995. Structural equation models with nonnormal variables:
Problems and remedies. In RH Hoyle (Ed.), Structural equation modelling. Concepts,
issues, and applications. Sage: Thousand Oaks, CA
Wiedmann K-P, Walsh G, Klee A. 2001. Konsumentenverwirrtheit: Konstrukt und
marketingpolitische Implikationen. Marketing ZFP(2): 83-99
Yesawich PB. 2000. National Leisure Travel Monitor. Yesawich, Pepperdine & Brown:
Orlando
Appendix: Measurement Scales
Consumer Confusion 5-Point-Rating Scale, (1=strongly agree, 5=strongly disagree),
Similarity confusion a) Information on Hotels in online-booking-systems is so similar
that you cannot do a bargain buy
b) It is difficult to recognized the differences between the hotels in
an online-booking-system
Overload confusion c) There are so many hotels you can book that you are really
confused
d) As there are so many reservation systems and hotel homepages
that it is difficult to decide which one you should chose
e) You get so much information about hotels in the internet that you
lose orientation
Unclarity confusion f) Hotels have so many additional services (e.g. swimming pool,
sauna) and characteristics (view, room size, etc.) that you can not
compare the hotels
g) If you book online you do not have the impression that you have
enough information on the hotel
h) Hotels have so many additional services (e.g. swimming pool,
sauna) and characteristics (view, room size, etc.) that you don’t
know on what you should focus
Reduction strategies 5-Point-Rating Scale (1=strongly agree, 5=strongly disagree)
Seek additional
information
i) If you seek additional information on hotels, it is easier to decide
j) The more information you have on hotels the easier it is to
decide
Abandon purchase k) If I am not sure which hotel to chose I decide not to book online
l) I search information on the internet and then I book with a travel
agency
Delegate decision m) Recommendations and tips of relatives and friends play a role
when I book my hotel online
n) If I am not sure which hotel to chose I leave the decision to
others (partner, family)
Rely on familiar brands o) When I book online I consider only known and strong brands or
hotel chains
p) When I book online I book only expensive brands or hotel chains
(e.g. Hilton, Marriott, etc.)
