iBusiness, 2013, 5, 96-99
http://dx.doi.org/10.4236/ib.2013.53B020 Published Online September 2013 (http://www.scirp.org/journal/ib)
Usability Evaluation of Independent-Sales B2C Fashion
Website Based on Consumer’s Perspective*
Min Li, Yue-ying Ren, Zhu Zhu
Fashion institute, donghua university, shanghai, China.
Email: fidlimin@dhu.edu.cn
Received July, 2013
Based on consumer’s perspective, according to Microsoft Usability Guidelines (MUG), evaluation system of independ-
ent-sales B2C fashion website usability was established. Six independent-sales B2C fashion websites were selected to
be tested. The questionnaire results were analyzed by gray correlation analysis, two-step cluster analysis and hierarchi-
cal cluster analysis, leading to a usability classification and characteristic description, which indicated the quality of
usability of th e selected fashion websites.
Keywords: Consumer’s Perspective; Usability Evaluation; B2C Fashion Website; Microsoft Usability Guid elines; Gray
Correlation Analysis; Cluster Analysis
1. Introduction
The development of China apparel e-commerce has a
history of more than ten years; in particular B2C area of
apparel online shopping grows rapidly. Foreign academ-
ics have conducted a lot of usability researches of
e-commerce website since 1970s; comprehensive evalua-
tion indicators have been developed and used for various
websites studying. Hong-In Cheng studied the usability
of input way, menu and navigation, drew a conclusion
that it is most efficient when there are 50 to 100 menu
options [1]. Ivory, Sinha and Hearst evaluated navigation,
content, visual design, functionality and experience of
the websites from different areas with more than 350
testers [2]. Compared to foreign success, the previous
domestic researches of the independent-sales B2C fash-
ion website have mostly discussed the design and im-
plementation. One of the successfu l studies is that Jinling
Chang and Guoping Xia evaluated the B2C e-commerce
websites of 5 companies based on Microsoft Usability
Guidelines (MUG)[3]. Yan Ge and Ronggang Zhou
studied the color preference of college students[4]. So it
is lack of the usability research from consumer’s per-
spective for independent-sales B2C fashion website.
Therefore, this article established usability evaluation
system of independent-sales B2C fashion website from
consumer’s perspective based on MUG. And then used
the system to classify six independent-sales B2C fashion
websites: VANCL, shishangqiyi, m18, xiu, menglu, togj.
2. Evaluation System
2.1. Microsoft Usability Guidelines (MUG)
Microsoft Usability Guidelines (MUG) is compreh ensive
usability evaluation indexes which proposed by Micro-
soft Corporation. MUG consists of five main indicators:
content, easy of use, promotion, made for the medium,
emoti on [5].
Content is used to assess the capacity of website and
information transfer. It consists of relevance, media use,
depth and breadth, current and timely information.
Easy of use refers to the requirement of ability to use
websites. It is composed of goals, structure and feed-
Promotion refers to publicizing ability on internet or
other media.
Made for the medium refers to the ability to meet the
request of user, which is made up with community, per-
sonalization and refinement.
Emotion could be defined as emotional reaction of
website. It is comprised of challenge, plot, character
strength and pace.
2.2. Evaluation System of Independent-Sales
B2C Fashion Website Usability
Based on the five main indicators of MUG, considering
the consumer culture, consumption habits and feature of
Chinese consumers, evaluation system was established as
*Supported by Innovation Program of Shanghai Municipal Education
Copyright © 2013 SciRes. IB
Usability Evaluation of Independent-Sales B2C Fashion Website Based on Consumer’s Perspective 97
shown in Figure 1.
System has 40 specific indicators which are shown in
Table 1.
3. Empirical Research
In order to achieve the complementary analysis result,
user testing, in-depth in terview and qu estionnaire method
were used during the empirical research together with
evaluation system[6].
3.1. In-depth Interview for Indicator Correction
Typically sample size of in-depth interview is five to ten
people, in addition sophisticated users can find more
problems during th e usability research[7].
For this reason 5 consumers whose “online shopping
age” was higher than two years were chosen. Through
the interview, 40 specific indicator s were cut down to 35
which would be used in Likert scale questionnaire later.
Usability of Independent-Sales
B2C Fashion Website
Easy of Use
Made for
the Medium
Figure 1. Evaluation system of independent-sales B2C
fashion website usability.
Table 1. Specific indicators of evaluation system.
Indicators Specific Indicators
commodity classification, website updates s pee d,
accurate commodity information, media information
expression, enrichment of com modity, image
orientation, detail shown from pictures, text credi
user evaluation, correct statistical and accounting,
fashion trend, commodity recommendation
Easy of Use
site layout ,interface design, website information
organization, rapidly commodity query, accurate links,
informa tion speed, navigation system, convenient
return, convenient payment and evaluation, feedback
timely, obvious exit path, succinct process
Promotion website promotion, sales promotion, membership
Made of the
instant communication tool, message system, BBS,
personal space, privacy protection, personalized
service, payment platform, customer self-management,
after-sale service
Emotion control of the amount of information, cost
performance, website attraction, website credibility
3.2. Selected Websites for Research
2009-2010 China Apparel B2C Online Shopping Re-
search Report indicated that the independent-sales B2C
fashion websites which ranked high of market share are
shown in Figure 2: VANCL, m18, menglu, Maso Maso,
shishangqiyi, togj, Xiu, HANY[8]. But Maso Maso and
HANY only sale men’s clothing, they are not suitable for
female consumer to do the test. Hence other six websites
were identified as test objects except these two.
r1: VANCL-- www.vancl.com
r2: m18-- www.m18.com
r3: menglu-- www.menglu.com
r4: shishangqiyi-- www.shishangqiyi.com
r5: togj-- www.togj.com
r6: xiu-- www.xiu.com
3.3. User Testing
Nilsen and Landauer had pointed out that 85% of the
problems fro most of the usability tests could be found
by 5 users[9].
To do usability testing, 30 consumers who were vet-
erans of B2C fashion website were chosen, and all 30
samples were proved to be effective. Among them stu-
dent/office worker ratio is 2:1, male/female ratio is 2:3,
19-24/25-30 years old ratio is 3:4, basically conform with
the distribution of Chinese online shopper in 2009 Chi-
nese Clothing Brand and Apparel Online Shopping Re-
search Report reported by China Intelli Consulting
Test time and place are up to testers in order to avoid
the influence from surrounding. Test is carried out as
following steps.
1) Step 1: Finish 4 tasks as follows during the test.
a) Task 1: Knowing the ranking of the independ-
ent-sales B2C fashion websites for reference.
b) Task 2: Female tester need to purchase a suitable
summer chiffon dress. Male tester need to purchase a
cotton business shirt.
Masa Masa
ot hers
Figure 2. Market share of independent-sales B2C fashion
Copyright © 2013 SciRes. IB
Usability Evaluation of Independent-Sales B2C Fashion Website Based on Consumer’s Perspective
c) Task 3: Female tester need to purchase a striped
sweater which is on promotion. Male tester need to pur-
chase a pair of jeans which is on promotion.
d) Task 4: Consult with customer service about size.
2) Step 2: Fill the form of indicator weight.
3) Step 3: Fill in the score of each indicator for each
4. Data Analysis
4.1. Gray Correlation Analysis
Subjective assessment for indicators of website are in-
fluenced by knowledge, experience, culture and many
other known or unknown factors, as a result, grey corre-
lation analysis can be used for evaluation[11].
Based on grey correlation analysis, the higher the cor-
relation coefficient is, the better the usability will be.
After analyzing scores given by testers with SPSS17.0,
here are correlation coefficients of websites: r1=0.902,
r2=0.821, r3=0.684, r4=0.827, r5=0.670, r6=0.72. It is
leading to the classification of usability: VANCL, shis-
hangqiyi, m18, xiu, menglu, togj.
Table 2 is the correlation coefficient of main indica-
tors, it is shown that shishangqiyi, m18 an d VANCL rank
high of content, which accords with their feature of fast
VANCL performs well in easy of use, and testers also
have pointed out the download time is very important
during the shopping.
It seems to make sense that the ranking of promotion
matches that of market share in the main.
VANLE get high praise of made of medium, from lo-
gistics to packing it does very well. On the contrary, af-
ter-sale service of menglu has been co mplained a lot.
At last it is not surprising that VANCL ranks high of
emotion, owe to th e good reputation.
4.2. Cluster Analysis
Cluster analysis of SPSS is composed by K-means clus-
ter, hierarchical cluster and two-step cluster, the latter
two have been used in this article.
Table 2. Correlation coefficient of main indicators.
coefficient Content Easy of
Use Promotion Made of the
Medium Emotion
r1 0.299 0.218 0.118 0.137 0.129
r2 0.295 0.193 0.098 0.123 0.111
r3 0.246 0.163 0.072 0.100 0.103
r4 0.325 0.205 0.073 0.106 0.118
r5 0.235 0.160 0.066 0.112 0.096
r6 0.239 0.187 0.064 0.132 0.102
Log-likelihood and Bayesian Information Criterion
(BIC) have been selected to do two-step cluster analysis,
as shown in Tab le 3, it comes to higher distance ratio in
step 2 and step 4.
Besides, each cluster should extremely represent the
characteristics of the website, consequently hierarchical
cluster analysis has been down under the premise of 4
clusters based on Between-groups Linkage and Squared
Euclidean. Figure 3 Dendrogram shows that 4 clusters
are: VANCL and m18, menglu and togj, shishangqiyi,
According to Table 4, Cluster 1 performs the best for
promotion, Cluster 3 is the best of content, and Cluster 4
does better than others in made of the medium.
As a result of the Cluster Analysis, the characteristics
of each cluster could be summarized and described as
blow in Table 5. It is observed that the clustering result
is basically conducted in accordance with the sorting of
the usability: r1>r4>r2>r6>r 3>r5.
5. Conclusions
1) The evaluation system of this article is based on
consumer’s perspective, and Microsoft Usability Guide-
lines (MUG), combined with the uniqueness of inde-
pendent-sales B2C fashion website.
2) On the basis of evaluation system, the usability
classification of selected websites is given by gray corre-
lation analysis: VANCL, shishangqiyi, m18, xiu, menglu,
Table 3. SPSS auto-clustering.
Clusters BIC BIC
Variance Variance
Ratio Distance
1 36.102
2 43.602 7.501 1.000 3.832
3 58.802 15.199 2.026 1.093
4 74.231 15.429 2.057 1.821
5 90.783 16.552 2.207 1.144
6 107.50616.723 2.230 0.000
CASE 0 5 10 15 20 25
Label Num+-------+-------+------+-------+-------+
menglu r3 -+--------------+
togj r5 -+ +------------------------------+
xiu r6 -----------------+ |
VANCL r1 -+------------+ |
m18 r2 -+ +--------------------------------+
hangqiyi r4 ---------------+
Figure 3. Dendrogram.
Copyright © 2013 SciRes. IB
Usability Evaluation of Independent-Sales B2C Fashion Website Based on Consumer’s Perspective
Copyright © 2013 SciRes. IB
Table 4. OLAP cubes for Hierarchical Cluster Analysis.
Clusters Content Easy of
Use Promotion Made of the
Medium Emotion
SUM 7.670 8.160 7.980 7.760 7.690
N 2 2 2 2
Mean 3.837 4.080 3.992 3.882 3.844
Std 0.051 0.161 0.294 0.069 0.191
SUM 6.710 7.210 5.620 6.880 6.940
N 2 2 2 2 2
Mean 3.355 3.604 2.810 3.440 3.471
Std 0.092 0.007 0.127 0.311 0.086
SUM 4.110 4.030 3.020 3.490 3.830
N 1 1 1 1 1
Mean 4.110 4.030 3.020 3.490 3.830
Std 0.000 0.000 0.000 0.000 0.000
SUM 3.390 3.930 2.630 3.960 3.530
N 1 1 1 1 1
Mean 3.390 3.930 2.630 3.960 3.530
Std 0.000 0.000 0.000 0.000 0.000
SUM 21.880 23.330 19.250 22.090 21.990
N 6 6 6 6 6
Mean 3.647 3.888 3.209 3.682 3.665
Std 0.327 0.238 0.635 0.287 0.214
Table 5. Character description of independent-sales B2C
fashion website.
Clusters Websites Characteristics
Attach great importance to both
commodity-promotion and
self-promotion, create good
impression and usability.
Homogenized websites, lack of own
characteristic, usability has to be
type shishang-qiyi
Gorgeous interface, good at capture
the fashion trend, attract the
consumer with précised fashion
nous and feat ured products.
friendly type xiu Provide convenient communication
platform with highly regarded
human-computer interaction
3) Then the six websites are clustered to 4 categories
by two-step cluster analysis. Through hierarchical cluster
analysis, the usability characteristic of each cluster is
respectively defined as sales-promotion type, bal-
anced-development type fashion-taste type, and termi-
nal-friendly type. As basis the development of the web-
sites can be professionally bring into force.
[1] H. I. Cheng, “Human-Computer Interaction in
E-business,” 2002, UMI: 3073438.
[2] M. I. Ivory, R. R. Sinha and M. A. Hearst, “Empirically
Validated Web Page Design Metrics,” CHI, 2001, pp.
[3] J. L. Chang and G. P. Xia, “Usability Evaluation of B2C
E-Commerce Website, ” China Academic Journal, Vol. 24,
No. 2, 2005, pp. 237-242.
[4] Y. Ge and R. G. Zhou, “Research on Color Preference of
Computer Interface of College Students,” Chinese Jour-
nal of Ergonomics, Vol. 10, No. 3, 2004, pp. 23-26.
[5] K. Keeker, “Improving Web 2 Site Usability and Appeal:
Guidelines Compiled by MSN Usability Research,” 2008.
http://msdn.microsoft.com/library/default.asp? url=/ li-
[6] W. Song, “The Research on the Usability Evaluation
System of E-mail Website in China-Take the Electronic
Website for Example,” Master Thesis, Shangdong Uni-
versity, Jinan, 2007.
[7] X. M. Chen, “Qualitative Research in Social Sciences,”
Educational Science Publishing House, Beijing, 2000.
[8] “2009-2010 China Apparel B2C Online Shopping Re-
search Report,” Transaction of IRESEARCH, 3, 2010.
[9] J. Nilsen, “Why You Only Need to Test with 5 Users,”
[10] “2009 Chinese Clothing Brand and Apparel Online
Shopping Research Report,” Transaction of China Intel-
liConsulting Corp. 3, 2010.
[11] J. L. Deng, “The Primary Methods of Grey System The-
ory,” Huazhong University of Science and Technology
Press, Wuhan, 1987.