See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/334734832

A System for E-Commerce Website Evaluation

Conference Paper · June 2019

DOI: 10.5593/sgem2019/2.1/S07.004

CITATIONS READS

8 7,890

1 author:

Snezhana Sulova
University of Economics - Varna
54 PUBLICATIONS 101 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Уеб базиран регистър "Научна дейност" View project

Digitization of the Economy in an Environment of Big Data (DEBD) View project

All content following this page was uploaded by Snezhana Sulova on 03 August 2019.

The user has requested enhancement of the downloaded file.

 Informatics

A SYSTEM FOR E-COMMERCE WEBSITE EVALUATION

Assoc. Prof. Ph.D. Snezhana Sulova

University of Economics – Varna, Bulgaria

ABSTRACT
E-commerce involves the processes of selling and buying goods and services through the
use of modern communication technologies and the Internet. The success of this form of
commerce largely depends on the website through which the sales are carried out. Modern
online e-commerce platforms are sophisticated applications that perform multiple
functions. They are both a marketing tool that attracts customers, dynamic systems that
allow interaction with the users, and the realization of transactions and a portal with useful
information about the sold goods and services. For online retailers as well as for software
developers, it's important to understand how effective the website, through which the
online sales are done, is. This paper proposes a methodology for evaluating e-commerce
websites. In order to understand the strengths and weaknesses of an e-shop, as a basic
tool for doing business, we believe it is good to make a comprehensive assessment by
means of a system of indicators grouped in the following sets: evaluation of the website’s
visitability; evaluation by specific e-commerce indicators; evaluation of the e-commerce
website’s functionality; evaluation of the e-commerce website as a marketing tool. Some
of the metrics are obtained as a result of expert judgment, while others are obtained from
the data of the e-shop’s database, data from analytical systems and data extracted from
Internet resources. The proposed metrics system for evaluation e-shops is of importance
to improving business processes, customer relationships, marketing, and management.
Keywords: e-commerce website, key performance indicators, evaluation, expert
judgment

INTRODUCTION
Nowadays, e-commerce is becoming a preferred form of making transactions.
Worldwide, incomes from this form of trade are rising steadily, for the period 2014-2017,
they have grown nearly 2 times and have reached 2.3 trillion US dollars, with forecasts
saying that in 2021 they will reach up to 4.88 trillion US dollars [1]. The success of e-
sales depends on many factors, but Web-based platforms are crucial for their
implementation. Modern online marketing applications are constantly being refined, they
are not only a tool through which customers make purchases but are also complete
systems that help and guide them during the time of the purchase. In order to be
competitive on the market and to maximize their online profits, retailers are constantly
looking for new ways to evaluate how effective their e-commerce websites are. It is
known that an activity is effective when it maximizes the results of the action [2].
This report provides a methodology for evaluating e-commerce websites based on a
system of indicators which helps to motor the activity and identify the strengths and
weaknesses of e-shops.

25
19th International Multidisciplinary Scientific GeoConference SGEM 2019

E-COMMERCE WEBSITE EVALUATION

Evaluation of e-commerce websites has a major impact on sales growth and the
generation of incomes from online marketing. It is part of the overall E-commerce
analytics process, which includes all the activities for analysis of systematic data that are
performed to improve the business results of companies that sell online [3]. Web site
evaluation should be planned in detail and then used as fully as possible. Тhe following
main phases of the e-commerce web site analysis process can be listed:
• Identification of the business objectives of the analysis;
• Selection of the criteria for the analysis;
• Collecting and integrating data;
• Evaluation according to the selected criteria;
• Forming reports and analyses of the obtained results;
• Development of a plan for making improvements.
The phase of choosing the right criteria for analysis has the greatest importance for the
overall process of analyzing the e-commerce website. Various studies use a wide range
of metrics to evaluate online stores. In his survey, Mark Hayes, presents 67 indicators to
evaluate eCommerce, grouping them into the following categories: Sales, Marketing,
Customer Service, Manufacturing and Project Management [4]. Oracle suggests
examining the e-commerce platforms by the following areas: Scalability, Product
Catalog, Business User Control, Search, Agility, Reporting and Analytics, Standards,
Integration, Interoperability, Synergy [5]. Other researchers use the OSSpal
methodology, which combines quantitative and qualitative software assessment measures
with seven different categories: Functionality, Operational Software Characteristics,
Support and Service, Documentation, Community and Adoption, Development Process
[6]. Bezes defines three main approaches to evaluating websites, analyzing them as:
information systems, communication channels and retailing channels [7]. According to
Davidaviciene, the five most important criteria are: easy to use, navigation, security
assurance, real time help, and content [8].
All these studies have many common aspects regarding e-shop analyses. Each of the
mentioned researchers prioritizes different criteria because they have set a specific task
for themselves, such as evaluating e-shops in a particular area, researching an activity-
related aspect, such as consumer satisfaction, use and adaptation of an existing evaluation
methodology. We believe that in order to understand the strengths and weaknesses of an
e-commerce web site, it is essential to analyze key indicators, such as those that result
from quantitative and qualitative assessment, and those that result from expert judgment.
This will allow a comprehensive assessment to be carried out in several areas such as:
evaluation of the website's visitability; evaluation by specific e-commerce indicators;
evaluation of e-commerce website's functionality; evaluation of the e-commerce website
as a marketing tool. (Figure 1)

26
 Informatics

Fig. 1. E-commerce website evaluation

The proposed system for evaluating e-commerce web sites is an analytical framework of
analysis indicators, with many characteristics included different areas of the analyses.
Key performance indicators can be calculated from stored data and used for the areas of
evaluation of the website visibility and evaluation by specific e-commerce indicators and
the analysis in the other two areas is based mainly on expert judgment.

KEY PERFORMANCE INDICATORS FOR E-COMMERCE WEBSITES

We have divided the key indicators for e-commerce website evaluation into two main
groups: common metrics that are valid for all websites and indicators specific for some
e-commerce sites.
The indicators for e-commerce website visits can be called common as they are essential
for e-commerce but are also valid for all other types of websites. These indicators are
calculated using existing web analytics tools, such as Google Analitycs. Out of them, the
following have a key meaning in the evaluation of online sales websites:
• Visits – the activation of the e-shop in a browser or through a mobile application
is considered the start of the visit, and exiting the site or interrupting the session
is considered to be the end. If the e-shop is not closed after a long period of time,
the user session is usually interrupted. After a 30-minute interval a new session
begins. The traffic to an e-commerce site is essential because the more visitors
there are, the more likely they are to become buyers. When traffic is targeted, it
brings more profits.
• Unique visitors – the calculation is required because a visitor can generate
multiple visits that are separated over time. To track this metric, users' IPs are
tracked, and cookie technology is used. It is worth mentioning that the number of
unique visitors cannot be calculated with absolute accuracy because the user can
use a web browser with a setting not to store the cookie information, yet this
metric is a good enough indicator of how many of all the visitors are different.

27
19th International Multidisciplinary Scientific GeoConference SGEM 2019

• Returns – shows the percentage of visitors who have already viewed the e-shop.
Calculation also uses cookie technology. This indicator shows that the e-shop is
liked, interesting, and useful to customers.
• Time on site – shows the approximate time a visitor stays on the website in
minutes. It is calculated from the time of accessing the page until the time of
opening the next page. The stay on the site is the sum of the time it takes to visit
each page. The length of staying on the last visited page is considered to be 0
minutes. This indicator also takes into account the extent to which the site
manages to attract the attention of visitors. However, if the length of the stay is
too high, it should also be considered whether this is not due to poor site
organization, poor navigation, difficult-to-understand content, or a complex
ordering system.
• Bounce rate – shows the percentage of visitors who view only one page and then
leave the website. Reporting the rate of the site's exit is important because an e-
shop can have many visitors and still a large part of them might not have become
its customers and may have immediately left it. An e-commerce site is successful
if the bounce rate is as low as possible.
• Top exit pages – shows from which pages the e-shop visitors most often leave.
An analysis of those pages can show the reasons why people left the website and
why the content of these pages is not appealing to users.
• Pages viewed per visitors – is calculated as the ratio between the total number of
pages viewed and the number of visitors to the website. In general, it is good to
have a large number of viewed pages, but in some cases, this may indicate that
the site needs to be reprocessed because in order to find the information you need
it is necessary to browse through many pages.
• Top visited pages – the most frequently visited pages of the e-commerce website.
This is an indicator that can draw insights into the interests of visitors, optimizing
the supply and make offers that best match their expectations and preferences.
• Traffic source – shows where visitors come from and how they discover the site,
whether through organic search, paid ads or social media.
• Top visitors per country – gives information about which geographical
locations the visits to the e-shop are from and this can serve to offer goods tailored
to the national characteristics, traditions, culture of potential buyers.
Measuring the proposed common e-commerce website evaluation metrics helps to
identify the strengths and weaknesses of the e-commerce web system. Based on the
common indicators listed, conclusions can be drawn about the dynamics of e-shop
visitability, about how the web site is interesting, the reasons for loss of interest in the
web site, errors
The specific e-commerce key performance indicators are such that e-commerce itself
requires. They can be an indicator not only of how successful the web site is for online
sales but also of the success of online business as a whole. The main ones are:
• Revenue per visitor – it is calculated by dividing the total revenue by the
total number of visitors to the site, the value of each additional e-shop

28
 Informatics

visitor is estimated. It can also calculate a visit’s value by the type of

customer and by the country.
• Conversion rate – it is calculated by dividing the total number of visitors by the
total number of conversions. It shows the value of a visit and is an important
indicator when determining the quality of traffic because an e-shop can have many
visitors but only a few of them could generate its income. The performance of ad
campaigns is most often measured via the value of a visit.
• Shopping cart abandonment rate – how many users add products to their
shopping cart but do not make a purchase. It is better for this criterion to have as
small a value as possible.
• Step completion rate – shows how well users use the site, successfully discover
a product, and make a purchase through the site.
• New customer orders – shows how many purchases have been made by new
customers and measures customer growth rate.
• Returning customer orders – shows how many of the purchases have been made
by previous customers, which shows their loyalty
• Customer loyalty – it is determined by monitoring the ratio of new and existing
customers. An increase in the number of loyal customers leads to a direct increase
in the income from commercial activity.
• Product relationship – identifying products that are viewed consistently, which
can help effective cross-selling tactics.
Specific metrics help managers make adequate business decisions. To them can be added
indicators that are used to measure the success of ad campaigns and promotional offers.
Based on these indicators, it is possible to draw conclusions about the main sources of
income for the e-shop, the satisfaction of customers with the goods and services offered
and the e-commerce system as a whole, the presence of related goods, the success of the
advertising campaigns, etc.

EXPERT EVALUATION OF E-COMMERCE WEBSITES

For a more complete and trustworthy evaluation of online sales websites, we think it is
expedient for them to also be evaluated by e-commerce experts. There are important
features of e-commerce websites that cannot be measured quantitatively. In our opinion,
it is best to do the expert evaluation of e-commerce websites in the following main areas:
• Evaluation of the functionality of the e-commerce web site, which includes an
overview of the presented catalog of goods and services, search capabilities,
comparison, filtering, functionality of the ordering system and tracking the order
status, integration of the web site with other systems.
• evaluation of the e-commerce website as a marketing tool, includes an opinion
about the content, design, structure and navigation of the website; the applied on-
page search engine optimization (SEO) techniques such as meta data analysis,
inbound links, the code, content, speed, domain, hosting, and the integrated off-
page SEO – the means to promote the website to external sources.

29
19th International Multidisciplinary Scientific GeoConference SGEM 2019

In table. 1 a sample survey questionnaire for expert evaluation of e-commerce websites

is presented.
Table 1.
A sample survey questionnaire for expert evaluation of e-commerce websites

Evaluation criteria Points

1. The functionality of the e-commerce web site: 50

1.1. Structure and design of the catalog of products or services 5
1.2. Possibilities for sorting and filtering goods and services 5
1.3. Opportunities to search and compare goods and services 5
1.4. Registration system 5
1.5. Shopping cart and order process 5
1.6. Integration with payment systems and bank credit cards 5
1.7. Integration with delivery systems 5
1.8. Means of personalization 5
1.9. Data transfer security 5
1.10. Multilingualism of the website and work with different currencies 5
2. The e-commerce website as a marketing tool 50
2.1. Graphic design of the website 5
2.2. Site content 5
2.3. Organization and navigation of the site 5
2.4. Responsive design 5
2.5. Clients’ instruments – chat systems, comments, etc. 5
2.6. Program code and meta data 5
2.7. Inbound and outbound connections 5
2.8. Site’s loading speed 5
2.9. Domain and hosting 5
2.10. Sharing in social media and on other web sites 5
In total: 100

After an evaluation of the selected criteria, a 6-step evaluation scale can be used,
according to which if the website has gathered from 0 to 49 points it is evaluated with a
Fail score; from 50 to 69 – Satisfactory; from 70 to 79 – Good; from 80 to 89 – Very good
and from 90 to 100 – Excellent.

CONCLUSION
In conclusion, it should be noted that the evaluation system allows for a comprehensive
analysis of the usability, content and functionality of online stores, not only take a look
at simple statistics on their attendance. It is not universal in character, but rather provides
a framework for evaluation, and if necessary, in the analysis process, a metric can be
excluded, or the system can be supplemented with other metrics. Web analysts are the
ones who define the main parameters for evaluation that best show the strengths and

30
 Informatics

weaknesses of the web site as a major tool for doing business. It is not by chance that the
well-known analyst Avinash Kaushik determines that, when analyzing web resources,
only 10% of the budget is spent on analytical tools and the remaining 90% on human
resources to carry out the process [9, p.81].
Future work on website evaluation may be focused on the use of artificial neural networks
(ANN). Known applications of ANN [10, 11] may be adapted to the topic of this paper.
A significant advantage in the assessment would be to include indicators based on data
mining technologies to process data of the usability of websites as they allow for deriving
extra valuable and useful information about consumer visits, buyers' interests, their
behavior and the operation of the e-commerce system. For example, clustering can
identify groups of clients and then examine the value of a site’s visit by the identified
groups. In addition to the proposed system and indicators that are calculated mostly by
statistical methods, data and web mining technologies can also be used to detect
additional hidden dependencies in the data. This would be a valuable competitive
advantage for the companies that use them.
Evaluation of e-commerce websites as part of the overall process of e-commerce analytics
creates business value and drives business growth.

REFERENCES
[1] Retail e-commerce sales worldwide from 2014 to 2021 (in billion U.S. dollars),
https://www.statista.com/statistics/379046/worldwide-retail-e-commerce-sales/,
[Accessed: 21.03.2019].
[2] Mihaiu, D.M., Opreana, A., Cristescu, M.P. Efficiency, effectiveness and
performance of the public sector, Romanian Journal of Economic Forecasting, 13(4), pp.
132-147.
[3] Philips, J. Ecommerce Analytics. Analyze and Improve the Impact of Your Digital
Strategy, Pearson Education, New Jersey, 2016.
[4] Hayes, М. 76 Key Performance Indicators (KPIs) for Ecommerce,
http://www.shopify.com/blog/7365564-32-key-performance-indicators-kpis-for-
ecommerce, [Accessed: 21.03.2019].
[5] Oracle, The Top 10 Technical Considerations for Evaluating E-Commerce Platforms.
Oracle White Paper. March 2011.
http://www.oracle.com/us/products/applications/atg/top-10-considerations-ecommerce-
333324.pdf, [Accessed: 21.03.2019].
[6] Ferreira1, T., Pedrosa, I and Bernardino, J., Evaluating Open Source E-commerce
Tools using OSSpal Methodology, In Proceedings of the 20th International Conference
on Enterprise Information Systems (ICEIS 2018) - Volume 1, pp. 213-220.
[7] Bezes, C., E-commerce Website Evaluation: A Critical Review, Journal of Electronic
Commerce Research, 2009.
[8] Vida, D., Jonas, T., Measuring quality of e-commerce web sites: Case of Lithuania,
Economics & Management, vol.16, pp. 723-729, 2011.
[9] Kaushik, A. Web Analytics. An Hour a Day, Wiley Publishing, 2007.

31
 19th International Multidisciplinary Scientific GeoConference SGEM 2019

[10] Shichkin, A., Buevich, A., Sergeev, A., Baglaeva, E., Subbotina I., Vasilev J.,
Kehayova-Stoycheva, M. Training algorithms for artificial neural network in predicting
of the content of chemical elements in the upper soil layer, AIP Conference Proceedings
pp. 060004-1 - 060004-5, doi: 10.1063/1.5082119, 2018.
[11] Tarasov, D., Vasilev, J., Sergeev, A., Mokrushin, A. Artificial neural networks
selection for soil chemical elements distribution prediction. AIP Conference Proceedings,
1978,440025, 2018.

32

View publication stats