Analysis and web structures optimization

(1)

Keywords—Analysis, optimization, search engines, SEO, websites.

Abstract—The aim of the presented paper is to familiarize the reader with methods and techniques of web structures optimization. The main benefit of the work is description of analysis of existing non-optimized website, demonstration of the practical website optimization and testing and evaluation of optimized web structures. The work is focused on SEO search engine optimization. It is a method that helps online presentations and websites succeed in today's competitive world.

I. INTRODUCTION

OWADAYS Nowadays it is very popular that every company or organization has its own websites. This trend has been developing for several years and therefore it is important to pay attention to it. With a number of websites increased a competition and a battle for first rungs in search results. It was necessary to begin to focus on the theme of how to optimize web structures. The main goal is to affect the placement of websites in search engines results. [1]

Generally, most visitors come to the websites just over search engines. For the rea-son that the search engines find our site, we have to inform them. One of relatively easy and fast way is to register our sites to the catalog link, which operates the most inland portals. Why register websites in some catalogs when they are separated from the search engines? A number of search robots as the first in its links hunt on the internet attending just these catalogs. The fact that most people coming to sites from search engines, we cannot generalize. More complex phrases are, however, less popular. Users prefer to search pages by scrolling through categories in the catalog. [2]

In contrast with full-text search engines in most catalogs we can buy display position. Search engines also offering the opportunity to visibility and display a link to your site, but for now there is a generally accepted custom that such paid sites are visually distinguished from the standard search results. Search engines are trying to present to its visitors especially Z. Prokopova is with the Department of Computer and Communication Systems, Faculty of Applied Informatics, Tomas Bata University in Zlin, nam. T. G. Masaryka 5555, 760 01 Zlin, Czech Republic (corresponding author to provide phone: +420-57-6035011; e-mail: prokopova@ fai.utb.cz).

R. Silhavy is with Department of Computer and Communication Systems, Faculty of Applied Informatics, Tomas Bata University in Zlin, nam. T. G. Masaryka 5555, 760 01 Zlin, Czech Republic (e-mail: [email protected]).

P. Silhavy is with Department of Computer and Communication Systems, Faculty of Applied Informatics, Tomas Bata University in Zlin, nam. T. G.

relevant results to their queries. On the contrary, catalog does not care about content of the site and de facto often does not know it. Catalog editors may sometimes go through the individual registrations, but often include references to its database automatically, without human control.

The goal of each website operator is to bring his own site to as many people as possible. Each visitor generates income. It is, for example, income from advertising or from sales of other services. Google (or Yahoo!) is trying to bring daily on their web-sites as many users as it is possible. And how this is accomplished? Just give users what they want - relevant search results. It is clear that on the problem of giving the most relevant answers, they are working “day and night”. [3]

II. HOW SEARCH ENGINE WORKS

Let us focus, for example, on the very popular Google search engine. In the following overview you can find some of the many reasons, why Google is so popular. In addition, general facts about what Google offers with a brief list of the best known and most widely used services:

• Very fast and comprehensive full-text search engine – primarily are sites, that mediate responses to what we are looking for on the internet, divided into catalogs and search engines. There are two basic categories of search engines. Search engines from the first category are searching titles and descriptions of the sites and the content of the sites is not taken into account. The second category represents full-text search engines. They are searching the contents of websites and one of them is Google.

• Accuracy of results – despite the huge number of scanned pages Google will al-ways return accurate results. There are results which contain the searched information. Google can be considered as one of the most successful search engine in terms of quality of results.

• No “annoying insect” (advertising) – in contrast with some of the other search engines Google does not offer paid links as a search results.

Google background consists of a server containing a list of URLs. Downloaded pages are sent to the storage server (warehouse). In the warehouse sites are compressed and stored in the depository. Each site receives a unique identification number, which is known as docId. The indexer and sorter take care about the inclusion in to the register (indexing). We could say that the indexer is the heart of search engines. Detailed description of its work can be found e.g. in [4].

Analysis and web structures optimization

Zdenka Prokopova, Radek Silhavy, and Petr Silhavy

(2)

A. Aspects affecting the location of website in search results

Search engines determine the placement of our website on the SERP (search engine result page) according to several criteria:

• The mutual positions of found words (searched for multiword phrases);

• Location of found words (position in the document); • Location of a word or phrase in the page title, meta

descriptions and titles;

• The page weight in the eyes of search engines (PageRank); • The number of links on the sites, and the quality of those

links.

1) Evaluation – PageRank (PR)

PageRank is an algorithm that shows authenticity of websites on a scale from 0-10. Display position in search results depends on the amount of PR. The authors of the original algorithm are former students at Stanford University, Lawrence Page and Sergey Brin the founders of Google. The algorithm is based on the Kandall-Wei evaluation theory from the fifties of the last century. The main idea of this theory is based on comparison of the importance of people and things based on their mutual influence. PageRank is the value of the credibility, i.e. how many pages simultaneously evaluated (using the same formula) refer to that page. PageRank is an important factor in determining the location of the site on the search engine result page (SERP). [1], [4]

2) SERP – search engine result page

SERP is an acronym, which we can sometimes meet on the pages dedicated to searching and search engine optimization. The main task of SEO is to get the best position in the SERP. The aim is to get the maximum number of relevant visitors - visitors who are interested in the information, products or services offered on our websites. [5]

III. SEO METHODS

SEO methods can be divided into two basic groups according ways of doing SEO. On-page SEO includes providing good content, good keyword selection etc. (on-page factors). Off-page SEO includes link building, link exchange etc. (off-page factors).

A. On-page SEO

On-page factors are factors that occur on one unique site. This means titles, headers, keywords, pure texts etc. The general rule in optimization is that every page on the optimized site must be unique. It is always important to keep in mind that search engines evaluate each page separately. So it is necessary to focus not only on the home page, it would be almost useless, but on all pages of the sites. [6]

1) Keywords

Keywords selection is the most important part of on-page SEO. If the site does not contain specific keyword, search engines cannot find it (except anchor text). If you are creating site from the beginning, just select the appropriate keywords

and distribute them well all over the site. Each page can be optimized for about 5-6 keywords. In the case of more competitive area is the number of keywords usually smaller (1-2). More detailed information we can find in [7].

2) Simplification of site navigation

Creation of file sitemap.xml is one of the way how to improve the position of site in the eyes of search engines. Using the sitemap we tell to the browser robot, which pages are on our site, and under what address there are.

The problem for a number of the authors is how to create a site map. Sitemap contains a list of pages in the form of XML structure. In addition, sitemaps contain their weight, date of last change and frequency of the page changes. Manual creation of such a list could be very frustrating or even impossible for very large sites. Curiously enough, a number of large sites do not have a sitemap for search engines. [8]

B. Off-page SEO

Off-page factors are connected with link building but not only that. Off-page SEO refers to activities outside the boundaries of the page e.g. social media (Facebook, twitter…), social bookmarking etc. It gives us a good indication on how other sites and users perceive our site.

Most popular off-page SEO method is link building. Links are like the streets be-tween pages. Link building it’s not only a matter of how many links are pointing to your site but it is more important from where these links are coming. [7]

IV. ANALYSIS OF THE EXISTING WEBSITE

The first step to good search engine optimization is the initial website analysis. It may be a current web presentation, which the company already has, but is not satisfied with it. We were asked for initial analysis of existing websites and their subsequent optimization by unnamed small manufacturing company. Initial analysis must be taken very seriously i.e. we cannot forget any important thing. The following sections briefly describe the shortcomings of the existing web server pages from the perspective of SEO optimization.

A. Analysis of index.html

When we look at the home page from the viewpoint of SEO optimization, we can find serious weaknesses on the very first and most important step. It can be found from the source code listing that missing basic meta tags as "descriptions" or “keywords”. When we go on, we can find the most important tag <title>. The title tag in original source code of the index page is completely wrong. The main problem is the vague description which will not say the necessary information about the content of the website to the search robot.

B. Analysis of on-page factors

The next step will focus on analysis of the several on-page factors. For help we took application Seo Servis [9] that offers a very useful SEO tools such as source code analysis, keywords analysis, etc. The steps are divided into some subcategories.

(3)

1) Descriptive information and document header

The first point of analysis is descriptive information. This function lists all the important information such as title, description, keywords, robots info etc. We found that some items of the document header are filled, but the label of page (description) is not filled. As is written above, some search engines use the mentioned label in the search results, but it is not in all.

2) Source code

Clear and high-quality source code is also very important for good SEO optimization. Any text that appears on the page should be in paragraphs, headings and subheadings marked with the appropriate style. Images that are used must include an alternative description called "alt". If we write source code, we should avoid mixing semantic highlighting with physical formatting. All should be defined using CSS styles (avoid using labels <font>). Very important is also to debug source code, to be free of errors and valid.

3) Content part

The last point in the analysis of on-page factors is part of the content. It is necessary to use keywords in headings and paragraphs of the text. Some search engines use tag <strong>, which is also good to use for each keyword.

C. Analysis using Google Analytics

Another very important step towards a good optimization is the analysis of the existing visitors of websites. Very powerful and useful tool is Google Analytics [10], which helped us determine e.g. how many visitors come to sites. After registration Google Analytics generates source code that must be inserted into each page that you want to monitor. In our case it was generated the following source code:

pageTracker._trackPageview();} catch(err) {} </script>

Analysis by using Google Analytics started on 8th September 2013 and ended on 8th October 2013. During the analysis we found very interesting information about the existing websites which we have used in the preparation of new sites.

1) Analysis of access to web site

Analytics offers plenty of options to monitor a web-site. The most important factors are the total number of visitors and traffic sources overview. In the Table 1 (column “Before”) you can find information received from the web server during the

measured time interval (30 days). The total number of web-sites visits for the test sample is 621 visitors. On average, there is about 15 visitors per day what is not very promising number. The bounce rate is about 57% what is on the other hand side a very high value. This number indicates that the visitor leaves the site from the main page. Average time on site (1:15 min) is sufficient, but the value should be larger. According to these indicators we can determine the poor quality of the website.

V. PRACTICAL WEBSITE OPTIMIZATION

Please note that the references at the end of this document are in the preferred referencing style. Give all authors’ names; do not use “et al.” unless there are six authors or more. Use a space after authors' initials. Papers that have not been published should be cited as “unpublished” [4]. Papers that have been submitted for publication should be cited as “submitted for publication” [5]. Papers that have been accepted for publication, but not yet specified for an issue should be cited as “to be published” [6]. Please give affiliations and addresses for private communications [7].

Capitalize only the first word in a paper title, except for proper nouns and element symbols. For papers published in translation journals, please give the English citation first, followed by the original foreign-language citation [8].

A. Index.php

The landing page is usually considered to be the most important page, because the user enters it mostly directly from search engines. Therefore, it is important to pay close attention to the landing page.

B. Meta tags

It is always better to use different descriptions and keywords on different pages. To search tips for new keywords we used a KeywordPlanner [11]. It is a tool which could help us with creating the best positioned websites.

C. Captions and headings

Title to be inserted into a website is very important. It should contain the keywords should not be overly long and mainly must be listed.

D. Optimization of subpages

It is important to stick to the strict rules for other pages and check for theirs SEO on-page factors. The most important factor is compliance with quality content and strict rules for optimization.

E. Structure of URL addresses

We assume that a well-structured and understandable address easily guide users to the page content. For users who want to refer to our website is a good idea to create a simpler URL address. The best way how to create a sitemap file is to download the special program (e.g. Sitemap Generator).

F. Quality content and useful services

The most important factor in terms of both users and search engines are high-quality content and provided services.

(4)

Appealing to users is the use of new useful services which no other comparable web offers. For users we created e.g. a service to compare indicators of price / performance ratio for the selected item.

G. Websites propagation

An important way to inform visitors about the new websites is to promote them on the appropriate fora. We decided to promote new websites on social networks (facebook, twitter…) and user’s blogs. The results of this promotion appeared later (after several months).

VI. TESTING AND EVALUATION OF OPTIMIZATION To evaluate the optimization of pages, we again used the Google Analytics application (from 15th November 2013 to 15th December 2013). In addition, Google offers another tool called Website Optimizer [12] where you can experiment with your site and find what will work and what will not work. In Table I we can see the positive development of websites after optimization. It is noticeable increase in number of visits and greatly reduced bounce rate, which is very important. We must also note that increased average time on site and percentages of access from the search engines.

Table I Comparison of website development before and after optimization

Before After

Number of visits 621 964

Pages viewing 1548 5976

The bounce rate 59,63% 20,22%

Average time on site 1:15 2:39

Access from referring sites 46,17% 42,33% Access from search engines 31,95% 45,70%

Direct visits 22,88% 10,55%

Since it takes several days or weeks until all search engines “reindex” theirs site content, it is clear that the keywords will have small representation yet. In the first case, indexes landing page index.php. Landing page is dominant and search engines take it as the highest authority.

On Fig. 1 we can see that the largest share of search engines has Seznam.cz (local search engine) – 71.35%. Right behind him is a search engine Google with 26.21%. There are also smaller full-text search engines like Bing and Search.

Fig. 1 Percentage distribution of various search engines

VII. CONCLUSION

The aim of the work was to propose a new structure of web sites in complying with the rules for SEO optimization. First, were analyzed all the websites factors which play a role in optimization. It was found that websites do not satisfy the basic conditions for a good search. Design of a new web sites structure was based on the per-formed analysis. The individual parts were optimized separately with the help of Website Optimizer. The aim was to reduce the bounce rate as low as possible, which succeeded. It was also necessary to increase the percentage representation of approach from search engines. Already the value increased from 31.95% to 45.70%. Individual sites were optimized by the most appropriate keywords and then tested in the search engines. Most of the results ranged from 1st to 10th places, what is the first page in the resulting search (users in most cases say that they are looking at the first maximum second page). From the test results it is apparent that e.g. optimization helped to reduce the bounce rate (the number of site exits) from 59.63% to 20. 22%. When we look at the work from a global perspective, it is clear that the optimization was successful and the newly designed websites are more searchable.

REFERENCES

[1] B. Croft, D. Metzler, T. Strohman, Search Engines: Information Retrieval in Practice, Addison-Wesley; pp. 552, (2009), ISBN:

978-0136072249.

[2] J. Grappone, G. Couzin, Search Engine Optimization: An Hour a Day,

Sybex, pp. 432, (2011), ISBN: 978-0470902592.

[3] D. Dover, E.Dafforn, Search Engine Optimization Secrets, Wiley, pp.

141, (2011), ISBN: 978-0470554180.

[4] P. Kent, Search Engine Optimization For Dummies, For Dummies,

Wiley, pp. 456, (2012), ISBN: 978-1118336854.

[5] M. H. Fleischner, SEO Made Simple: Strategies for Dominating the World's Largest Search Engine, CreateSpace Independent Publishing

Platform, pp. 158, (2013), ISBN: 978-1481838061.

[6] R. L. Adams, SEO Simplified: Learn Search Engine Optimization Strategies and Principles for Beginners, CreateSpace Independent

Publishing Platform, pp. 100, (2013), ISBN: 978-1484831007. [7] R. L. Adams, SEO White Book: The Organic Guide to Google Search

Engine Optimization, CreateSpace Independent Publishing Platform,

pp. 114, (2013), ISBN: 978-1484815083.

[8] R. L. Adams, SEO Black Book: A Guide to the Search Engine Optimization Industry's Secrets, CreateSpace Independent Publishing

Platform, pp. 176, (2013), ISBN: 978-1482665161. [9] Seo Servis. Available : http://seo-servis.cz.

[10] Google Analytics. Available : https://www.google.com/analytics/web [11] Keyword Planner Google AdWords. Available :

https://adwords.google.com/ko/KeywordPlanner.

[12] Google Website Optimizer. Available : https://services.google.com/websiteoptimizer.

Zdenka Prokopova was born in Rimavska Sobota, Slovak Republic in 1965.

She graduated from Slovak Technical University in 1988, with a master’s degree in automatic control theory. She has received in technical cybernetics doctor’s degree in 1993 from the same university.

She worked as assistant at Slovak Technical University from 1988 to 1993. During years 1993-1995 she worked as programmer of database systems in Datalock business firm. From 1995 to 2000 she worked on position lecturer at Brno University of Technology. Since 2001 she has been at Tomas Bata University in Zlin, Faculty of Applied Informatics. She presently holds the position of associating professor at the Department of Computer and Communication Systems.

(5)

Her research activities include programming and application of database systems, mathematical modeling, computer simulation and control of technological systems.

Radek Silhavy was born in Vsetin in 1980. He received a B.Sc. (2004),

M.Sc. (2006), and Ph.D. (2009) in engineering informatics from Faculty of Applied Informatics, Tomas Bata University in Zlin.

He is a senior lecturer and researcher at the Computer and Communication Systems Department.

His Ph.D. research was on the verification of the distributed schema for the electronic voting system. Major research interests are software engineering, empirical software engineering and system engineering.

Petr Silhavy was born in Vsetín in 1980. He received a B.Sc. (2004), M.Sc.

(2006), and Ph.D. (2009) in engineering informatics from Faculty of Applied Informatics, Tomas Bata University in Zlin.

He is a senior lecturer and researcher at the Computer and Communication Systems Department.

His Ph.D. research was on the electronic communication and services in medical information systems. Major research interests are data mining, database systems and web-based services.