ONLINE MARKETING MODEL FOR EVALUATING THE EFFECTIVENESS OF COMPANIES' RELATIONSHIPS WITH END CUSTOMERS

Abstract

A model for creating effective relations, based on inbound marketing, enables established companies which operate in online consumer markets to assess the strengths and weaknesses of their marketing approach at each stage of their communication with customers. It also outlines the funnel scheme of companies' sales. In addition, the evaluation of global indicators gives a comprehensive idea of the effectiveness of marketing activities and the chosen marketing tools. The importance of the factors influencing the local business environment in the model can be assessed by preparing a matrix of uniform criteria applicable to the various sectors and organizations selling their products and /or services on consumer markets in the B2C segment. The main aim of the article is to present a model for evaluating the effectiveness of companies' relationships with end customers, and justify theselection of criteria and the opportunities for their application.

Headnote

Abstract:

A model for creating effective relations, based on inbound marketing, enables established companies which operate in online consumer markets to assess the strengths and weaknesses of their marketing approach at each stage of their communication with customers. It also outlines the funnel scheme of companies' sales. In addition, the evaluation of global indicators gives a comprehensive idea of the effectiveness of marketing activities and the chosen marketing tools. The importance of the factors influencing the local business environment in the model can be assessed by preparing a matrix of uniform criteria applicable to the various sectors and organizations selling their products and /or services on consumer markets in the B2C segment. The main aim of the article is to present a model for evaluating the effectiveness of companies' relationships with end customers, and justify theselection of criteria and the opportunities for their application.

Key words: inbound marketing, marketing relationships,online model

JEL Classification: M3

(ProQuest: ... denotes formulae omitted.)

INTRODUCTION

Nowadays, consumers are constantly "flooded" with hundreds of marketing messages, which leads to a growing intolerance of such marketing tactics. (Godin,2010)A recent trend in the constantly evolving competitive globalised online environment is the application of a new approach called inbound marketing, which has become one of the most successful forms of marketing practiced by organizations involved in diverse economic activity. Its rapid development turns inbound marketing into a preferred form in a dynamic and deeply personalized online environment. The Internet can be considered a driving marketing environment in which companies try to attract the interest of their potential customers who are already looking for information on various solutions. (Holliman and Rowley, 2014,269-293)At the same time, companies try to retain their consumers, to such an extent as to develop long-lasting loyalty-based relations of bilateral nature, which benefit both parties. Establishing good relationships with customers is a time-consumimg process. Organizations themselves come to the conclusion that relationships with customers yield results not only in terms of revenue but also in terms of productivity. Loyal customers can not only bring more income than new customers, but also contribute to attracting new customers and improving the product. Studies shows that 44% of companies focus on attracting customers and 18% focus on retaining them while attracting customers is 5 times more expensive than retaining them, and increasing the number of loyal customers could lead to an increase of revenues from 25% to 95% (Visually, 2020)

In today's dynamic online environment, there are software applications for researching the marketing approach in real communication campaigns. Such applications were created by the company HubSpot, which develops and presents real solutions. Their disadvantage is the lack of consideration and focus in the marketing relationships with regular customers. However, the various forms and tools used by organizations on the Internet do not answer the specific questions companies might have about their appropriateness and their use. The need for a model for evaluating these relationships stems from the trend in the development of the online environment and the desire to develop interconnections in order to improve the results of companies' work.

The article aims to present a model for evaluating a marketing approach, based on inbound marketing and relationship marketing. The main aimof the presented model is to determine the role of the various forms and tools used by organizations which operate on the Internet. For this purpose, a matrix with uniform criteria was developed. The model can be applied to the evaluation of any company that communicates with real and potential customers on the Internet, and end customers who might place new orders, so that the results allow comparability. The potential users of the model involve:

* Companies and organizations with a permanent presence on the Internet - for assessing their marketing approaches and forms of presentation;

* Institutions, branches and other organizations - when choosing their marketing orientation and building trust in the respectiveinstitution.

METHODOLOGY

The aim of the study was to develop an online marketing model for evaluating the effectiveness of business organizations' relationships with end customers. In view of this, a combination of methods was applied, which allowed to provide sufficient necessary information at the different stages of creating the model, and to conduct adequate discussions. The indicators for measuring the effectiveness of existing digital campaigns were examined through a critical analysis of research articles in the field of digital marketing.

An online survey was conducted to collect quantitative data. It was sent to a sample of 200 randomly selected emails of departments or specialists in the field of marketing, advertising, PR, digital media, or to managers, as well as to 200 random online stores. The questionnaire included 21 questions, 20 of which with a specific indicator and a choice between three main stages - the stage of attracting customers, the stage of customer retention and the stage of developing relationship with customers, as well as an option if the respondent considers the indicator inapplicable. Through the method of expert assessment, combining indirect observation and opinion poll, an expert opinion and assessment were obtained, which are necessary for the successful functioning of the model.

The limitations of the study can be seen in relation to some types of parameters in terms of sales, especially in companies with a permanent presence on the Internet. The model considers only the relationships with end customers, ie. building relationships in consumer markets; it includes indicators visible only to companies using analytic software applications to track the effectiveness of marketing campaigns (such as Google Analytics). Other indicators can be derived from software systems for customer relationship management, marketing automatisation and company databases in generalcontaining information confidential to the company.

The model uses real data and information obtained from the companies. It can be used for each year (campaign), taking into account the change of indicators for improving the marketing approach and building effective relationships.

1.GENERAL CHARACTERISTICS OF THE MODEL

The model comprises three main parts:

* Part 1 - Evaluation of the marketing approach through analysis of the tools used

* Part 2 - Analysis and evaluation of global performance indicators.

* Part 3 - Integral assessment - a combination of the assessments of the first two parts.

The model requires an analysis of qualitative and quantitative indicators. Each of these indicators has a weight of significance determined by expert assessments.

Most of the quantitative indicators are derived from analytical systems for tracking the results of marketing campaigns on the Internet, such as Google Analytics, and the quality is monitored and evaluated by specialists responsible for online communication with customers in business organizations.

In order for the results to fall into a matrix, a form of the checklist type is filled in, where the specific results from the filled in data are displayed.

Table 1 and Table 2 present the general structure of the first and second part of the model.

The methodology for calculating weights is presented in item 3 - Toolkit for determining relative weights

Part 3 - Integral (complex) assessment of the two criteria - marketing approach and global performance indicators.

2.REGULATION OF THE EVALUATION LEVELSOF INDICATORS IN THE MODEL

The criteria in the matrix are mainly structured in two large groups: qualitative and quantitative. The qualitative criteria affect the information security with the respective communication tools, and the quantitative ones cover the value indicators, which are realized as a result of the activity.

* Accessibility - Web presence is guaranteed by the number of digital platforms through which companies inform the public and their potential users.

For the purpose of the research, three levels of web presence were regulated, depending on the number of owned digital platforms:

o From 0 to 2 platforms - low level

o From 3 to 5 platforms - intermediate level;

o Over 5 platforms - high level.

* Visibility - the extent to which the company is visible in search engines. Due to the difficulty of measuring this indicator, it is necessary to implement additional tools for optimization evaluation. The evaluation is taken from the website website.grader.com, by entering the url address of the company's website.

o From 0 to 10 - low level

o From 11 to 20 - intermediate level;

o Above 20 - high level

* Organic search - The results of the study by Nayar and Pandey 2016 (Nayyar and Pandey, 2016,5-9), show that 60% of organic traffic is due to search engine optimization. The following levels of organic search are determined:

o up to 30% - low level

o from 31 % to 60% - average level

o over 60% - high level

* Dropout rate - an indicator whose values give an idea of the attractiveness of a website's content. Bounce rate is measured as a percentage (%) of dropouts, and is reported when users leave the website by visiting only one page, without interacting in any way, and without visiting another page of the same site. These one-page sessions have a relatively short duration which cannot be traced.(Google support, 2019)

According to a study by RocketFuel, the dropout rate of most websites varies between 26% and 70%. (Payton, 2019)The following levels of dropout rate are defined:

o At Bounce rate values lower than 40% - high level;

o At values of Bounce rate between 40% and 60% - average level;

o At Bounce rate values higher than 60% - low level.

* Content - The content evaluation indicator includes two main sub-indicators - a web performance evaluation indicator and a website relevance indicator.

Optimizing website performance is critical to increasing traffic, improving information exchange, generating more leads, and increasing revenue. The indicator for measuring the effectiveness of web content is also complex, as it takes into account the weight of the website the loading time, etc., and will consequently be taken as a basis for the indicator of web performance on website.grader.com. The evaluation of web performance varies from 0 to 30. We outline three main ranges in determining the level of web performance.

o From 0 to 10 - low level;

o From 11 to 20 - intermediate level;

o Above 20 - high level.

In order to determine the extent to which the indicator of timeliness of marketing content is within certain limits, it is necessary to obtain data on the average frequency of updating content on the Internet.

We consider web content up-to-date if it has been updated within the last 3 months. If the information is updated within a period of 3 months to 1 year, we have an average level of timeliness; if the content has not been updates for up to 1 year, we have a low level of timeliness.

For the purpose of the study, the following levels of timeliness are determined:

o Over 12 months without update - low level

o From 3 months to 12 months from the last update - medium level

o Up to 3 months from the last update - high level

Falling into the specific levels of timeliness and web presentation, the results for the indicators are distributed in the following matrix(Figure 1), regulating the respective levels of evaluation of web content:

When using the matrix, a high level (green zone) of web evaluation is assigned when the indicators of timeliness and web performance fall into the three squares in the upper right corner of the matrix. Alternatively, we have a high level in combinations of high level of web performance, with high and medium level of timeliness of content, and medium level of web performance with a high level of timeliness.

A medium level (yellow zone) is assigned when the indicators of timeliness and web performance fall into the three squares along the diagonal of the matrix. Or, in combinations of low level of web performance, high level of timeliness, low level of timeliness with high level of web performance, and medium levels of web performance and timeliness.

A low level (red zone) is assigned when the timeliness of content and web performance indicators fall into the three squares at the bottom left corner of the matrix. Or in combinations of low level of timeliness with low and medium level of web performance, and medium level of timeliness and low level of web performance.

* Functionality - this indicator gives a clear idea of the functionality of a company's internet communication, depending on the presence or absence of specific functional tools.

The evaluation of the indicator, namely functionality, is determined depending on the presence or absence of listed functional tools. The following levels are defined:

o From 0 to 2 tools - low level

o From 3 to 5 tools - intermediate level;

o Over 5 tools - high level.

* Retention rate - The retention rate indicator depends on two main metric indicators that companies can enter from their analytics applications, unique visitors and returning visitors.

Due to the nature of the model, which emphasizes the retention of consumers who perform reliable interaction with the company, we defined the following equation:

... (1)

where Rvr stands for retention rate, Uv -for unique visitors, and Rv - for returning visitors

According to Neil Patel, at Rvr values of around 30%, web content can be considered engaging. (Patel, 2019) In this sense, the assessment of the retention coefficient according to its value is presented as a percentage (%) and is defined as follows:

o When the value of the retention coefficient is less than 10% - low level

o At a value of the retention coefficient between 10% and 30% - average level;

o At a retention factor value higher than 30% - high level.

* Conversion rate - this indicator shows the number of visitors who have performed a certain action - made an inquiry, registered, subscribed to an e-newsletter, placed an order, etc. The conversion rate is measured as a percentage (%).

Google Analytics provides the option to display the conversion rate after a setup and selected goals for specified conversions. According to a Wordstream study (Baadsgaard,2019) conducted among hundreds of Google Ads accounts, the average conversions come to 2.35%. The lowest conversion values are from 0-1%, and the best values are for accounts with conversion rate above 5%. The following levels are established:

o At conversion rate above 0% but less than 1 % - low level;

o At conversion rate from 1% up to 2% - average level;

o At conversion rate over 2% - high level.

* Additional content - this indicator determines the extent to which companies provide additional free content in their leads. The additional content indicator is determined on the basis of the presence or absence of additional content such as webinars, podcasts, e-books, e-newsletters, e-mail offerss, video content, and other free online content;

The level of the indicator is determined depending on the availability of additional content:

o From 0 to 2 - low level;

o From 3 to 4 - intermediate level;

o Above 5 - high level.

* Reference - the indicator of reference is a system of recommendations, comments, rating scales, feedback surveys and interaction assessment (when buying, assisting and helping customers).

Three levels of reference are specified:

o In the absence of a system of recommendations, comments section, rating scales and feedback surveys - low level;

o In the presence of one to two of the listed - intermediate level:

o In the presence of three to four of the listed - high level.

* Trustworthiness - The indicators for trustworthiness represent the two-way relationship between the surveyed sites, the presence of which helps to generate factors to promote trust. In the four groups of indicators - for reliability, responsibility, flexibility, and integrity, the respective measured values are determined. Then the points for each indicator in the presence or absence of relevant tools are summed. The resulting maximum score is 13, and the minimum score is 0.

According to the assessments made and the factors for promoting trust, the following levels are specified:

o For total assessment in the interval from 0 to 4 - low level;

o For total assessment in the interval from 5 to 8 - average level;

o For total assessment in the range from 9 to 13 - high level.

* Activity - the indicator shows the extent to which the company's customers are engaged with the content of the site and are active in their interaction with the company. In terms of interaction and direct communication, the indicators of average duration of 1 visit (session), and average number of pages viewed per 1 visit (session) can be considered.

A study by Littledata (Little data, 2019)conducted in August 2019, which surveyed 3,623 sites, found that an average of 3 pages per visit are viewed per session. More than 4.7 pages viewed per visit, is considered a benchmark for content engagement and good user activity. 20% of the surveyed sites fall into this category.

A relatively poor result is registered when 1.8 pages are reviewed in one session. Another 20% of the surveyed sites achieve such an unsatisfactory result.

The average session duration is an indicator used by Google Analytics, which takes into account the average amount of time users spend on the company's website. A study by Databox shows that the average duration of 1 session is in the range of 2 to 3 minutes (55% of the surveyed sites). (Albright, 2019)

Based on the presented information, an activity indicator is determined, and can be expressed by the following equation:

A = Pps-AvsD (2)

whereA - activity indicator

Pps - an indicator of the average number ofpages viewed per visit

Avsd - an indicator of the average duration of one 1 session

For the purpose of the work we determine levels of evaluation of this indicator as follows:

In option 1, when the average duration of a session is less than 2 minutes, we determine the minimum as 120 seconds. The minimum number of pages viewed is 1.8. The activity indicator, provided that values below the two minima are realized, is equal to the multiplication product of 1.8 and 120, ie. the activity indicator in the first variant is less than or equal to 216, which is determined as low level.

In option 2, when the average duration of a session is longer than 2 minutes and less than or equal to 3 minutes, the minutes are again converted to seconds. The average number of pages viewed for 1 session is in the range of 1.8 to 4.7. The values of the activity indicator in the range from 216 to 846 determine an average level.

In option 3, when the average duration of a session exceeds 3 minutes and the average number of pages viewed exceeds 4.7,the values of the activity indicator of over 846 determine is determined as a high level of the indicator.

o At A <= 216 - low level;

o At 216 <A <= 846 - average level;

o At A> 846 - high level.

* Compatibility ratio (CLV vs CAC ratio) - the indicator presents the ratio between the cost of acquiring customers (CAC), to the lifetime value of customers (CLV) for the period under review. This indicator shows what part of the funds is spent on acquiring customers, and what the revenues from customers for a certain period of time are. The calculation of the indicator supports the orientation of the companies regarding the correctness of the applied marketing approach.

* Coefficient of marketing return - one of the indisputable advantages of the Internet is the possibility to accurately measure the results of advertising campaigns, such as: clickthrough rate (CTR), conversions (Conversions), range (Reach), frequency (Frequency), position (Position), cost-per-click (CPC), cost per 1000 impressions (CPM), etc. The coefficient of marketing return is the most important indicator for any advertiser, because it gives the clearest idea of the real effect of advertising, both in the digital environment and outside it. The coefficient of marketing return is the ratio between the net cash flow from the investment and the costs incurred for the investment.

* Coefficient of loyalty - the term loyalty itself implies a wide range of components. The choice of components is tied to the purpose of the study.

3.TOOLS FOR DETERMINING RELATIVE WEIGHTS

For each criterion in the model, a weight is determined, which adjusts the specified measure. The overall weight of the model is 2.

One tool for determining the relative weights is a survey conducted with 100 experts in the field. Each respondent gave a score from 0 to 10 to the importance of each criterion. The values in the assessment are determined by 0 - for the lowest level of significance of the criterion, and 10 - the highest level of significance.

The survey data were summarized through Excel. When conducting the survey, the main requirements to the sample were met - the respondents were not biased, and were selected at random.

The measurement of the weights for each individual criterion is based on the relationship between the sum of the scores for the given criterion and the total sum of the scores for all criteria.

The evaluation of the i-th criterion is the sum of the evaluations of all respondents for the criterion:

Bi = Bi1 + Bi2 +. + Bin, where: (3)

Bi - total score for the whole criterion i (selected criterion)

Bi1 - the assessment for the i-th criterion given by the first respondent

Bi2 - the assessment for the i-th criterion given by the second respondent

Bin - the assessment for the i-th criterion given by the n-th respondent

The total amount of the assessment for all criteria can be determined by the formula:

... (4)where:

B - total amount of the evaluation for all criteria

Bi - total score for the whole criterion i (selected criterion)

Based on the formulas described above, we can derive a formula for measuring the weight of each criterion:

... (5) or

after simplifying the denominator, according to formula (4), the formula will have the form:

... (6) where:

Wi - weight of the i-th criterion (selected criterion)

Bi - total score for the whole criterion i (selected criterion)

B - total sum of the evaluation for all criteria

Measured by formulas (5) and (6), the weights for the relevant criteria are as follows:

...

4.TOOLS FOR DERIVING INTERIM AND FINAL EVALUATION OF BUSINESS ENVIRONMENT IN THE MODEL

The interim grades in the model are formed by summing the scores obtained for the individual criteria. This allows us to obtain unambiguously defined quantitative estimates. The range of possible results for each intermediate level is defined between the minimum and maximum possible value. For each level, intervals are defined with a corresponding median assessment. Five possible grades were made, namely: excellent, very good, good, unsatisfactory, and bad.

An arithmetic principle was used to determine the width of the intervals. According to the same width for each interval is the ratio between the difference formed by the maximum and minimum value of the feature and the number of groups.

"A distinctive feature of the arithmetic principle is the uniform width of the group intervals. If k is the number of groups, then the width h is obtained from the ratio of the difference between the maximum Xmax and the minimum value Xmin of the attribute and the number of groups: "(Mihalev, 2016,21)

... (7) where:

h is the width of the interval;

Xmax - themaximum value of the feature shown in the matrix;;

Xmin - the minimum value of the feature shown in the matrix;

k - the number of groups.

The model displays 2 interim grades - for qualitative and quantitative criteria. For their part, partial gardes can also be made for the qualitative criteria, respectively for the legal and administrative criteria.

Determining the width of the interval for the individual interim grades Interim evaluation of the criteria for Part 1

The scale of results is in the range from 0.00 to 1.2. The most pessimistic option allowed in the model gives a final grade of 0.00. The maximum score in this part gives a grade of 1.2.

Determining the width of the assessment interval for Part 1 is done by the formula

(7):

...

The scale of the interim results obtained in the first part of the model can be graded in ascending order as follows:

...

Interim evaluation of the criteria for Part 2

The range of scores for the criteria of an administrative nature is from 0.00 to 0.8.

To determine the width of the assessment interval in Part 2, formula (7) is used again:

...

The scale of the interim results obtained in the section Analysis and evaluation of global performance indicators can be ranked in ascending order as follows:

Tools for deriving final evaluation in the model

Producing a general evaluation of the model in part 3 forms the integrated evaluation according to the two criteria - the marketing approach and the global efficiency indicators. Calculating results in both parts of the model generates scores in the form of grades as a final result. The definition of an integrated (complex) assessment makes it possible to determine strategies depending on the results for the respective criterion and the effectiveness in part 2.

The final grades in the model can be formed by summing the grades obtained for the interim levels. The limits of possible results are defined between the minimum and maximum possible value for the results from the two parts of the current model. The range for each final grade is determined. To ensure consistency and comparability between the interim and final evaluation, the same assessment scale is used, namely: excellent, very good, good, unsatisfactory, poor.

An arithmetic principle based on the formula (7) was used to determine the width of the intervals.

...

From the width of the interval thus calculated, the estimates should have the following limits:

An integral assessment can also be defined in the form of a combination of the results of the two parts of the model in the form of a matrix(Figure 2):

5.RESULTS AND SOLUTIONS

The derived integrated assessment objectively analyzes the marketing techniques and their effectiveness, examining the impact of applied marketing tactics, tools and strategies used to improve the effеctiveness of buisnesses working online. An analysis of online approaches and their effectiveness can be done on an annual basis in order to improve customer relationship strategies. The evaluation can be conducted annually, showing changes in the improvement of unsatisfactory results from baseline or previous analyzes.

* Poor evaluation is given in case a company does not show much concern about the means and approaches to its internet marketing. It is possible to implement the strategy "Structuring", in which new user structures are formed for advertising, taking into account the specifics of the emotional impact, enriching the expressive means of advertising in order to communicate faster and more clearly so as to be retained longer in the memory of the users. The strategy is characterized by clear domination of the image over the text in advertising, as well as by superiority of visual over verbal advertising. The "Information" strategy is also relevant inits seeking to create an image and build an opinion about an organization, the directions of its work, its organizational condition and financial situation, social groups, contractors, reference structures, etc. Its persuasiveness is achieved through a natural and objective presentation, sophistication and competence in communication.

* In the case of unsatisfactory evaluations company might have taken a marketing approach with obvious shortcomings and inaccuracies regarding the applied tools. In this case it is advisable for it to use the strategy "Linking"(Banchev,2010,556-594), which seeks to build alternative solutions for direct feedback. It is achieved through direct advertising which targets potential buyers, and the development of address files and data banks of customersto be contacted directly.

* In the middle "good", "yellow" zone the results vary. They can be improved significantly provided that the model recommendations are largely followed. Strategic "Friendliness" is applied, which relies on friendly attitude and kindness to buyers, and full understanding of their problems. It is based mainly on personal acquaintances and connections. The emphasis is on a personalized approach and creating relationships. Strategically, the most sustainable combination is achieved in the implementation of the strategy "Mutual benefit", oriented to understanding the problems of buyers, engaging in their solution, and seeking mutual benefits from the purchase. It is hoped that this will establish long-term business relationships.

* In the "very good" assessment, it is recommended to implement a "Maintenance" strategy. It aims to maintain public interest in the organization and its products by constantly reminding customers about them. The aim is to preserve the positions won, and to prevent a decline in interest and reorientation to rival companies.

* In the green zone, the companies rated "excellent" clearly follow the recommendations of marketing managers in applying an inbound marketing approach, which improves customer relationships and, as a result,their online activities are highly effective. The implementation of the "Prestige" strategy is recommended, which is based on maintaining and sustaininghigh-level business reputation (using modern technologies, offering innovative products, etc.).

The model is built as an open system that can be easily changed and adapted to any change in technology, as digital technologies change with significant dynamics over time. The convenience of the model lies in the ability of each of its users to implement it, by analyzing the approaches that are applied on the Internet.

References

REFERENCES

Banchev,P., Marketing, Faber, Veliko Tarnovo, 2010, pp.556-594

Godin, S., Permission Marketing: Turning Strangers into Friends and Friends into Customers,Izdatelstvo Focus , Sofia, 2010

Holliman, G. & Rowley, J., Business to business digital content marketing: marketers' perceptions of best practice. Journal of Research in Interactive Marketing, 8(4), 2014, pp.269-293

Mihalev, D.,Selected lectures on statistics, pp.21, http://math.vtu.bg/Stat.pdf, 12.09.2016

Nayyar, T., Pandey, A., An Analytical Study of Search Engine Optimization (SEO) Techniques: To Maximize Number Of Visitors On An E-content Website, International Journal of Technology and Science, ISSN (Online) 2350-1111, (Print) 2350-1103 Volume VIII, Issue 3, 2016 pp. 5-9

Albright, D., Benchmarking Average Session Duration: What it Means and How to Improve It, https://databox.com/average-session-duration-benchmark Available at: [Accessed 5.09.2019]

Bounce rate, https://support.google.com/analytics/answer/1009409?hl=en Available at: [Accessed 14.08.2019]

Baadsgaard, J. "How to Improve Your Conversion Rate By 50% in One Day" https://neilpatel.com/blog/improve-your-conversion-rate-50-percent/ Available at: [Accessed 31.07.2019]

Visually, Customer Acquisition Statistics And Trends, https://visual.ly/commumty/Infographics/business/customer-acquisition-statisticsand-trends?utm_source=visually_embed, [last available at 26.04.2020]

Patel, N., How to Measure Reader Engagement and Loyalty Using Google Analytics https://neilpatel.com/blog/how-to-measure-reader-engagement-and-loyalty-usmggoogle-analytics/ Available at: [Accessed 15.08.2019]

Payton, J., "What's the Average Bounce Rate for a Website?", https://www.gorocketfuel.com/the-rocket-blog/whats-the-average-bounce-rate-ingoogle-analytics/ [Accessed 14.08.2019]

Little data, What is an average pages per session? https://www.littledata.io/average/pages-per-session-(all-devices) Available at: [Accessed 5.09.2019]