The methodology is divided into three parts: interview with responders about their level of prior knowledge in both industries - gastronomy and pharmaceutics, observation of participants` path with the coding key-decisions to a database, then interview with the responders to understand the factors, that may effect on his decision making path. The reason of choosing distinctive industries is that the gastronomy sphere, where a person should book a table in a restaurant for a family dinner is a part of people`s daily life, while pharmaceutics need a specific knowledge and experience. Two different industries allow to see are there any differences in the impact of prior knowledge in decision making process, as it is supposes that for responders will be easier to choose a restaurant for dinner, than pills for stomachache.

2.1 Data: sampling issues

The research has population, consisting of population of Russia that is 144.5 million people, as the web site will be in Russian and has Russian domain. To analyze this, the convenience nonprobability sampling method should be used. The reason of choosing such sampling design is that there is no ability to get the whole list of general population and calculate the probability of each responder being included in the sample.

The responders` age should be from 18 years old, because there is a need to book a table in a restaurant and buy a specific medicine online and teens have not got these opportunity, to 55 years old, as older people, who use the Internet, represent the smallest percentage of all Internet users; the participants might have a device, with the help of which they will complete the task; the device should have the Internet connection and functions of screen recording in order observe responder`s route.

By the way, such demographic characteristics, as gender, marital status and education are not important, because it does not reflect on the level of usage devices and self-confidence, while searching the Internet, so such demographics characteristics will not be analyzed.

The sample size is 50 participants because the observation will be collected by qualitative approach and qualitative and quantitative analysis will be presented in the research.

The collected sample consists of 50 observations for each product category. All the participants are located in Saint Petersburg, Russia. The participants are between 18 and 55 years old, use a phone or computer with the Internet connection and have a screen recording on their devices. Responders in the sample are represented by diversity in order to avoid biases and to ensure that the sample is representative.

2.2 Research design, data collection & processing tools

In order to have full image of current situation cross-sectional timeframe is most consistent because it investigates information about prior knowledge and online decision-making process at one point in time.

In order to achieve the empirical aim of the paper, the data base was created from several resources: data from interviews and observations. The data collection process from the gastronomy and pharmaceutics industries are presented in details for groups of evaluators and loyalty loopers in Appendix 1, Appendix 2, Appendix 3 and Appendix 4. Each participant was asked to complete the following tasks: to book a table for family dinner via the Internet and to purchase a pill for a stomachache with the help of the Internet. Before the observation there is a need to understand the level of prior knowledge of the participant by an interview, as prior knowledge is the latent variable and it cannot be measured directly. The level of prior knowledge consists of two items, according to Bruck`s model from the literature review, where an individual has knowledge about the industry and the product itself. Therefore the responders were asked two several question for each industry: “What do you know about the gastronomy sphere in Saint-Petersburg? - The answer on this question gives the information about the level of prior knowledge of the participant for the gastronomy industry. “Do you know any family restaurant to organize a family dinner?” - The answer on this question provides information about the level of prior knowledge of the participant for product, as a specific restaurant. “What do you know about the pharmacy market? - The answer provides the level of prior knowledge about the pharmaceutic industry. “Do you know any pills for a stomachache?” - The answer on the question gives information about the level of prior knowledge for a product - pills for stomach pain. After the interviewing the participants start to complete the task and all their decisions were recorded on the video screen and coded to a data base. So, the method of data collection on the second stage is observations. There are several reasons of choosing this method. First of all, one of the best ways to track the process of online decision making in order to knowledge of the final result does not influence on perception. Secondly, observation of consumers facilitates identification different behavioral patterns. Thus, the method that permits to keep a record of the process in real time suits the best and reveals patterns of decision making process. Further, there is a need to conduct extra interviews with the respondents in order to understand why they choose a specific restaurant or pills, what other factors, which are not included in the data base, were significant in decision making process. Finally, the qualitative analysis was done by creating an online decision making path of the participants. All collected data is coded as categorical variables.

2.3 Data: operational definition

The database is analyzed with the coding system, where dependent variable is presented, as stages of decision making process, excluding post-purchase behavior because the aim of the research is to find the way how the purchasing decision process unfolds and leads to product selection, and independent variables, as the level of prior knowledge, time spent, participant`s age, use of feedbacks and search for benefits.

Table 5Variables coding

It was found that all responders have such steps, as problem recognition, initial consideration set and moment of purchase, after collecting the data. It was decided to divide the dependent variable into three categories, for which there were differences in the responders` decision making steps (Table 6).

Table 6 Coding the consumer decision making path

According to McKinsey Customer Journey customers, those who are in loyalty loop, was once as evaluators and went through all stages of decision making process. Thus, the loyalty looper is estimated as the highest point - 2, the evaluator as 1 and participants that are indifferent as 0.

2.4 Descriptive statistics of variables

While conducting descriptive statistics for each independent variable for both industries, there were following results, according to the table 7:

Table 7Descriptive statistics of variables

3.Results

3.1 Quantitative Analysis

To analyze the correlation between dependent and independent variables the correlation matrix for gastronomy sphere was created (Table 8).

Table 8. Correlation matrix for gastronomy industry

* - significance at the p<0,05 level

The multicollinearity between the variables from the gastronomy sphere was checked and there was found that there is a reverse medium correlation with the 95 percent confidence interval between prior knowledge for product and feedbacks. The relationship can be explained as the more people look for a feedbacks and recommendations the lower prior knowledge about the restaurant they have. Moreover, the significant direct middle correlation was found between prior knowledge for product and prior knowledge for industry. The correlation between these variables can be clarified because the more a person knows restaurants for family dinner the higher prior knowledge about the gastronomy sphere he has. The reverse significant medium correlation between the time which is spent by the participant and the usage of feedbacks can be explained as the more feedbacks a participant are searching for the less time he spend on decision making process, as feedbacks can improve participant`s prior knowledge. The reverse significant correlation between the time which is spent by the participant and his age can be clarified as the older the person the higher knowledge and experience he has and for this reason a participant spend less time on purchase decision making process. The direct significant relationship between feedbacks and benefits might be interpreted as if a participant looks for a feedback about the restaurant to get extra knowledge, benefits also might be analyzed.

To understand is there the significant impact of prior knowledge for product and industry on the online purchase decision making process the analysis of variance for gastronomy industry was constructed (Table 9).

The null hypothesis is presented as there is no difference between groups. The model is significant due to Prob>F is 0.0000 and the determination coefficient is high and compose 0.7431, so the model matches the data. According to ANOVA the null hypothesis is rejected due to the significance of deviation. It can be seen, that such independent variables as prior knowledge for product and time are significant. Evaluators, loyalty loopers and indifferents are different from each other in the level of prior knowledge for a specific restaurant for family dinner and time, which is spent by the participants on online purchase decision making process. In other words, the level of prior knowledge for product and time influence on online decision-making process.

It is interesting to go further and find out exactly where the differences lie, i.e. which groups differ from each other. Due to the fact that ANOVA does not show in what group the difference is, there is a need to perform pairwise comparisons of the average values of existing groups. However, the student's criterion for such comparisons is not applicable due to the effect of multiple comparisons. Therefore to perform a large number of pairwise comparisons of group averages without loss of statistical power it is necessary to make a Tukey`s honestly significant difference test, which help to identify on what decision making steps the level of prior knowledge for restaurants and time are influenced the most.

Table 9 ANOVA for gastronomy industry

Through the Tukey`s test can be seen that there is difference between loyalty loopers and evaluators, as P>0.05. Thus, the level of prior knowledge for gastronomy sphere are influenced on loyalty loopers and evaluators. The level of prior knowledge for gastronomy sphere of loyalty loopers is lower than evaluators` level (Table 10).

Table 10 Tukey`s test for gastronomy sphere for prior knowledge for product

The Tukey`s test on the difference between group in time spent shows that loyalty loopers and evaluators; loyalty loopers and indifferents are different from each other in time, which participant spend on online purchase decision making process, because P>0.05. Loyalty loopers spend less time on online purchase decision making process, than indifferent and evaluators do (Table 11).

Table 11Tukey`s test for gastronomy sphere for spend time

Taking everything into account, the level of prior knowledge for industry and spend time have a significant impact on online decision making process. The level of prior knowledge for industry in a gastronomy sphere influences on loyalty loopers, where loyalty loopers have a lower level of prior knowledge than evaluators; the spend time in gastronomy sphere influences on loyalty loopers, where this group spend less time on online purchase decision making process, than evaluators and indifferents.

To analyze the correlation between dependent and independent variables the correlation matrix for pharmaceutics sphere was created (Table 12). The multicollinearity between the variables from pharmaceutic sphere was checked and found out the reverse significant medium correlation between the feedbacks and prior knowledge for product. It can be explained as the more participant searches for feedbacks the lower prior knowledge for product in pharmaceutics he has. The direct significant middle relationship between time and prior knowledge might be clarified because if a participant has a high level of prior knowledge for stomachache pills, the more things he try to analyze, for instance, the pill`s composition and its instruction manual. The direct significant medium correlation between prior knowledge for industry and prior knowledge for product can be explained as the higher level of prior knowledge about the stomachache pills a person has the higher his level of prior knowledge for pharmaceutics. The reverse significant medium relationship between time and feedbacks can be explained as the more feedbacks and recommendations the participant searches for the less time he spend on purchase decision making process, as feedbacks improve the level of prior knowledge.

Table 12Correlation matrix for pharmaceutic sphere

* - significance at the p<0,05 level

To understand is there the significant impact of prior knowledge for product and industry on the online purchase decision making process the analysis of variance for both industries was constructed (Table 13).

The null hypothesis is presented as there is no difference between groups. The model is significant due to Prob>F is 0.0000 and the determination coefficient is high and compose 0.7539, so the model matches the data. According to ANOVA the null hypothesis is rejected due to the significance of deviation. It can be seen, that such independent variables as prior knowledge for product, prior knowledge for industry and time are significant, as Prob>F is lower than 0.05. Evaluators, loyalty loopers and indifferents are different from each other in the level of prior knowledge for a pharmaceutics sphere and the level of prior knowledge for specific pills for stomachache. In other words, the level of prior knowledge for product and prior knowledge for industry influence on online decision making process.

Table 13ANOVA for pharmaceutic sphere

In order to find out on what group the level of prior knowledge for product and industry influence a lot the Tukey`s test was made (Table 14). The significant difference can be seen between indifferents and evaluators; loyalty loopers and evaluators. The level of prior knowledge of pharmaceutic industry is higher for evaluators than indifferents` and loyalty loopers` level.

Table 14Tukey`s test for pharmaceutic sphere for prior knowledge for industry

Tukey`s test on a level of prior knowledge for a stomachache pills shows that there are significant difference between indifferents and evaluators; loyalty loopers and evaluators. People who do active evaluating of alternatives have higher level of prior knowledge for stomachache pills, than loyalty loopers and indifferents (Table 15).

Table 15Tukey`s test for pharmaceutic sphere for prior knowledge for product

To sum up, the prior knowledge for product and industry in the pharmaceutic sphere affect the decision making process significantly. The level of prior knowledge for pharmaceutic industry is higher for evaluators, as well as the level of prior knowledge for stomachache pills.

According to analysis of variation and Tukey`s test for gastronomy and pharmaceutic industry the several similarities and differences. First of all, the level of prior knowledge for industries influences on online purchase decision making process in both spheres. By the way, decision making process for gastronomy sphere also is affected by spend time, where loyalty loopers make a decision more quickly than people who make an active evaluation of alternatives or do not go through all the steps of decision making process. However, the pharmaceutic industry is influenced by the level of prior knowledge for the stomachache pills, where people who evaluate the alternatives have higher prior knowledge for a product, than those who are loyal or indifferent.

3.2 Descriptive data analysis

In order to evaluate customers' online purchase decision-making process descriptive type of analysis was conducted. The analysis of the online purchase behavior for the 50 participants was made on the basis of two tasks: to choose a restaurant for the evening and a pharmaceutical from a stomachache. As the research is considered to be about online purchase decision-making process the “moment of purchase” was stranded as booking a table in a restaurant a and putting goods in the online pharmacy basket. Firstly, each responded conducted two tasks using his or her gadget and then was asked some questions in order to make a decision-making path more detailed and fully explained. The video data was coded using cues to identify activities that were undertaken by the participants, which were then assigned to stages. Figure 2 illustrates the combined online purchase decision-making model for all individuals. The black line shows the path of the respondents which were grouped as “evaluators” - people, who include in their decision-making process the stage of comparing and evaluating alternatives. Dotted line illustrates the cycles of the path. It means that the actions or stages were repeated by the respondents. The blue line shows purchase decision- making process of “loyalty loopers” - people, who are aware with the product or category and skip the evaluative alternatives stage.

Table 16 illustrates the division of the respondents into groups, based on their online purchase behavior. The visual combines all the respondents both for choosing a restaurant for evening task and choosing a drug for the stomachache in the online pharmacy store.

Table 16Description of the respondents

Table 16 illustrates that the type of products' category does not have a strong influence on the online purchase behavior of the respondents. It could be seen that the number of evaluators in both gastronomy and pharmacy category is pretty much the same. The main diversification could be only seen concerning such group of the respondents as Indifferents. The restaurants category contains 8% of the indifferents, however the pharmacy sphere only 2%. This could illustrate the trend among Internet users that the process of choosing pills is more carefully proceeded for theme comparing with the choosing a restaurants' task.

The visual Figure 2 l illustrates the online purchase decision-making path for both tasks: for choosing a restaurant and booking a table afterwards and choosing a pill for the stomachache and put the product into the online basket. Video recordings of the tasks which were conducted by the respondents illustrated the online behavior and the decision-making path. Firstly, the visual creation was started with the first task - to choose a restaurant for the evening and book a table. Then, one by one extra options were included in the Figure 2. It is essential to mention that the overall online purchase decision-making path of the respondents were not diverse enough concerning the category of the product. It seemed that the average combination of action does not literally depend on the sphere of purchase, as it is much more affected by the person which was chosen as a respondent.

The Figure 2 was created according to the McKinsey consumer behavior model. The whole visual is divided on decision-making steps: problem recognition, initial consideration set, evaluation of alternatives, moment of purchase and loyalty loop. As there was a special task for the respondents, their step in a frame of problem recognition was “received task”, then respondents moved to the next part of the decision-making process, which is initial consideration set. During this step respondents were choosing online platform for making an information search. The key platforms were: World Wide Web, Instagram and Interactive online maps like Yandex maps, Google maps and 2GIS. On this stage respondents were choosing their way to find necessary information concerning restaurants list or existing online drugstores and the types of medicine which would be helpful in case of stomachache. It is essential to mention, that for the restaurants tasks respondents used all three types of online platforms, however for the medicine task they used only World Wide Web and the interactive online maps. On the evaluation of alternatives stage such respondents as “evaluators” were analyzing the evaluation criteria, which are crustal for them, like other people feedbacks in order to rely on their experience, photos of the product: in case of restaurant choice respondents were evaluating the atmosphere, interior design and photos of the dishes. In a pharmaceutical case they were analyzing the packaging and the drugs. The next sub-step here was also “analyzing benefits”. In this part respondents were searching for extra bonuses like sales and promocodes on sales in order to affect the purchase and make it less expensive. If this path was not enough for the respondents, they moved to the “initial consideration set” stage again and repeat the cycle. The number of cycles were diversified from one to 5 depends on the respondents' preferences and their level of prior knowledge concerning category of the task and product itself. When the solution was found, respondents moved to the final stage “moment of purchase”. Weather it was just an experiment and the respondents were not needed to make a real purchase, the final step for them was choosing a restaurant and their readiness to book a table or putting a drug in a bucket. However, the online purchase decision-making path for respondents that were coded as a “loyalty loopers” is completely opposite. Is clearly seen from figure 2 that the whole decision-making path for this group for the respondents is much shorter. They also start their path in a stage of “problem recognition”, as they were received a special task - to choose a restaurant for the evening and medicine for stomachache. Then they move to the “initial consideration set” in order to choose the most convenient platform for them. The variety of online platforms for the loyalty loppers is also the same as for the evaluators as they chose it among World Wide Web, interactive online maps and Instagram. It is needed to remind that for the restaurants tasks respondents used all three types of online platforms, however for the medicine task they used only World Wide Web and the interactive online maps. As this group of the respondents is considered to be loyalty loopers which are people with extensive level of prior knowledge of product and category, they are already provided with the products on the exact market. It means that they do have their own preferences and previous experience, what makes their online purchase decision-making path less time consuming. When they made the final decision, they were falling into the loyalty loop stage. It means that they repeated the purchase that was made previous, previous experience with the product. Then they also moved to final step of the online consumer behavior - “moment of purchase”. Weather it was just an experiment and the respondents were not needed to make a real purchase, the final step for them was choosing a restaurant and their readiness to book a table or putting a drug in a bucket. The third group of the respondents were “indifferent”. They are people with the deviate online purchase decision-making path. This group of people were not included in Fid.1 as their amount was quite small and they made their decision without any criteria: they were not sensitive with the evaluation criteria or were choosing the first option that they found in a minimal period of time. That is why it was much more convenient and reliable to illustrate the online consumers' behavior basing only on two groups of the respondents.

The Figure 2 illustrates the average online purchase decision-making path for all the respondents. It is clearly seen from the visual that the stage that was combined both evaluators and loyalty loopers was the “initial consideration set”. On this stage both groups were involving in the online decision-making process. It means that if brands and companies want to level up their competitiveness and gain new audience or make their customers to stay with the bland and become loyalty loppers, it is better for them to concentrate their marketing strategies especially in this stage. This step could be characterized as a brand recognition stage and highlighting among competitors. In this stage online consumers spend their time on choosing a platform where to make a search of information which is needed for them to achieve a task.

When respondents were choosing World Wide Web as an online platform where to search for a product, they used Google or Yandex search for example in order to find a necessary web-site. During coding video files, it was clear that all the respondents were choosing higher links on the search page. It means that companies need to include additional expenses on SEO promotions if they are interested in enlarging the visits of the site page and increase reaches. Moreover, during restaurants' choosing, respondents were keen on such web-sites that ae specializing on collections and ratings of existing restaurants. Respondents explained their choice of these web-sites as it was much more convenient for them to use such web-sites, as they include information which combines all the criteria that they are interested in and the variety of option to choose from and compare with. That is why, it would be a great for companies to collaborate with such platforms in order to gain new customers and increase their chances on extra revenue. The next alternative for the respondents during choosing an online platform was such social network as Instagram. This platform is also convenient for the respondents as it combines fast information searching tools and the photos which are useful for the respondents in case of evaluating contents and option. The next convenient point in Instagram is an availability of other real consumers' feedbacks. User Generated Content (UGC) is a unique tool for the respondents to analyze company's performance as Instagram users share photos and post their feedbacks concerning exact product. Moreover, the function which allows users to comment posts enlarge number of feedbacks what helps to evaluate product itself based on different criteria which customers are more interested in. The next tool which customers are also interested in are the Instagram accounts with collections and ratings of special products in a particular category. It could be the same situation with the web-sites. Respondents prefer to use pages with a great variety of product alternatives, like Restorating, bloggers' accounts where influencers share their personal experience and perception with the product or category. The next useful Instagram tool for the respondents was saving a post in a necessary folder. They create collection of products what helps them to reduce time spending on information search. Consumers just need to open the folder with the collection of posts concerning particular product and then start their evaluating process. This experience illustrates brands and companies that they need to invest in their brand positioning and promoting in Instagram platform, to devote more attention and money to Social Media Marketing, if they are interested in competing in e-commerce market.

The final online platform which is included in respondents' initial consideration set is online interactive maps. Respondents prefer to use Yandex and Google maps and 2GIS also in case of searching primary information concerning product. They highlight this type of online platform due to its convenience as these maps combines companies based on their location. If respondents were interested in a restaurant in a special district or a drug store near their place of living, they opened these maps for information search. This tool helps consumers to highlight the set of companies that could be suitable for them and make a primary analysis of them. Such interactive maps provide their customers with the necessary information concerning brands like photos, web-site of the company, links to their pages in Social Networks and all necessary contacts like e-mail and phone number. If brands are interested to compete with other e-commerce market players in their segment, it could be useful for them to collaborate with these interactive maps in case of gaining their audience and raising brand's recognition.

If companies are interested in gaining new consumers and increase their level of the competitiveness among market players, it could be useful for them to stress their attention on the such customers' decision-making stage as “evaluation of alternatives”. On this step of p=online purchase behavior respondents paid more attention on the contents that describes the product or service. The more information concerning product consumer has, the more interested in and loyal customers are to the particular product or service. According to the Figure 2 it is better for companies to provide their target audience with the such extra product's information as high-quality photos of the product, detailed menu or the nutrition explanation and other consumers' feedbacks. These points push customers to rely on brand and even become loyal to if, they are interested and satisfied with the received data.

Moreover, consumers which are grouped as loyalty loopers tend to have more than 1 circle in their decision-making path. The average time for making an online purchase was from 2 to 6 minutes, that shows that it is better for brands to provide their customers with the necessary information as fast as they can. Thirdly, the step of searching for benefits is also essential part for such group of respondents as evaluators. According to the Figure.2 is the second sub-step in the evaluative of alternatives stage. If respondents are not satisfied with some conditions concerning particular product, they could search for some sales and promocodes. There is a tendency to repeat a circle after this step. It means that consumers are likely to make a purchase if there are some extra benefits for them which will enlarge their level of satisfaction and pushes to make a purchase of the product.

To sum everything up it shown that as loyalty loopers, as evaluators do have the specific online purchase behaviors that could be illustrated in a Figure 2. The result illustrates the variations in the behavior concerning groups. Such type of the respondents as “indifferent” was not included in the Figure 2, as the amount of the respondents which were described as this group is quite small and their online purchase decision-making actions are quite derivative. It is essential to remember that both evaluators and loyalty loopers do have a similar path up to the evaluative of alternatives stage. In the step of initial consideration set both groups for all tasks were choosing the most convenient online platforms that could be useful for them to make a purchase and to achieve a final step of the task.

On this stage loyalty loppers tend to find out their final option of the task, however such respondents as loyalty loopers could only start their path. After choosing a platform among World Wide Web, online interactive maps and Instagram, they switch to the stage of the evaluating of alternatives. It is essential to remind that Internet users implement search in such social network as Instagram only for the restaurant case. Loyalty loppers in both cases move than to their loyalty loop stage and end their task despite evaluating other alternatives, while the group of respondents as evaluators only begin their online purchase decision-making path.

It is also essential to mention that evaluators could repeat the circle of the decision-making path and use another platform for information search in the initial consideration set stage. It means that one consumer could be gained by the companies in several platforms, so it could increase the level of brands' recognition what could lead to the increased brand' loyalty of the target audience.

The research paper was directed to the identification the effect of prior knowledge of consumers on online purchase decision-making process. It essential to remind that in 2018, the number of Russian buyers making purchases on the Internet every month reached 69% and by the end of 2018, the volume of the online advertising market in Russia exceeded the indicators of the television advertising market and amounted to $ 3.1 billion. (PwC 2019) The results of the research were gained by the detailed evidence from a sample size of 50 respondents who are Internet users and online purchases makers. They were provided with two simple tasks.

Figure 2. Combined online purchase decision-making model for all individuals

They were asked to choose a restaurant for the evening and a pharmaceutical from a stomachache. As the research is considered to be about online purchase decision-making process the “moment of purchase” was stranded as booking a table in a restaurant a and putting goods in the online pharmacy basket. Firstly, each responded conducted two tasks and then was asked some questions in order to make a decision-making path more detailed and fully explained.

The analysis of variance revealed the influence of primary knowledge on the decision-making process for online purchases. The analysis was conducted in two independent samples two independent areas such as gastronomy and pharmacy. It was found that different levels of primary knowledge affect the decision-making process about online purchases in different areas. So, only one level of prior knowledge affected in the process of choosing a restaurant for a family dinner - primary knowledge of the field of gastronomy. In the process of choosing pills for abdominal pain, both levels of prior knowledge were affected, such as the level of prior knowledge about the product and the level of prior knowledge about the field. An additional influence on the process of choosing a restaurant for dinner was revealed. This variable turned out to be the time that the observation participant spent on completing the task. Variables such as the gadget used to select the product, the age of the participant, recommendations and reviews from other customers and relatives, and various discounts and promo codes did not affect the decision-making process for online purchases. In a further study of which group of people is strongly influenced by a particular factor, it was found that in the field of gastronomy, loyal customers spend less time on decision making process, but have a low level of knowledge about the industry, than people who actively evaluate alternatives or skip both steps in the decision-making process. In the field of pharmaceuticals, it has been found that people who make an active assessment of alternatives have a higher level of knowledge of the industry and the product than loyal consumers or indifferent ones. Thus, the hypothesis that knowledge of product has an effect on purchase decision-making process outcomes is confirmed in one industry; another hypothesis about the extensive level of the prior knowledge customer has, the less time he or she spends on the decision-making process is confirmed in one sphere too.

Based on the qualitative approach of data analysis, it was identified that the more extensive level of the prior knowledge consumer has, the less he or she is engaged in decision making processes with a smaller number of cycles. (Hb1) The Fig. 1 illustrates the combined online purchase decision-making model for all individuals. As it was identified that people with extensive level of prior knowledge are considered to be grouped as a “loyalty loopers”. This type of the respondents tends to have a product experience and a strong perception with the brands. Their online purchase decision-making path was much more linear compared with “evaluators”, as they skipped the stage “evaluation of alternatives”. This result support and proves the H1b.

However, Fig. 2 shows the online purchase decision-making behavior of “evaluators” too. This type of respondents could be characterized as people who are nor familiar enough with the particular product and have a longer path than “loyalty loopers”. As it could be seen from the visual, the online behavior of “evaluators” has an “evaluative of alternatives” stage. On this step respondents were analyzing the evaluation criteria, which are crustal for them, like other people feedbacks in order to rely on their experience, photos of the product: in case of restaurant choice respondents were evaluating the atmosphere, interior design and photos of the dishes. In a pharmaceutical case they were analyzing the packaging and the drugs. The next sub-step here was also “analyzing benefits”. In this part respondents were searching for extra bonuses like sales and promocodes on sales in order to affect the purchase and make it less expensive. If this path was not enough for the respondents, they moved to the “initial consideration set” stage again and repeat the cycle. This support the H1c: consumers with the limited level of prior knowledge engage in an online purchase decision-making process with a higher number of alternatives, than consumers with the extensive level of prior knowledge. It is essential to mention that the overall online purchase decision-making path of the respondents were not diverse enough concerning the category of the product. It seemed that the average combination of action does not literally depend on the sphere of purchase, as it is much more affected by the person which was chosen as a respondent.

4. Implementation

The key idea of the research is to conduct a deep analysis of consumers' online behavior, which undoubtedly could be useful for companies which are represented on e-commerce market. The results and findings of the research could be implemented in brands' marketing strategies, if they do specialize or have a division on e-commerce market. The results could help companies to analyze their positioning and promoting strategies in order to evaluate their competitiveness and full in the gaps in their ways of communications with potential customers. If brands and companies want to level up their competitiveness and gain new audience or make their customers to stay with the bland and become loyalty loppers, it is better for them to concentrate their marketing strategies especially in this stage. This step could be characterized as a brand recognition stage and highlighting among competitors. In this stage online consumers spend their time on choosing a platform where to make a search of information which is needed for them to achieve a task. It is also essential to mention that evaluators could repeat the circle of the decision-making path and use another platform for information search in the initial consideration set stage. It means that one consumer could be gained by the companies in several platforms what could increase the level of brands' recognition what could lead to the increased brand' loyalty of the target audience.

4.1 Limitations

However, the research paper has several limitations which could affect the results of future researches in an online consumer behavior field. In order to identify how does the level of prior knowledge affects the online purchase decision-making process, especially on every stage of the purchase behavior, it is better to conduct a kind of regression models. However, sample size of 50 respondents is not enough for such tool of quantitative analysis. The multinomial regression model does not provide the authors with the stable and significant results what could be explained as the limits of the sample. Secondly, the type of data collection could also have an influence in the received results. Data for this research paper was received by video files and then coded for the further analysis. Then some interviews and extra questions were asked for the respondents in order to specify their decision-making path. It could be a useful idea to create a questionnaire for more detailed results. This type of data collection could illustrate some other trends on online consumer behavior which were not obvious from the analysis of this research. The next limitation of the research paper is considered to be the following: the level of prior knowledge of the respondents is a latent variable, which does not have the exact scale of measuring. This means that the evaluation of this variable is not objective enough as there is opportunity for respondents to provide the authors with modified information. This could affect the results of the analysis.

4.2 Discussions

It could be a great idea for the further researchers to identify the average time that respondent spends on each stage on online decision-making process and what sub-steps of these actions are more crucial for them in case of online behavior. These results could be implemented by companies in order to adjust their positioning and promoting strategy, to make some changes in their content if is needed and start gaining more consumers. Secondly, it could be also interesting for the researchers to examine how do consumers perceive information which is provided by brands. This could also be useful for companies in case of modifying their content, based on their target audience preferences. Thirdly, it could be a great opportunity for the researchers to examine the behavior of loyalty loopers and found out what factors affected them in such a way that they moved from evaluator group to loyalty loopers.

Reference list

1. Langley, H. Mintzberg, P. Pitcher, E. Posada, J. Saint-Macary, Opening up decision making: the view from the black stool, Organization Science 6 (1995) 260-279.

2. Alba, Joseph and J. Wesley Hutchinson (1987), “Dimensions of Consumer Expertise,” Journal of Consumer Research, 13 (March), 411-454.

3. Armano D. (2007) “The Marketing Spiral”, Logic+Emotion

4. Belch G. & Belch M. (2009) Advertising and Promotion. An Integrated Marketing Communication Perspective. McGraw-Hill/Irwin, New York, p. 77

5. Boyd, H.W., Walker, O.C., Mullins, J. and Larre´che´, J-C. (2002), Marketing Management, A Strategic Decision-Making Approach, McGraw-Hill/Irwin, Columbus, OH.

6. C. Ranaweera, G. McDougall, H. Bansal, A model of online customer behavior during the initial transaction: moderating effects of customer characteristics, Marketing Theory 5 (2005) 51-74.

7. Cho, N. and Park, S. (2001), “Development of Electronic Commerce User-Consumer Satisfaction Index (ECUSI) for Internet Shopping”, Industrial Management & Data Systems, Vol. 101, No. 8, pp. 400 - 6.

8. Constantinides, E. (2002), “The 4S Web-Marketing Mix Model, E-Commerce Research and Applications", Elsevier Science, Vol. 1 No. 1, pp. 56 - 76.

9. Court D., Elzinga D., Mulder S. & Vetvik O.J. (2009) “The consumer decision journey”, McKinsey Quarterly.

10. Dumler Michael P. & Skinner Steven J (2007). Primer for management. (2nd ed.). South-Western College Pub

11. E. Cowley, A.A. Mitchell, The moderating effect of product knowledge on the learning and organization of product information, Journal of Consumer Research 30 (2003) 443- 454