Online purchase - Free Essay Example

Search and Purchase Correlation in German Online Market Dynamics and Trends Purchasing Drivers for Customers Abstract More and more people make online purchase; buy different kind of goods in online stores. Targeting on these people becomes opportunity as well as challenge for marketers. Understanding what are the main drivers of people preferring to buy products and services online rather than offline has become main area of research for many stakeholders. Researchers analyze the topic form various angles, they try to find main online purchasing drivers for customers, which make them prefer to go to online stores rather than real ones. Building a reliable model on that factors which will make predictable the future of online purchase dynamics, is one the purposes researchers aim to achieve. My thesis will be one more contribution in that area. It will be unique in the way that its analysis is based on most recent facts which cover 4 years dynamics and tends in that field on German market, starting from 3rd quarter of 2005 until 2nd quarter of 2009. Looking at facts through this time frame should give us opportunity to observe if there is any recognizable change in â€Å"customer behavior†: if the trust towards online purchase is increasing which is reflected on the increased amount of sold products in online stores, and which factors or purchasing drivers can be applicable for explaining these trends. To be more specific, I will divide my thesis in several important parts: first, I will try to find out if there is any connections between the amount people spend with the internet stores and the amount of products the buy, or in other words, if there is any evidence of correlation between the â€Å"product search† and â€Å"product buy† amounts. This will be done by running regression analysis based on internet traffic data provided on www.AGOF.de web site (in more details it will be discussed in the INTRODUCTION part). Second, after tracking the statistically significant relationships between these two factors: â€Å"search† and â€Å"buy†, where second factor depends on the first one, I will try to dived this relationship between â€Å"explainable with search amount† and â€Å"explainable with other factors† parts. Then I will try to cover already famous literature, which is explaining the motivators of people searching goods online. Finally, we will talk abo ut other factors, which might explain online purchase behavior but they are not prone to the underline logic of our model, or just simply are beyond or the factor â€Å"search† as an explanatory of â€Å"buy† rate. As a conclusion I will summarize my findings in the topic. 1. Introduction Developing of technologies and internet has born for people new opportunities changing our behavior and our attitude to many things. Without any exaggeration it has dramatically changed our lives. In many activities Internet has given us opportunity to save our efforts, time and sometimes even money. One of the areas internet has made dramatic affect is shopping. Online shopping becomes more and more important part in our everyday life. Referring to the Chen Chih-Chung and Su-Chao Chang common research in 2005, â€Å"if in 2004 online shopping was only 7% of retail sales in USA, in 2010 it is expected to reach 13%†. Marketers are paying more and more attention to the opportunities lying in online world trying to analyze and predict the trends and movements there, in order not to miss their chance, while in the fierce and almost saturated competition in offline shopping, online market is almost only way of fast growth. In order to succeed in online shopping one of the m ost important aspects for marketers is to understand their potential clients _ what makes them buy goods in virtual stores rather than in real ones, what are the key factors and drivers which initiate them to conduct online purchase. The final goal of the thesis cant be to find the only model with only factors which describe pretty exactly the determinants and intensions of customers. Many researchers try to explain the phenomenon such as online purchase intension and I will one of the also. Hope to make my little contribution in that field. In this Thesis I apply some academic researches, online traffic data analysis available on AGOF, bring some insights from ESMT practice project conducted with my per classmates in Axel Springer consulting team and also will refer to one questionnaire conducted for the same practice project. Based on these information and my personal observations I will try make some more insights not only in above mentioned online purchase determinants for customers, what makes them to pay money online, but also I will try to analyze overall trend and dynamics in online shopping during last four-five years. I will try to make some insights which products were sold mostly through internet several years ago and which are sold now, through which web sites. Which Internet-content-areas were most employed years ago and which are emerging now. Finally, I will try to build a model explaining the intension of purchase through different factors. Even more challenging is converting offline shoppers into online shoppers, but before achieving that, marketers in the industry have to understand basic drivers _ what make people buy different things on-line instead of going to supermarket, understanding whether it is general trend that due to the technologies development people change their purchase habits and more and more people transform form off-line into on-line shoppers and hence the future of e-markets is sunny with plenty of room and e ver growing possibilities, or there is something special which catches peoples attention, which affects their decision making process in particular situation and makes them preferring buying things in virtual world. Something which makes them â€Å"take more risk† and without having possibility to touch the product and checking its quality pay money in it and at the same time be sure that his decision is right. My thesis will be one more contribution in understanding above mentioned online-purchase drivers. Besides referring to already published academic researches and newspaper/magazine articles, I will try to analyze internet traffic data, to look at the developments and trends over last four-five years, if there is any content-wise change what people were buying several years ago and what they are buying now. This observation should give us additional information if there is substantial difference in amount of purchased quantity of different products, if it so, we have to analyze what induces these differences, why are definite kind of products sold through internet more compared to others. Finally, defining key online purchasing derivers, I will try to classify them with respect of relevance and importance, try to understand their correlation and final affect on purchasing decision. 2. Literature Review Literature, different researches and publications, were basically used for explaining the driving factors of people who search and buy products online. By analyzing various sources I tried to come up with common understanding of drivers which motivate people in conducting online purchase. Besides academic and non-academic researches, some insights from area experts were also taken in consideration. These experts were interviewed during consulting practice project, which I and my three other peer class mates conducted for one of the European leading media companies, about the topic â€Å"how to monetize online content†. Kevin Wises and Hyo Jung Kims publication in Cyber Psychology Behavior in November 2008, named â€Å"Searching versus Surfing: How different ways of acquiring content online affect cognitive processing† elaborates about distinctions between different methods of search people apply online. Authors claim that there are two different types of search: s earching and surfing. In first case people are aware of their will. They look for something objectively, whether in another case, they just wander in internet, without aiming on finding something specific. Testing differences between these two models with different type of examples and quantitative models and assessment tools, gave evidence that, depending on that whether people surf or search internet, differences not only final content of found information and then they way how these people remember details about the searched product, but differences also the biological processes which take place in these people while looking for information. Depending on the model used, they hade different level of arousal and different pace of hart beating. Authors conclude that, â€Å"being there† and â€Å"getting there† makes big affect on the final outcome of searched data. M. Bosnjiak, M. Galesic and T. Tuten in their â€Å"Personality determinants of online shopping: Expl aining online purchase intensions using a hierarchical approach† published in Science Direct in 2007 make focus on personality determinants of online shopping behavior and the study is based on the on the work of Mowen about hierarchical approach to personality (Mowen, 3M Model, 2000 will be described below). Besides Mowen, authors widely apply to some other researchers and academics works. It could be perceived as almost perfect summarization of most valuable researches conducted on this topic by other researchers. One of the theories they apply is Pachuris 2002 work, where he tries to explain approach to online shopping behavior through four categories: 1) economics of information approach, 2) cognitive costs approach, 3) lifestyle approach, and 4) contextual influence. 1) Economics of information approach elaborates on perceived efficiency of buying online. It focuses on consumers preferences on shopping channels taking in consideration and comparing different alternativ e subjective costs including time costs. Perceived cost of shopping online should be less then shopping offline, if that is not the case, then consumers will not apply to inline purchase. 2) Cognitive costs approach elaborates on search-related cognitive processes and claim that consumers take decisions based on price and quality of the products, as well as reliability and credibility of online stores. Their final aim is to minimize cognitive costs associated with the actual purchase. 3) Lifestyle approach focuses on the type of life consumers lead, and their socio-demographic habits of money spending, interests, motives, values and opinions. Here they quote Alerck and Settle (2002) investigate regarding the time saving as a motivator of online shopping, where they conclude that surprisingly online consumers do not shop because Internet is perceived as the time-saving factor during shopping. 4) Contextual influence approach makes focus on navigational aides as well as atmos phere. Quoting Senecal et al. studies (2005) they repeat that recommendations in buying process to the customers who are exposed to the product make decision more complex but in fact it does not change final outcome. The work claims that despite all four points give us some insights and knowledge about online purchase motivators, it does not exactly examine the traits affecting the willingness of shopping online and still makes hint on the psychological portrait of every individual, and it seems that online and offline shoppers psycho-portrait differs from each other (Donthu and Garcia, 1999). Online shoppers seem to be more innovative people who are willing to take more risk and experience new feelings rather than non-Internet-shoppers. Mowens 3M Model of Motivation elaborates more on personality, takes I consideration individual differences and tries to draw the portrait of online shoppers. It applies four hierarchical levels: surface traits, situational traits, compound tra its and elemental traits. All these are hierarchically dependent on each other with the order they are presented. Situational traits (affective involvement, cognitive involvement) affect surface traits (intention to shop online), then come compound traits which often are developed through socialization (need for cognition, need to evaluate, need for arousal, need for material resources), but they are affected not only from external factors but also with personal experiences and socialization history with the traits at the highest level of model (elemental traits neuroticism, conscientiousness, extraversion, openness, agreeableness). â€Å"Effects of consumer Trust and Risk on online purchase decision making: A comparison of Finnish and United States students† (authors: C. Comegys, M. Hannula, J. Và ¤isà ¤nen) published in August 2009 by International Journal of Management, as we understand from label, makes focus on consumers trust and risk perceiveness during online pu rchase. They go through Kotlers five stage Consumer Buying Decision Process (need recognition, information search, evaluation, purchase decision, post-purchase behavior) and investigate how risk and trust theories work together through these five stages. Based on hypothesis testing method they also examine such factors as gender, changes over time and purchase volume. Finally they suggest managerial implications. Chen Chih-Chung and Su-Chao Changs â€Å"Discussion on the behavior intention model of consumer online shopping† tests and extends the Theory of Planned Behavior (TPB) models by examining variables of past experience and purchasing channels knowledge, in predicting consumer shopping. TPB elaborates that not all human decisions are explained and motivated by personal will, but rather they are somewhere between completely and incompletely controlled by personal will. In this research, authors exercise different factors, such as attitude, subject norm, perceived beha vior control, behavior intention, past shopping experience and consumer channel knowledge through multi-regression analysis model with level of significance p0,01. Jonna Jà ¤rvelà ¤inens â€Å"Online purchase intensions: An empirical testing of a multiple-theory model† (2007, Journal of Organizational Computing) concentrates on understanding how consumers choose their purchasing channel when the environment is relatively stable. The factors tested are: preferring conversation with customer service, complexity of product, prior online shopping experience, social influence or perception of system usefulness or ease of use. This model also offers to test some hypothesis and then builds correlation matrix to identify extent of final influence of each factor. Finally it summarizes theoretical and practical implications. â€Å"A model of the determinants of purchasing from virtual stores† by R. Bakhri, F. Belanger and J. Hiks (2008, Journal of Organizational Computin g and Electronic Commerce) describes the consumer purchase decisions in a virtual stores. It makes focus on how individual visiting online stores could be affected to purchase from such a stores. Factors such as perceived usefulness, perceived behavioral control, perceived peer influence are tested through several hypothesis built in a quantitative model Chi-square analysis. The research discusses the design of stores as a tool leading to a purchase decision. As a conclusion I would say that literature contributed me not only in understanding intensions and drivers of online customers which make people start search for products objectively, but also to understand what are some other factors which are not among the factors motivating search, but are in the â€Å"un-explainable by search† part of our regression analysis and affect final decision of purchase. 3. Identifying correlation between the quantities of searched and purchased products. Analyzing AGOF data 3.1. What is AGOF? Founded in 2002 by German online marketers and online advertising media, AGOFs area of interest include activities such as: advertising market, meaning media agencies and online media planners, marketing decision makers at advertisers, marketers as well as other market partners. It has 17 members and 60 licensees and its studies are audited by independent third parties. AGOF collects and analyzes data about customers usage of internet, analyzes it through different criteria and segments (differentiated by sex, age, education, working status) and thus makes internet advertising more transparent and predictable. AGOFs approach is to sort and analyze everything through â€Å"unique user†, and the methodology applied is multi-method model, which means: collecting data about unique users (unique computer) with three different ways. Main source is the electronic measurement (data collection) is a basic, and the other two, telephone interviews and on-site surveys are complementa ry for fulfilling the data. Individual clients account measures more technical inputs such as page impressions, visits, time (when), content related topics (content on IVW level), whereas online survey collects the information about clients themselves and their computer usage habits: socio-demographic data (age, gender, education†¦), place of use, information on computers (number of users, login, etc). AGOF renews its information on the quarter basis, publishing the internet facts of previous three months, tries to catch with latest dynamics to make internet more predictable and plannable for marketers. Also, it constantly tries to further develop internet facts study. Analyzing relationship between â€Å"search† and â€Å"buyâ€Å". Conversion rate Before I move to the main part of the thesis, and use the â€Å"search† data as the main explanatory of â€Å"buy†, I will try to conduct multi-step analysis, first find out if there is any statistically significant relationship among these two factors. For this purpose I will test two hypotheses: 1. H0: B = 0, there is no statistically significant relationship between these two factors 2. H1: B 0, there is statistically significant relationship between two factors We will test these hypotheses on p0,01 significance level. Then if there is any correlation between â€Å"search† and â€Å"buy† we will interpret this confirmed correlation through the time-line across all products. Finally we will try to find main explanatory drivers and motivators of online purchase, academics and different publications talk about. Before testing the hypothesis lets review what it means for two variables being correlated to each other: to claim on correlation be tween X independent and Y dependent variables, the following should occur: * X proceeds Y * Y must not occur if X does not * Y must occur whenever X occurs As we know, the last point is not always the case, and very often people search products but do not buy them. One of the explanations of this â€Å"phenomenon† is nicely described in Kevin Wises and Hyo Jung Kims research named â€Å"Searching versus Surfing†, published in 2008, with the purpose to find out whether the content of acquired online media changes or not depending on the way how it acquired. The â€Å"surfing† part describes situation when people just â€Å"wander in the internet† without aim to buy anything. For us it is important to understand what makes these wanderers not aiming to purchase, suddenly change their mind and conduct online shopping. More important for us is to understand the purchasing drivers of those people who go shopping online consciously and predetermined . Or as this article explains, we are going to describe different motivators of â€Å"being there† and â€Å"getting there†. For the first step of our analysis, we are taking 26 different products from AGOF database (list of products in Appendix 2), for which we have different observations for every quarter, and then we run one linear regression for 4 year time observations. In total we will have 416 (26 products multiplied on 16 quarters) observations in the regression analysis (Appendix 4). As the results of regression show on appendix, R Squared is 0,8422 which means that about 84% of buy data can are explainable with â€Å"search†, and if we explain what are the motivators for people to search or surf on internet, then we will make one more step towards making internet purchase process more explainable and predictable. Other 14% of data can be explained with some other factors rather than search, and in later stage of the thesis we will have a talk on the se factors either. Besides R Squared, there are some more important coefficients in our results of linear regression, which are worth to pay attention. P-value as well as F-value in ANOVA table, is almost zero with 168 level, which tells us that from two hypothesis previously developed by us, we can reject first H0 hypothesis with p0,01 significance (1% significance level) and claim that there is statistically significant relationships between search and buy data. Some other coefficients which are worth to interpret are slope and the intercept: * Y intercept B0, in our case 1,13 says how many products are bought when â€Å"search† rate equals to zero. Once we have not any observations on zero â€Å"search† level, then its interpretation will be dubious practical value. * Slope B1, 0,53 in our regression, can be understood as change in half-year â€Å"buy† quantity, when â€Å"search† is increased by one degree. On the second stage we should inte rpret this confirmed correlation through the timeline, across different quarters across all 26 products participating to understand if the conversion ratio of search into buy changed over time. Because it is suggested to have more than 26 observations for running more reliable regression analysis, we will unite two quarter data under one group/label â€Å"half year†, thus doubling our sample of observations and running regression on 52 observed products. In 4 year time period we will 8 different half-years and thus 8 more new regressions should be run. After running eight different regressions, to facilitate overview of results I gathered major coefficients in one table and then built some graphs to indicate trends better for R Squared, Intercept, Slope and P-value (Appendices 6 and 7) As table on Appendix 6 and the slight but constant increase over time of R Squared show (P-value decreases over time), correlation between search and buy rates increase, which means, more and more actual purchases become explainable with search rate, so more and more actual purchases become prone to logical explanation by different purchasing motivators researches have investigated. Other 14% which on its side is also decreasing over time, which means less and less purchase are non-explainable by search factor and its motivators. 4. Defining purchasing drivers based on academics researches, publications and expert insights As we already saw in literature review part, different academics have different approaches toward this problem; they all try to find the most appropriate criteria which will be most accurate measure of on-line purchase willingness. They try to test their theories and models form various angles, starting from personal psychological characteristics and portrait of people, continued with previous purchase experience, poor IT skills, affect of different external factors such as socio-demographic or the shopping channels used during process. Lookin g at the models of different publications and researches we can structure general criteria and factors which might be main motivators for conducting online purchase. Despite that these works never are the same they still have quiet much in common. We can categorize these factors as those, which might be logical motivators for searching products in internet and those which affect final â€Å"buy† quantity but are not explanatory factors for search motivation. As we already ran regressions on AGOF data for half-year time periods to look at the changes through 4 years time frame and we have clearly visible relationships between search and buy factors, where search as an independent variable mostly explains buy rate outcome with even increasing tendency, we can now try to correlate quantitative outcome with theory and try to interpret figures in purchasing drivers. But, lets first discuss separately what each article and publications tell us and then let us try to unify the m under one structure. In â€Å"trust and risk† theory already discussed in literature review, authors categorize eight different risk factors associated with internet purchase; these factors play important roles in every stage of the buying process. 1) performance risk related to product performance 2) time risk involved in transaction timed needed for shopping 3) financial risk absolute cost may be higher than in conventional shopping 4) delivery risk- the product may not arrive or may arrive late 5) social risk how friends and family react 6) privacy risk whether personal information is secure or not 7) payment risk financial consequences from sharing the credit card number 8) source risk reliability of the information on web-site Authors conclude that privacy, source, performance and payment risks were considered as the most crucial of risk in electronic business. Some of the hypothesis developed in the research found justification and also some other findings were: a) There is a positive relationship between the amount of consumer online trust and the online purchase volume b) There is a positive relationship between the amount of consumer online trust and the possible increase in purchase volume. There was evidence that people with little trust towards e-vendors significantly reduced purchased volume of products over time. c) Amount of risk-taking does not have any influence on purchased volume d) Risk and trust factors were perceived independently from five stage purchase process during buying process itself. â€Å"Behaviour intention model† which came up with following factors as purchasing drivers: attitude, subject norm, perceived behavior control, behavior intention, past shopping experience and consumer channel knowledge, finally concluded: a) Consumers attitude will affect their shopping behavior b) Consumers subject norm to online shopping will affect his/ser purchase beha vior c) Consumers perceived behavioral control to online shopping will affect his/her purchase behavior d) Consumers past online shopping experience will affect his/her purchase behavior e) Consumers internet channel knowledge will affect his/her shopping behavior f) The perceived behavioral control is not significant Further more, research agrees that consumers increased channel knowledge will grow in increased shopping intention. In fact the same concludes another, â€Å"online purchase intension model†, or â€Å"multiple-theory model†, which studies the affect of online channel in a relatively secure environment. This theory comes up with some interesting findings, but before quoting those findings, I will describe the theory approach itself: In the model, author looked to the case of purchasing behavior through eight different factors: PU, PEOU (Appendix 1), conversation preference, prior online experience, previous online booking experience, prod uct complexity, task ambiguity and social environments attitude. Covariance table was built for checking the relationship of each factor towards the final outcome. As covariance table showed, none of above mentioned criteria was significant determinant of result on its own; however some combinations like PU-PEOU might have become not big but statistically significant. Also, the model made evidence that: a) Prior online purchase experience has a significant positive effect on behavioural intention b) With increased experience a customer begins to trust online channel more c) The more customers are willing to repurchase form any given site, the more tend to shop on-line in general positive relationships with repurchase intension and purchase volume d) Trusting customers tend to be more loyal than who are risk-neutral. Loyal customers are typically satisfied with the service/products they receive â€Å"Experience with traditional and online channels had a conside rable impact on channel choice, as well as PU. The effect of social environment was smaller but statistically significant. Inexperienced customers who do not find online shopping quicker, cheaper, suitable, for shopping, or easier to find information are likely to abort the online shopping process and finish it offline.† Experience had such a big impact that special efforts should be done by companies to create first pleasant shopping experience. Some other important findings were: a) Some people, who do not want to shop elsewhere than in a store, do not perceive any convenience in for example online shopping. This again comes to the individualistic approach of different people. b) The social environment has a significant positive effect on behavioural intention and PU of the system Theory claims that the main reasons to purchase online could be: convenience, broader selection, competitive pricing, product choice, product value, greater access to information Do IT skills matter? What really matters rather than IT skills is trust towards the online shopping, which is something which comes with experience of internet usage. As conducted research among Fin and American students show, any difference in trust factors among consumers lay beyond the IT skills. It something deeper, which in most cases is inexpiable because perceived risk is very individual and comes from personal characteristics * Described in of Jonna Jervelà ¤inens Multiple-theory Model, â€Å"Technology acceptance model (TAM), which is based on two basic concepts, perceived usefulness (PU) and perceived ease of use (PEOU) tries to somehow answer that question. Experience with traditional and online channels rather than pure IT skills, had a considerable impact on channel choice, as well as PU. The effect of social environment was smaller but statistically significant† General insights in academics researches Lets sum up also some more general knowledge also, different research ers and stakeholders got about the internet: Here are some important findings: * Those who have less risk perception associated to the internet purchase buy more but at the same time we should consider that risk perception reduces due to the usage the internet over time. More experience customer has more he starts to trust the online channels. Hence heavy internet users buy online more than those who use internet less. Women and men perceive risks differently no matter the level of their technical or any other skills and no matter the environment * First online purchase experience has a significant influence on customers behaviour in the future. If a consumer had a negative experience during first purchase process, e.g. during booking tickets online, it less reliable that he will come back and try some other time. This explains why companies try to make positive first experience in their consumers buying behaviour and very often with different approaches try to incentivise them to conduct the purchase * Having positive prior experience increases the tru st of customers towards channels. There is always less possibility that trusting, loyal customer will be unsatisfied with the product/service he receives * Experience is so important that it is recommended to be taken in consideration while making the design of site. Even more as researches show, very often people take purchase decisions based on emotions rather than reasonable judgment and actual need of the product * There is positive correlation between total volume purchased by a customer and his repurchase intention rate from any other site. * Increased security perception in terms of provided personal information safety does not increase significantly consumers willingness to buy more. At the same time risk perception for customers is on peak when they enter their personal information. Thus we can conclude that companies should not make barriers to the customers during purchase procedure * One of the main factors which can make believe people to by products online is having possibility of returning them back, in case they do not like it will not meet their expectations after receiving it * Overall costs perceived during online purchase should less than during offline parches. If is more or the same, than consumer will not by online. In other words it should be more â€Å"convenient† for client to prefer buying online Interpreting increasing dynamics of regression analysis coefficients As we saw in previous topics reviewing academic and non-academic researches and their findings regarding purchasing motivators, purchasing channel preferences and etc. we can conclude that the trend we came up in our regression analysis (Appendices 6 and 7) is very logical reflection of what researches say. Increased R Squared value means that hypothesis mentioned by us in previous topics as the explanatory factors gain more and more credibility. They become more prone to the logic, which tell us that, over time, people get more experience of using Internet, thus their fear and risk averseness towards online purchase decreases: two main factors benefit two that: shifting the proportion of generation distribution, thus more and more old people become skilled in internet usage (those who were young several years ago) and another is socio-demographic affect or in other words peers or friends influence: once more people buy start trusting online purchase at least because of gained experience in the area, more people they will recommend. So, the conversion ratio of search to buy data will increase, as well as absolute number of online shoppers. Some other factors explaining this phenomenon might be the marketers themselves who are now more aware about online customers and their behaviour rather than several years ago, when industry was new and emerging. Many marketers possibly also realized that sometimes it might be nice, attracting design and correctly delivered massage, rather than customers actual need which will make them buy products online. Marketers started â€Å"attacking† the customers by asking with different flexible approaches to buy for some products and services online, and in many cases they become successful, once as we mentioned they are now more educated, and customers also become more trustful towards internet shopping. Factors not explaining the search amount As Mowens 3M Models one of the surprising findings is that sometimes important determinant which makes people to purchase goods online, is affective rather than cognitive involvement. In other words, decision sometimes is made purely on emotion merit rather than reasoning. This might be partial explanatory factor of those 13% of error term we got in our regression analysis last â€Å"half-year† coefficients. The other explanatory factor might be e.g. easing of payment for the good online. Because mostly this is the factor that customer is not aware the specifics about from the very beginning of purchase decision and when comes at the stage of real purchase, way of payment (easiness, clarity, quantity of information required) might become either motivating factor or discouraging one. 5. Conclusion As dynamics show slowly but objectively more people become online buyers, and also more consciously they become regarding online purchase (which is reflected in regression analysis coefficients) meaning that they need less and less page impressions (search clicks) to conduct actual purchase. I think this trend will continue because many facts already mentioned by us, contributes to this dynamic. As we mentioned, more experience with online has positive relationship with increased purchased volume. Another factor is the successful prior experience which also positively contributes towards increased purchased volume and finally the affect of your social network. Marketers should take in consideration some important factors which affect either negatively or positively the buying decision: * Sometimes people make online purchase on emotional basis rather than cognitive, so attractiveness of massage and design of site has big importance * Ease of payment system is another facto r contributing to increased amount of online sales * Increased security measures regarding filling the personal information, in not transformed in increased perception of safety. Even more, risk perception from customer is on the peak while entering online his name or his credit card number. Thus marketers should target on minimizing risk perception for the customers and hence boosting sales * Once successful prior experience with online purchase has invaluable impact on customers further behavior, marketers should aim to create first pleasant experiences of online purchase for their customers * Customers should have opportunity to return product back, is they do not like it after receiving As we managed to show in our analysis, there is a strong correlation between search and buy amounts in internet. Marketers should target not only â€Å"buy† traffic records, but also, â€Å"search†, because as we mentioned customers need time to become more loyal towards online shopping, time and experience converts them more trustful. Besides, sometimes purely emotional arousal might become the basis of consumers decision, thus give them opportunity to sea a nicely designed massage, might convert them in from neutral to active customers. References M. Bosnjak, M. Galesic, T.Tuten (2006) Personality Determinants of Online Shopping: Explaining Purchase Intentions Using a Hierarchical Approach ScienceDirect (www.sciencedirect.com), Journal of Business Research 60 (2007) 597 605 Jonna Jà ¤rvelà ¤inen, Online Purchase Intensions: An Empirical Testing of a Multiple-Theory Model Journal of Organizational Computing and Electronic Commerce 17(1), 53 74 (2007) Chen Chih-Chung, Su-Chao Cahng (2005) Discussions on the Begaviour Intension Model of Consumer Online Shopping Journal of Business and Management vol. 11, (2005) Reza Barkhi, France Belanger, and James Hicks, A Model of the Determinants of Purchasing from Virtual Stores Journal of Organizational Computing and Electronic Commerce, 18: 177 196, (2008) Copyright  © Taylor Francis Group, LLC; ISSN: 1091-9392 print/ 1532-7744 online; DOI: 10.1080/10919390802198840 Charles Comegys, Mika Hannula, Jaani Và ¤isà ¤nen, Effects of Consumer Trust and Risk on Online Purchase Decision-making: A comparison of Finish and United States Students Vol. 26 No. 2; (August 2009) Kevin Wise and Hyo Jung Kim, Searching Versus Surfing: How Different Ways of Acquiring Content Online Affect Cognitive Processing, volume 11, Number 3, (2008) Traffic data analysis Power Point presentation from internet web-site: www.agof.de graphiken if 2009 ii Traffic data from internet web-site www.agof.de Appendix 1: Terminology Explanation of some words or phrasal words, which in everyday life could be understood with many different meanings, but in my thesis they bear one and only one underlined meaning: AGOF German internet web-site tracking internet traffic data based on multi-method model AGOF clients German audience more than 14 years old using internet Conversion ratio the ratio of searched and purchased products in internet, indicating the rate by which â€Å"search† are converted in â€Å"buy† Customer behavior customers attitude towards online purchase on average, their intention to search and then by products online. The pace they search and buy products at a definite period of time Changing customer behavior any changes in amount of searched and purchased products in particular period of time compared to other period IVW German internet web-site tracking internet traffic data Unique user individual user in internet, individual computer Technology acc eptance model (TAM) consists from two major concepts: the perceived usefulness (PU) and ease of use (PEOU) of a system. PU users subjective perception of the ability of a computer to increase job performance when completing a task PEOU persons subjective perception of the effortlessness of a computer system Appendix 2: List of products on which analysis are conducted Alcoholic bavareges other alcoholic beverages and spirits Books computer and videogames computer hardware and accesoories computer software without games Tickets for movies, theatre, classical concerts, pop concerts or sporting events Movies on DVDs, Videos cheap air tickets from airlines Pay to music or movies from the internet as a download Investments, shares, securities, funds Health products and medicines DIY or DIY tools Hotels for leisure and business trips Charges Lottery Games Life and annuity insurance as private pension rental cars Music CDs Furniture or objects on the furniture Perfumes, fragrances for men or women shoes Sporting goods, sports equipment Sweets and salty snacks Telecommunications products, e.g. Cell phones or cordless phones Frozen foods and ready meals Holiday travel and last-minute travel Appendix 3: Number of Internet users in Germany in the second qurter of 2009 Source: www.AGOF.de, graphiken-if-2009-ii.download Population in Germany (people more than 14 years old) _ 64.82 million People who use Internet _ 43.20 million Internet users observed for surveys (people who used internet last three months) _ 42.22 million Appendix 4: Regression analyses for finding out relationship between â€Å"search† and â€Å"buy† data Appendix 5: Uniting two quarter data under one group for the sake of regression analysis Appendix 6: Important coefficients from eight different period regression analyses Columns: SS explainable Regression sum of square. Portion of â€Å"buy† explainable by â€Å"search† SS Residual Residual sum of square. Non- explainable part of â€Å"buy† factor with â€Å"search† SS Total Total sum of square R Squared Portion of â€Å"buy† explainable by â€Å"search†, as portion form 1 Non Explainable Non- explainable part of â€Å"buy† factor with â€Å"search†, as portion from 1 Appendix 7: Graphs representing trends in regression analysis coefficients over 4 years time period R Squared increases over time, which means, more and more „buyâ€Å" data are explainable with „searchâ€Å" data, thus the purchase motivating factors described by academics become more and more prone to the logic. It also is a matter of experience, social affect, improved IT skills, and trust towards channels as well as online purchase as the whole P-value approaching to zero means that hypothesis about â€Å"search† and â€Å"buyâ€Å" data correlation gain more and more viability