The Internet has tremendously changed the way in which businesses and consumers interact with each other. Consumers are becoming increasingly demanding, and naturally expect products and services to be suited to their individual needs. Derived needs change at an even faster pace nowadays and therefore firms recently discovered the Internet as a means of interactive marketing. For consumers the Internet creates new opportunities for buying at higher speed, discount prices and greater overall convenience.
Probably one of the most important developments in the previous years is e-commerce. Interactive home shopping is considered to be one of the fastest growing markets in the future, given the substantial benefits to both firms and consumers. However, the failure of many business-to-consumer start-ups shows, that recent innovations have been company-driven. The new consumer has yet to be educated to use the Internet as a shopping device. Especially sensitive issues such as data security as well as payment security seem to contribute to the reluctance of a great majority of potential Internet users to purchase via the Internet.
Despite its recognized importance, knowledge of the critical success factors in e-commerce is lacking. For once, the mind-set and expectations of the consumer have to be taken into account. Secondly, companies have to redesign their business processes in order to deliver superior customer value via the Internet. However, knowledge of consumers’ Internet buying behavior is limited. This paper will provide a deeper insight into the factors influencing the decision process.
We, a group of five students from Maastricht University, found that the problem has not been adequately addressed in the existing literature. For this reason, we conducted a market research study investigating Internet buying behavior. Our objective is to identify the factors and variables critical to a successful e-commerce strategy for business-to-consumer companies. This paper will focus on products from seven frequently quoted categories: books, audiovisual media, software, electronic equipment, travel services, clothing and groceries. We assume a consumer perspective in order to advise firms making market-driven decisions.
Although forecasters expect e-commerce to be one of the fastest growing markets in the near future we feel that firms can only gain sustainable competitive advantage if they can identify and control the critical determinants of success.
1. Introduction
The Internet has tremendously changed the way in which businesses and consumers interact with each other. Consumers are becoming increasingly demanding, and naturally expect products and services to be suited to their individual needs. Derived needs change at an even faster pace nowadays and therefore firms recently discovered the Internet as a means of interactive marketing. For consumers the Internet creates new opportunities for buying at higher speed, discount prices and greater overall convenience.
Probably one of the most important developments in the previous years is e-commerce. Interactive home shopping is considered to be one of the fastest growing markets in the future, given the substantial benefits to both firms and consumers. However, the failure of many business-to-consumer start-ups shows[1], that recent innovations have been company-driven. The new consumer has yet to be educated to use the Internet as a shopping device. Especially sensitive issues such as data security as well as payment security seem to contribute to the reluctance of a great majority of potential Internet users to purchase via the Internet.
Despite its recognized importance, knowledge of the critical success factors in e-commerce is lacking. For once, the mind-set and expectations of the consumer have to be taken into account. Secondly, companies have to redesign their business processes in order to deliver superior customer value via the Internet. However, knowledge of consumers’ Internet buying behavior is limited. This paper will provide a deeper insight into the factors influencing the decision process.
We, a group of five students from Maastricht University, found that the problem has not been adequately addressed in the existing literature. For this reason, we conducted a market research study investigating Internet buying behavior. Our objective is to identify the factors and variables critical to a successful e-commerce strategy for business-to-consumer companies. This paper will focus on products from seven frequently quoted categories: books, audiovisual media, software, electronic equipment, travel services, clothing and groceries. We assume a consumer perspective in order to advise firms making market-driven decisions.
Although forecasters[2] expect e-commerce to be one of the fastest growing markets in the near future we feel that firms can only gain sustainable competitive advantage if they can identify and control the critical determinants of success. We hope to close the existing information gap and aid firms embarking on e-commerce in future decision making.
2. Research Process
Problem Formulation
The purpose of this study is to approach the research topic answering the following problem statement:
“Which products are most likely to be bought via the Internet and why?”
We subdivided the general problem statement into separate research questions guiding our hypothesis formulation.
- “Which product categories are likely to be bought via the Internet?”
- “How much weight do consumers put on the distinct features of e-commerce?”
- “Which product characteristics determine where a product is bought (Internet vs. real marketplace)?”
- “Do the factors measured in the abovementioned problem statements display differences with regard to demographic variables?”
Determining the research design
In the following paragraph we will elaborate on the steps taken to answer the questions stated above. We describe how we used exploratory research to get a general idea of the subject matter, make an inventory of the facts already known, generate hypotheses and find related questions to be asked.
“…exploratory research will be used to find the most likely explanations that will then be tested empirically.”[3]
The group started out with a brainstorming session and an in-depth focus group discussion on the subject. In addition, we interviewed knowledgeable persons on the topic and consulted websites of market research agencies and consultancies specialized in e-commerce (e.g. IDC, Gardner Group, Forrester Research, ProActive International). Additional secondary data came from academic literature, though limited in scope due to the fact that little prior research has been conducted in this field. Personal briefings, focus group discussions and the Internet search yielded the most valuable information.
The findings guided us in defining the product categories, distinct features and critical product attributes we investigated in our research. For a comprehensive overview, see the table below.
Table 1 – Results of Exploratory Research
illustration not visible in this excerpt
Hypotheses
After distinguishing these three categories, we came up with the following groups of hypotheses:
- Hypotheses related to the distinct features of e-commerce
The first hypotheses group deals with the characteristics that the Internet possesses and that have a major influence when buying on the Web. Exploratory research among friends and relatives showed that many aspects influence their attitude towards e-shopping. These aspects are: security/privacy, price, extra services, speed of delivery, cost of delivery, ease of Payment, user-friendliness of the website, after-sales services (insurance and guarantee).
- Product-category-related hypotheses
According to our findings, we have defined seven product categories: books, audiovisual media, software, electronic equipment, travel services, clothing and groceries. The reasons why certain products have Internet potential deserve further investigation.
- Hypotheses related to product characteristics
The nature of the products more likely to be bought via the Internet is described by a third set of hypotheses. Exploratory research showed that following attributes deserve further investigation, namely: higher value, standardization, customization, size and weight, and hard-to-find products.
- Hypotheses on demographic differences
Spending patterns, perceptions of importance and attitudes might differ across nationalities, age groups, education levels and genders. The managerial usefulness of this hypotheses-group is to provide guidelines for socio-demographic segmentation, product positioning and marketing programs aimed at specific target groups.
Data Collection Method
Data collection was administered using a questionnaire closely related to the hypotheses. The questionnaire was distributed in several different ways to achieve an overall coverage of the targeted sample population. As a starting point, the questionnaire was distributed via e-mail. The advantage of this method was the fast reply and little time invested on behalf of the research group. The problem inherent in this method was the limited access to data from different age groups and income classes as well as people from without convenient access to the Internet. Therefore, a street survey was carried out in Germany and the Netherlands including personal face-to-face interviews. These were especially helpful in finding more extensive explanations for the responses given to the questionnaire. On the other hand we tried to minimize the bias originating from the presence of the interviewer who might influence the respondent in his decision process.
Data Collection Form
In order to obtain survey results the design of the questionnaire is of great importance. The time to reply to the questionnaire should be as short as possible. Due to a study the refusal rate of a questionnaire is 41% when it takes 6-12 minutes to answer and only 21% when the time taken is less than five minutes. We therefore kept the questionnaire short, the questions being answerable in ca. five minutes. Furthermore, several aspects have been considered regarding the content and structure of the questionnaire. The language of the questionnaire is easy to understand and should avoid vague or ambiguous words. In our case this was especially important in the English version of the questionnaire, which was also distributed to Dutch respondents. The answers options are mutually exclusive and at least one answer is applicable to every respondent.
Concerning the structure, the sequence of the questions is important. A screening question allows sorting out respondents, which do not have an opinion on buying via the Internet. We continued with questions related to our hypotheses using a Likert-scale from 1 to 5. Finally, personal questions were asked concerning the place of living, age, income class, education level and gender.
Descriptive Research
After having conducted the exploratory research phase successfully we clarified the nature of the problem and gained a first picture of the market situation on which we based our research objectives. Subsequently, in this section we perform a brief descriptive investigation of our dataset in order to enlarge preliminary understanding of the research topic and to describe the characteristics of our sample population.
“Descriptive research is concerned with determining frequency with which something occurs or the relationship between two variables”[4]
As stated earlier in this report our main collection methods comprised e-mails, street surveys and face-to-face interviews. The collection methods limited the extent to which the sample statistics are representative. However, the fact that we focus on internet-active persons counteracts the likelihood of a biased sample.
Regarding the limited sample size of 209, the distribution of gender indicates a representative sample consisting of 56 % of males and 43 % of females. The distribution of age raises doubts that the sample might be biased with reference to the fact that 70% of the respondents are in the age group of 18 to 25. Although this could be due to the high occurrence of this age group in our emailing collection method, another explanation seem to be much more convincing: we incorporated a screening question asking for the respondents’ attitude towards the Internet. A large amount of older internet-averse persons might have been filtered out, thus being irrelevant for our analysis and subsequently affecting the dataset. This suspicion gains further support by regarding the distribution of education. The major education level of our respondents was clearly academic with 52,2%, college with 20,6% and High-school level with 14,4%. Obviously, these education-levels are closely related to the above mentioned age interval and a relatively high degree of Internet affinity implying logical consequences due to the dataset. The distribution of income implies a similar reasoning, with having a high amount of people still in the educational phase we find 56% earning below 15.000 Euro and 15,8% with an income between 15.000 and 22.500 Euro.
However, bearing in mind those characteristics, we will analyze the resulting cause and effect relationships in greater detail in the inferential statistics part later in this paper.
3.1 Statistical Analysis
After collecting, entering and editing the data from our questionnaires, we screened the dataset for outliers and irrelevant respondents. The dataset was established using Excel, for it provided a more convenient tool for descriptive data analysis and data screening than SPSS in the eyes of the research team. The 43 items of our questionnaire were coded, assigning a specific number to each possible response. Interval scales indicating the relative importance attached to a certain feature could be easily converted into data, leaving the numbers as they were (1 = very unimportant up to 5 = very important). Nominally scaled variables have been entered as dummies (1 = yes; 0 = no), while ordinal variables carry a tag for every possible category (e.g. for education: 0 = none, 1 = High School, 2 = College…). Ratio variables were entered as they appeared in the questionnaire. With this consistent labeling scheme, the computer software was able to perform tabulations, graphs and classify the data according to the categories.
3.2 Statistical tests performed
The research design and the nature of the scales used in the questionnaire (nominal, ordinal, interval and ratio) allows using all the statistical techniques at hand. The analytic methods can be subdivided into two categories: 1. analysis of one separate variable, 2. joint analysis of several variables.
Since most of the hypotheses are concerned with proving the relative (un)importance of specific factors in e-commerce, one-sample t-tests are used. To observe the mean of a single variable (e.g. data security on a 1 to 5 Likert scale) and prove any significant difference to the assumed population mean (assuming neutrality = 3), the null hypothesis whether the mean is equal to 3 is tested against a one- or two-sided alternative. The mean difference establishes a possibility to rank relative importance. To detect whether there is a statistically significant difference in the answers of two groups, an independent t-test or ANOVA can be conducted. Levine’s test for the equality of variance gives decisive hints how to interpret the results, despite possible deviations from the assumption of equal variance among the samples compared. The tests which can be used for joint analysis hinge on the nature of the measurement of the variables involved. A simple regression relates one independent factor to the dependent variable while a multiple regression relates the simultaneous impact of various independent factors on the dependent variable.
All tests are performed at a 95 % confidence level, leaving a 5 % possibility of a type I error, that is to reject a valid null-hypothesis on the basis of the sample statistics.
Table 2 – Hypotheses, Data and Inferential Methods Used
illustration not visible in this excerpt
4. Hypothesis testing – tests and assumptions
The hypotheses stated below are tested on the basis of a one-sample t-test, a two-sample independent t-test or ANOVA. In our questionnaire we used a 1 to 5 Likert scale in order to indicate the importance of certain product attributes when buying via the Internet and the likelihood of buying certain product categories. A value between 4 and 5 indicates high importance, while values between 1 and 2 signify unimportant features. We assume the population to be normally distributed around a mean of 3 (indicating indifference). Furthermore, we assured that our respondents were sampled at a random basis and therefore assume the selected samples to be independent.
Hypotheses group 1 – distinct features of e-commerce
Hypotheses
1.1 H0: Consumers are indifferent concerning security and privacy aspects when purchasing via the Internet.
Ha: Consumers consider security or privacy aspects important or explicitly unimportant when purchasing via the Internet.
- One-sample t-test: H 0 of m = 3 is tested against Ha of m ¹ 3 (Appendix p. 22).
1.2 H0: Consumers are indifferent concerning price advantage when purchasing via the Internet.
Ha: Consumers consider price advantage important or explicitly unimportant when purchasing via the Internet.
- One-sample t-test: H 0 of m = 3 is tested against Ha of m ¹ 3 (Appendix p. 22).
1.3 H0: Consumers are indifferent concerning the cost of delivery when purchasing via the Internet.
Ha: Consumers consider the cost of delivery important or explicitly unimportant when purchasing via the Internet.
- One-sample t-test: H 0 of m = 3 is tested against Ha of m ¹ 3 (Appendix p. 22).
1.4 H0: Consumers are indifferent concerning time of delivery when purchasing via the Internet.
Ha: Consumers consider time of delivery important or explicitly unimportant when purchasing via the Internet.
- One-sample t-test: H 0 of m = 3 is tested against Ha of m ¹ 3 (Appendix p. 22).
1.5 H0: Consumers are indifferent concerning extra services when purchasing via the Internet.
Ha: Consumers consider extra services important or explicitly unimportant when purchasing via the Internet.
- One-sample t-test: H 0 of m = 3 is tested against Ha of m ¹ 3 (Appendix p. 22).
1.6 H0: Consumers are indifferent concerning ease of payment when purchasing via the Internet.
Ha: Consumers consider ease of payment important or explicitly unimportant when purchasing via the Internet.
- One-sample t-test: H 0 of m = 3 is tested against Ha of m ¹ 3 (Appendix p. 22).
1.7 H0: Consumers are indifferent concerning after-sales services when purchasing via the Internet.
Ha: Consumers consider after-sales services important or explicitly unimportant when purchasing via the Internet.
- One-sample t-test: H 0 of m = 3 is tested against Ha of m ¹ 3 (Appendix p. 22.)
1.8 H0: Consumers are indifferent concerning user friendliness of the website when purchasing via the Internet.
Ha: Consumers consider user friendliness of the website important or explicitly unimportant when purchasing via the Internet.
- One-sample t-test: H 0 of m = 3 is tested against Ha of m ¹ 3 (Appendix p. 22).
Data Highlights: measured on 1 to 5 Likert scale (1 = unimportant, 5 = very important)
illustration not visible in this excerpt
Analysis
Regarding the t-test results of our first dataset one can examine that the perception about all tested buying process attributes differ significantly from 3. This observation is reflected most evidently with the security/privacy aspect holding a mean score of 4,6 and an overwhelming t-ratio of 36,025 leaving us with the conclusion that people regard the security/privacy issue as a very important one. A large group of other aspects comprising better service, speed of delivery, cost of delivery, ease of payment, user friendliness and after sales service is clustered around a mean of four. For this group, there is significant evidence against the null with t-ratios between 5 and 15 - buying process attributes are regarded as important. The only exception within our first hypotheses group is the extra services aspect that reveals a t-ratio of -5,448 and a mean score of 2,53, which lies significantly below our assumed mean of 3. Therefore, we can reject the null in favor of the alternative that consumers do not regard extra services to be important, we like to point at the low magnitude of the deviation. The question itself caused some non-decisive answers since respondents had problems imagining possible valuable extra services that would be important to them.
Managerial implications
Observing these test results one can conclude that customers developed a demanding evaluation basis when shopping on-line. Moreover, this implies that decision makers should handle these relevant aspects with the appropriate consciousness and technical expertise. Special importance should be directed to the security/privacy issue. Companies ideally provide a secure and convenient shopping environment to the potential customer.
Hypotheses Group 2 – Product Categories
Hypotheses
2.1 H0: Customers are unlikely to buy books, newspapers and magazines via the Internet during the next year.
Ha: Customers would seriously consider buying books, newspapers and magazines via the Internet during the next year.
- One sample t-test: Ho of µ ≤ 3 is tested against Ha of µ > 3 (Appendix p. 23).
2.2 H0: Customers are unlikely to buy audiovisual media via the Internet during the next year.
Ha: Customers would seriously consider buying audiovisual Media via the Internet during the next year.
- One sample t-test: Ho of µ ≤ 3 is tested against Ha of µ > 3 (Appendix p. 23).
2.3 H0: Customers are indifferent whether to buy software/updates via the Internet.
Ha: Customers prefer or dislike buying software/updates via the Internet.
- One sample t-test: Ho of µ= 3 is tested against Ha of µ ≠ 3 (Appendix p. 23).
2.4 H0: Customers are indifferent whether to buy electronic equipment via the Internet.
Ha: Customers prefer or dislike buying Electronic Equipment via the Internet.
- One sample t-test: Ho of µ= 3 is tested against Ha of µ ≠ 3 (Appendix p. 23).
2.5 H0: Customers are likely or indifferent to buy travel services via the Internet.
Ha: Customers dislike buying Travel Services via the Internet.
- One sample t-test: Ho of µ ≥ 3 is tested against Ha of µ < 3 (Appendix p. 23).
2.6 H0: Customers are likely or indifferent to buy clothing via the Internet.
Ha: Customers dislike buying clothing via the Internet.
- One sample t-test: Ho of µ ≥ 3 is tested against Ha of µ < 3 (Appendix p. 23).
2.7 H0: Customers are likely or indifferent to buy groceries via the Internet.
Ha: Customers dislike buying groceries via the Internet.
- One sample t-test: Ho of µ ≥ 3 is tested against Ha of µ < 3 (Appendix p. 23).
Data Highlights
illustration not visible in this excerpt
Analysis
For books, audiovisuals, software, hardware, clothing and groceries, the observed p-values are sufficiently close to zero. Therefore, a conclusion can be drawn that the population mean is considerably different from 3 in these product categories. The categories books, audiovisuals and software have a mean significantly above three indicating that these categories play a great role in e-commerce. Especially books, are most likely to be bought via the Internet which is reflected in a mean of 3,95 and a higher t-ratio compared to the other categories. The same applies to audiovisual media, which also belongs to the categories which are more likely to be bought. In both cases the null hypothesis can be rejected stating that customers do not like or are indifferent when buying these products via the Internet. Concerning software, the null hypothesis can be rejected that customers are neither likely nor unlikely to buy software via the Internet in the near future. In contrast, electronic equipment, clothing and groceries are rather unlikely to be bought via the Internet given that the mean difference is negative in these categories.
As far as travel services are concerned, the statistical analysis shows significant results. The null hypothesis cant be rejected. As a conclusion, people are neither very likely nor unlikely to buy travel services via the Internet in the future, in other words, they do not have an opinion on the issue or buy via the Internet and the real marketplace.
Managerial Implications
As the analysis concerning the product categories shows, books, audiovisual media and software are the products that are most likely to be bought via the Internet in the near future.
illustration not visible in this excerpt
An explanation for this is the high degree of standardization in these categories. Customers feel familiar with their purchasing decision and repeat purchases are more likely to occur. Books and audiovisual media such as CD’s are already the most common products to be bought via the Internet as compared to products such as clothing, food and cosmetics. And indeed, as the statistical analysis reveals, highly customized products such as clothing are unlikely to be bought via the Internet in the near future. An additional explanation might be the importance for the consumer to touch and feel the product before the purchasing decision. Internet sellers realize this issue more and more and therefore take unwanted products back and offer money back guarantees within a certain time frame . If consumers are made more aware of these options they would feel safer to buy more customized, individual products via the Internet. Improving the overall services offered to customers would open new possibilities for certain industries not yet successful in the world of e-commerce.
4.3 Hypotheses group 3 – Product-feature related hypotheses
Hypotheses
3.1 H0: Consumers are indifferent or likely to buy higher-value products in the real marketplace or the Internet
Ha: Consumers prefer to buy higher-value products in the real marketplace
- One-sample t-test: H0 of m ³ 3 is tested against Ha of m < 3 (Appendix p. 24).
3.2 H0: Consumers are indifferent to buying standardized products in the real marketplace or the Internet
Ha: Consumers prefer to buy standardized products in the Internet or the real market.
- One-sample t-test: H0 of m = 3 is tested against Ha of m ¹ 3 (Appendix p. 24).
3.3 H0: Consumers are indifferent or likely to buy customized products in the real marketplace or the Internet
Ha: Consumers prefer to buy customized products in the real marketplace
- One-sample t-test: H0 of m ³ 3 is tested against Ha of m < 3 (Appendix p. 24).
3.4 H0: Consumers are indifferent to buying higher weight/size products in the real marketplace or the Internet
Ha: Consumers are not indifferent to buying higher weight/size products in the real marketplace or the Internet.
- One-sample t-test: H0 of m = 3 is tested against Ha of m ¹ 3 (Appendix p. 24).
3.5 H0: Consumers are indifferent or unlikely to buy hard-to-find products in the real marketplace or the Internet
Ha: Consumers prefer to buy hard-to-find products in the Internet.
- One-sample t-test: H0 of m £ 3 is tested against Ha of m > 3 (Appendix p. 24).
Data Highlights: measured on 1 to 5 Likert scale (1 = real market, 5 = Internet) 5
illustration not visible in this excerpt
Analysis
Besides for hard-to-find products, the statistical analysis points toward a clearly significant deviation from neutrality which is, however, in two cases negligible in magnitude. The table above signifies a clear importance ordering of the product features.
Consumers clearly prefer buying customized commodities (clothing) and higher value products in the real marketplace. The mean differences of –1,43 for “Customized products” and –1,41 for “Higher-value products” is significant at the 95 % confidence level with the very low p-values pointing towards possibly even higher confidence levels. In both cases, the values for the standard error of the mean highlight the small spread of the sampling distribution. This can be viewed as further evidence of the relative importance of the characteristics in question. Surprisingly, consumers are rather indifferent where to buy hard-to-find products. With a mean difference from 3 of only 0,35, which is, however, significant at the 95 %-level. Response on this item appears to be undecided. This might be due to the fact, that the respondents associated different products with the hard-to-find category. Face-to-face interviews in Aachen and Maastricht provided the researchers with some insight into people’s understanding in the question, with the quoted products ranging from Cuban cigars to special motorbike equipment. Because of the diversity of the answers, the inconclusive result on this item might arise from the perceptual bias involved on the respondent’s side. The same conclusion holds for products of high weight/size. With a mean difference of –0,61, the t-ratio is highly significant and the null-hypothesis can be rejected at a 5 % level. However, interviews showed equally diverse responses as it was the case with hard-to-find products. Statements ranged from “Furniture is clearly heavy and bulky. But I would like to touch and see my new couch before I buy it”, thus clearly opting for 1 = real marketplace, to “Well, if the delivery is quick and reliable, I would order heavy and bulky products via the Internet.”. The differential response on this item in the questionnaire explains the small mean-difference and the large standard deviation. From the underlying data, no clear-cut conclusions could be derived.
Finally, for standardized products, where the researchers expected the greatest bias towards the Internet, a mean-difference of 0,75 (significant at 5 % level) does support our expectation, though not to the desired extent. This might be due to the fact that while consumers are more inclined to buy standardized commodities via the Internet still purchase large quantities from the shelves.
Managerial Implications
The insights gained are rather disappointing. Although standardized products seem to have some e-commerce potential, the data analysis shows that higher-value products, customized products, products of high weight/size and hard-to-find products are bought in the real marketplace (or indifference). Consumer behavior still seems to be dictated by the habit to go to the store. Especially products requiring limited or extensive problem-solving (higher-value products) or else including a highly individual and emotional response (customized products) seem to have virtually no chance to be sold via the Internet. Despite the promises of high profit margins and a huge potential for product differentiation in these product categories, investments should be postponed until buying behavior becomes more favorable. “Hard-to-find” products might eventually provide some potential for market-nichers, if more differentiated results can be obtained that eliminate the perceptual bias recognized in this research and reduce the degree of uncertainty. Further research in this direction is needed.
4.4 Hypothesis group 4 – Demographic differences
Hypotheses
4.1 H0: There is no difference in the perception of the factors influencing buying behavior between males and females.
Ha: There is a significant difference in the perception of the factors influencing buying behavior between males and females.
- Two-sample independent t-test (Appendix p. 25-27).
4.2 H0: There is no difference in the perception of the factors influencing buying behavior across age groups.
Ha: There is a significant difference in the perception of the factors influencing buying behavior between age groups.
- Two-sample independent t-test (Appendix p. 31-33).
4.3 H0: There is no difference in the perception of the factors influencing buying behavior across education levels.
Ha: There is a significant difference in the perception of the factors influencing buying behavior between education levels.
- ANOVA (Appendix p. 34-36).
4.4 H0: There is no difference in the perception of the factors influencing buying behavior between Americans and Europeans.
Ha: There is a significant difference in the perception of the factors influencing buying behavior between Americans and Europeans.
- Two-sample independent t-test (Appendix p. 28-30).
4.5 H0: There is no difference in the perception of the factors influencing buying behavior across income levels.
Ha: There is a significant difference in the perception of the factors influencing buying behavior between income levels.
- ANOVA (Appendix 37-39).
Data highlights (See Appendix)
Analysis
Looking at the overall picture, the data supports the view that only minor differences between demographic groups persist. At the five percent significance level, the majority of the null-hypotheses could not be rejected. However, some interesting findings deserve to be mentioned:
- Between males and females, it seems that the main divergences are about buying software and hardware, and buying “higher-value”[6] products. There is a favorable difference for males, which seem to have higher interest in buying such items via the Internet. In contrast, attitude of both genders towards sensible issues such as security/privacy, user friendliness of the web site, cost of delivery is similar. The tests show that both sexes are equally alike to buy books and travel services via the Internet[7].
- Americans and Europeans have different buying intentions as regards travel services, clothing, “hard-to-find” and “higher-value” products (see appendix): For travel services, the American mean of 4.38 on the Likert-scale is 1.21 point higher than the European mean with a p-value of 0.7 %. Here, we should state statistical values. However, the two groups are totally indifferent when it comes to other items, such as user friendliness, security/privacy, better service, or purchases of software, groceries or books.
- Generally, age groups do not display significant differences in their preferences when shopping via the internet. Significant results were obtained with regard to the following distinct features: security/privacy, speed of delivery, ease of payment and user-friendliness of the website. E-buyers older than 24 year consider user-friendliness of the website more important than younger buyers. Concerning the younger age group, features such as security/privacy, speed of delivery and ease of payment which are relevant for overall convenience are more important. Underlying reasons could be that younger e-buyers are less experienced with payment issues, especially with credit cards. Consequently, younger people also consider the privacy issue as important, which signifies a new and more sensitive perception of how to deal with personal data. They tend to rely more on their parents to deal with this topic and are therefore insecure in handling these issues.
- Education really matters. Differences have been observed with regard to books, CDs and travel services. The results indicate that academics tend to buy books, magazines and reading material over the internet. The strong relation between education level and this buying pattern complies with basic intuition. Academics tend to be more price sensitive. This conclusion is straightforward, since many students are included in the sample. Students are more seriously affected by income constraints, especially for books, travel services, and audiovisuals.
- Personal yearly income has a negligible influence on the product choices and perception of several items. However, some aspects yield highly significant results, namely security and privacy, speed of delivery, and ease of payment, which are more important to the lower income group. Evidently, lower income groups also represent the younger generation, which has a different perception of data security mentioned above. Above all, people with a higher income are often more educated and Internet-savvy.
Managerial Implications
Suffusing all else, despite the fact that most of the demographic factors proved irrelevant, specific differences are extremely relevant to management. According to this report, managers should definitely use different target strategies, when it comes to gender, education level, and country. Security/privacy, speed of delivery, and other Internet-related issues are mostly equally perceived by all genders, by all ages, and in both America and Europe. From here, we can conclude that security/privacy issues, speed of delivery, user friendliness and cost of delivery are global when buying via the Internet. Could we conclude from here that Internet using and buying involves people with the same spirit or cultures among different groups, or is it that Internet by itself helps shaping homogeneous minds and cultures within many social groups? The most surprising managerial information is the general indifference of age groups toward their buying preferences via the Internet. This homogeneity of expectations is certainly a good message to management that does not need different marketing strategies adapted to age groups.
4.5 Multiple Regression
After analyzing the data descriptively, via t-tests and ANOVA, we would like to draw some further conclusions using a multiple regression. We expect that it will allow us to reach a deeper understanding whether and how certain variables like demographic or internet-specific factors influence customers’ buying behavior.
One important consideration in deciding on the way of analysis used for this research was a shortcoming of the above described methods. While testing the significance of a single variable, effects of other variables might be incorporated, resulting in incorrect conclusions. Since the proposed regression model includes several variables, the respective effects will be covered by the accompanying variable, if the model is constructed properly. Another merit of this analysis lies in the capability to predict customers’ behavior on the grounds of his or her characteristics. We will go further into this subject, when evaluating the results.
Along with those advantages, there are several disadvantages to be brought up. Because of including more variables a higher number of degrees of freedom will be lost. In this research, it should not be too problematic, since we obtained quite a large number of observations. A second factor to be mentioned is the problem of multicollinearity. In a poorly constructed model, variables that are highly correlated might take over effects of the other variable, resulting in dubious conclusions. To prevent this several tests can and will be conducted, which are beyond the scope of this paper.
A couple of assumptions must be made for this method of analysis. For one is the necessity to translate some indications made in $ US into NLG. To do so, a real exchange rate is needed.[8] On the other hand there are several assumptions implicit in the model, which can be looked up in “Optimization Techniques”, Part II Statistics written by Drs. C. Kerckhoffs.
Hypotheses
There are five main hypotheses and respective expectations of interest to be tested here:
1. H0: There is no difference in purchasing behavior between Europeans and US Americans.
Ha: Nationality causes a difference in buying behavior via the internet.
We expect Americans to purchase more via the Internet.
2. H0: There is no difference in purchasing behavior between men and women.
Ha: Gender causes a difference in buying behavior via the internet.
3. H0: Purchasing behavior does not differ between income groups.
Ha: Income has an influence on the shopping behavior.
An expectation here, would be that the more people earn, the more they are willing to
spend on the Internet.
4. H0: Purchasing behavior does not differ between age groups.
Ha: Age has an influence on the shopping behavior in the Internet.
We believe that age would be a significant factor in the determination of purchases in the
Internet.
5. H0: Purchasing behavior does not differ between groups of different education.
Ha: Education causes a difference in buying behavior via the internet.
Model[9]
The most appropriate model turned out to be:
m.produc = b0 + b1*gender + b2*country + b3*Inc30 + b4*Inc45 + b5*Inc60 + b6*Inc100 +
b7*age
where: m.produc = money spend on products purchases via the Internet
gender: if gender = 1, the person is a woman, base level: male
country: if country = 1, the person lives in the USA, base level: European
Inc30 – Inc100: dummies for characterizing income groups, base level: below
NLG 30,000: Inc30 = Income group between NLG 30,000 and 44,999
Inc45 = Income group between NLG 45,000 and 59,999
Inc60 = Income group between NLG 60,000 and 100,000
Inc100 = Income group above NLG 100,000
Evaluation
1. H0: b2 = 0 Ha: b2 ¹ 0
As can be seen by the t-ratio of 2,052 and a significance level of 0,041 we can reject the null hypothesis. Since 2,052 > 0, we can conclude that US Americans purchase significantly more via the Internet that Europeans do.
2. H0: b1 = 0 Ha: b1 ¹ 0
From a t-ratio of 1,119 we fail to reject the null hypothesis. Therefore, the inference can be made that there is no difference between men and women in their purchasing behavior via the Internet.
3. H0: b3 = b4 = b5 = b6 = 0 Ha: one of the coefficients differs from zero
For determining this result, a general Wald test needs to be conducted. As the results in the appendix show, we can reason that the null hypothesis has to be rejected, concluding that the income of the particular person influences the desire to buy products in the Internet. Judging from the SPSS output, we can see, that an income between NLG 45,000 and 59,000 results in the highest purchases in the Internet. Surprisingly, as people earn a higher salary, the Internet consumption decreases.
4. H0: b7 = 0 Ha: b7 ¹ 0
Contradicting our expectations, we can assume from a t-ratio of 1,086 and a p-value of 0,279 that age does not matter in determining amounts of money spend on the Internet.
5. There are no t-ratios or other statistical data included in this report, since the problem of multicollinearity between income and education arose. From perform regressions during the testing phase of the proper model, we can conclude that education did not play a role in this context.
Managerial Implications
One important benefit from such an analysis is the possibility of defining a reference market and controlling, whether efforts taken toward a predetermined reference market were successful. There is no conclusive advice, we can give here, since we found little evidence resulting from the performed analysis. This arises from the relatively small sample size compared to the wide range of variables used, which were generated from the fact that the base population is represented by all the people using the Internet and considering buying products via the World Wide Web.
5. Limitations and Suggestions for further Research
The results presented in this paper is subject to certain limitations such as the lack of time-series data, the sample size, and the data collection method.
Since the short time frame did not allow us to collect data at different points in time, the research conducted does not mirror possible trends and restrains the predictive power of the inferences drawn. Besides the constraints imposed by the cross-sectional nature of the study, inferences based on the relatively small sample size compared to a total potential target population of millions of people can barely be generalized. The research’s findings are not representative, its external validity remains small. Finally, the data collection method in the USA and Europe differed. A bias originated due to the fact that Americans could only respond via email, while face-to-face interview additionally took place in Europe.
Continuing research effort should thus be devoted to time series analysis, enlarged samples and standardized collection methods among a more international cross-section of potential e-buyers. Moreover, the wide scope of the problem statements, sub questions, and hypotheses limited the researchers’ ability to gain profound insight into specific motivational factors, a wider set of e-commerce process characteristics and product attributes. Future research should investigate those specific factors more deeply.
6. Conclusion
The managerial guidelines stated beneath each set of hypotheses should be used to develop an e-commerce strategy for B2C-companies.
Decision makers should pay great attention to the shopping process as such. In order to attract potential e-buyers, shopping has to be safe and convenient. Especially customer retention is very likely to be heavily affected the way customers’ online experience. Given negligible switching cost and low brand loyalty on the consumer side, lowering acquisition and retention costs as well as encouraging repeat purchases on the firm’s side is crucial for any e-commerce project in order to pay off. Manager’s have to fine-tune e-commerce processes, keep up with the latest trends in data security, procurement and billing technology in order to succeed.
However, technological leadership may not be sufficient in a market, that is distinguished by a strong demand for standardized products carrying a low margin. Potential for differentiation is small and even market and cost leadership do not seem to guarantee profitability (see amazon.com). Still, huge marketing efforts will be necessary to induce customers to purchase via the Internet, while future payoffs remain uncertain. The lack of significant differences between demographic groups limits the firm’s ability to segment the market, enhance marketing effectiveness, differentiate product lines and gain competitive advantage along these dimensions.
E-commerce will always take place in an environment, where functions, customer groups and technologies change at an even faster pace. Competitive forces will be strong. Consumers have a large bargaining power and can compare prices. Products can be easily substituted. Successful business processes are reengineered and barriers to entry are low. Fast changing needs and short product life cycles will challenge existing players’ position in the market and provide opportunities for innovative firms.
Looking at the list above, it becomes apparent that strategic marketing is vital to the success of e-commerce. In the words of Henry Yang, founder of Yahoo!: “The Internet is just the tool-case, and as such incapable of solving a single problem.”. We hope that our research provided some useful information for future academic research and business practitioners alike and that we took some tools out of the tool-case. Next, customers’ problems are to be solved.
References
Aker, Kumar, Day; “Marketing Research”; 6th edition; Wiley.
Bagozzi, Richard P.; “Principles of Marketing Research”; Blackwell Business.
Churchill, Gilbert A.; “Marketing Research – Methodological Foundations”; 6th edition, Harcourt Press Publishers.
Kalakota and B. Whinston; “Electronic Commerce, A manager’s guide”, Adison Wesley.
Lambin, Jean-Jacques; “Market-driven Management: Strategic and Operational Marketing”; London, Macmillan.
www.wiwo.de
www.forrester.com
7. Appendix
7.1 Hypotheses Group I: Distinct Features of e-Commerce
Output 1: One-sample t-test on the distinct features of e-commerce
illustration not visible in this excerpt
7.2 Hypotheses Group II: Product Categories
Output 2: one-sample t-test on product categories
illustration not visible in this excerpt
7.3 Hypotheses Group 3: Internet vs. Real Marketplace
Output 3: one-sample t-test on product characteristics
illustration not visible in this excerpt
7.4 Hypotheses Group IV: Demographic Differences
Output 4: Differences in Hypotheses-Group I between Genders
illustration not visible in this excerpt
Output 5: Differences in Hypotheses Group II between Genders
illustration not visible in this excerpt
Output 6: Differences in Hypotheses-Group III between Genders
illustration not visible in this excerpt
Output 7: Differences in Hypotheses Group I between USA and Europe
illustration not visible in this excerpt
Output 8: Differences in Hypotheses Group II between USA and Europe
illustration not visible in this excerpt
Output 9: Differences in Hypotheses Group III between USA and Europe
illustration not visible in this excerpt
Output 10: Differences in Hypotheses Group I between Age Groups
illustration not visible in this excerpt
Output 11: Differences in Hypotheses Group II between Age Groups
illustration not visible in this excerpt
Output 12: Differences in Hypotheses Group III between Age Groups
illustration not visible in this excerpt
Output 13: Differences in Hypotheses Group I between Education Levels
illustration not visible in this excerpt
Output 14: Differences in Hypotheses Group II between Education Levels
illustration not visible in this excerpt
Output 15: Differences in Hypotheses Group III between Education Levels
illustration not visible in this excerpt
Output 16: Differences in Hypotheses Group I between Income Groups
illustration not visible in this excerpt
Output 17: Differences in Hypotheses Group II between Income Groups
illustration not visible in this excerpt
Output 18: Differences in Hypotheses Group III between Income Groups
illustration not visible in this excerpt
7.5 Multiple Regression – Model I and II
Output 19: Multiple Regression Model I
illustration not visible in this excerpt
Output 20: Multiple Regression Model II
illustration not visible in this excerpt
7.6 Graphs and Figures
illustration not visible in this excerpt
7.7 Questionnaire – English Version
Questionnaire Survey
Purchase Behavior @nd the Internet
Dear participant,
we are five students conducting an international marketing survey about purchasing via the Internet. The aim of our survey is to describe the products that are most likely to be bought in the future and why.
The following questionnaire includes 10 easy-to-answer questions asking for your general opinion on online-shopping. It will take you only 5 minutes to answer the questions. We would appreciate your co-operation, which is of crucial importance to the success of this market research.
The survey is anonymous: all data will be treated highly confidential and for academic research purposes only.
Thank you
If you are interested in receiving a free report on the findings of this research, just leave your email address: . We will be glad to send you a copy when ready.
1. Would you personally consider buying products (order books, CDs) or services (e.g. travel services, etc.) via the Internet?
illustration not visible in this excerpt
2. How important are the following points to you when buying via the Internet?
illustration not visible in this excerpt
3. Which products would you most likely buy via the Internet?
illustration not visible in this excerpt
4. How often would you consider buying the following products via the Internet during the next year? Please give an estimate in terms of the number of items.
illustration not visible in this excerpt
5. Given your own estimates from your previous question, how much money would you spend on the following products during the next year?
illustration not visible in this excerpt
6. Where are you more likely to purchase products with the following characteristics?
illustration not visible in this excerpt
7. In which country do you live?
8. What is your age?
9. What is the level of education you have currently attained?
illustration not visible in this excerpt
10. Please give a rough indication of your yearly personal net income.
illustration not visible in this excerpt
11. What is your gender?
illustration not visible in this excerpt
7.8 Questionnaire – German Version
University of Maastricht
Faculty of Economics& Business Administration
Marketing Research 2001
Guten Tag,
wir sind fünf Studenten der University of Maastricht und führen ein internationales Marketingprojekt zum Thema „Einkaufen im Internet“ durch. Wir würden uns freuen, wenn Sie sich einen kurzen Moment Zeit nehmen würden, um mit Ihren Angaben den Erfolg des Projektes zu ermöglichen.
Selbstverständlich werden Ihre Angaben vertraulich behandelt und dienen lediglich der statistischen Auswertung im Rahmen unserer akademischen Aktivitäten.
Falls Sie Interesse an den Auswertungen und der endgültigen Analyse der Studie haben, notieren Sie einfach Ihre Email- Adresse am Ende des Fragebogens. Wir würden uns freuen, Ihnen frühstmöglich die Ergebnisse zukommen zu lassen.
Mit freundlichen Grüßen,
Benjamin Hittel, Cyrille Tagne Tamo, Maria Kimme, Mathias Wengeler, Hannah Wilke
12. Würden Sie generell Produkte über das Internet kaufen?
illustration not visible in this excerpt
13. Welchen Stellenwert würden bei Ihnen folgende Faktoren beim Einkauf im Internet haben?
illustration not visible in this excerpt
14. Welche der folgenden Produkte würden Sie am ehesten im Internet kaufen?
illustration not visible in this excerpt
15. Wie häufig würden Sie schätzungsweise folgende Produkte im Internet kaufen (pro Jahr)?
illustration not visible in this excerpt
16. Welche Gesamtsumme würden Sie maximal im Internet für folgende Produkte ausgeben (pro Jahr)?
illustration not visible in this excerpt
17. Wo würden Sie folgende Produktkategorien eher kaufen?
illustration not visible in this excerpt
18. In welchem Land leben Sie zur Zeit?
19. Wie alt sind Sie?
20. Wie hoch ist Ihr derzeitiger Bildungsstand?
illustration not visible in this excerpt
21. Bitte ordnen Sie Ihr derzeitiges, jährliches Netto-Einkommen folgenden Kategorien zu.
illustration not visible in this excerpt
22. Was ist Ihr Geschlecht?
illustration not visible in this excerpt
Vielen Dank für die Beantwortung des Fragebogens.
[...]
[1] www.wiwo.de
[2] www.forrester.com
[3] Lambin, p. 144
[4] Lambin, p.143
[5] One-sided p-value = two-sided p-value from output divided by 2.
[6] see table.
[7] see appendix
[8] We took a real exchange rate of $US 1 = NLG 1.79, computed with the nominal exchange rate and the purchasing power of one $US in the Netherlands, see www.economist.com
[9] see appendix p. 40-41.
- Quote paper
- Maria Kimme (Author), 2001, Purchase Intentions @nd the Internet - Which Products are most likely to be bought via the Internet and why?, Munich, GRIN Verlag, https://www.grin.com/document/110652
-
-
-
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X. -
-
Upload your own papers! Earn money and win an iPhone X.