The Internet is widely expected to become a channel for electronic commerce. The study discusses what kind of similarities Internet shopping has with conventional non-store shopping. It also discusses factors influencing Internet shopping, the benefits, motives, and risks. In an empirical WWW survey, a number of factors were found to increase the likelihood to shop on the Internet: previous activity in in-home shopping, computer or Internet related work, Internet experience, active Internet use, and product uniqueness. Risk due to inability to inspect the product, payment method and slowness of buying were found to decrease the likelihood to shop.
1. Introduction
1.1 Background for the study
1.2 The study objectives
1.3 Terminology
2. The Internet as a shopping medium
2.1 The new media
2.2 Media properties
2.3 Consumer's point of view
2.4 Marketer's point of view
3. In-home shopping and the Internet - Literature review
3.1 User profile
3.2 Benefits
3.3 Disadvantages
3.4 Electronic shopping channels
3.5 Explaining Internet shopping behaviour
4. User survey
4.1 Hypotheses
4.2 Survey
4.3 Hypotheses tested
4.4 Additional findings
4.5 Limitations of the results
5. Conclusion
5.1 Results of the survey
5.2 Explaining factors network
5.3 Implications for marketing
5.4 Suggestions for future research
6. Summary
7. References
Appendix A: Questionnaire
Appendix B: Limitations of WWW surveys
The Internet has grown exponentially in a couple of years. Internet commerce is expected to boom in the near future. However, little is known about what Internet commerce can be like in reality. Who are the potential customers, what are their motivations?
The number of Internet users has increased exponentially in the last few years, and this is expected to continue (Network Wizards 1996, Rutkowski 1996, Taloustutkimus 1996b). Although one cannot exactly know how many users there are, a minimum of 13 million users can be approximated by the number of Internet host computers connected. The figure is probably manifold, as users share the computers.
There is no consensus about what "Internet user" means. In some surveys it is WWW users, in others it is email users, weekly users, users who have tried at least once etc. In 1996, some user estimates are 23 million (WWW users, International Data Corp. 1996) to 35 million (U.S. Internet users, Hodges 1996). Forecasts are 120-200 million users by year 1999 (International Data Corp. 1996, Morgan Stanley 1996).
According to Yankelovich Partners (1996), 22% of the U.S. population use online services including the Internet, a +150% increase in 2 years. Yankelovich Partners predict the growth rate will slow down. According to Taloustutkimus (1996a, 1996b), 18% of the Finnish population has used the Internet at least once, a +40% increase in 6 months. According to Nielsen (1996), +17% of the U.S. and Canadian population have used the Internet in the past 6 months. But still, there are countries in which food is scarce, let alone telephones and Internet connections.
The amount of sales on the Internet has increased too (Gupta, Pitkow & Recker 1995) being USD 130 million in June 1996 (ActivMedia 1996), and it is expected to soar (USD 23 billion in year 2000, Hambrech & Quist 1995, ref. CyberAtlas 1996) in the coming years as technical solutions mature and new customers enter the market.
A typical Internet user is very interesting from the perspective of marketing because the user has buying power and is willing to buy. A great majority (over 2/3) of active Internet users are positive towards Internet shopping (GVU 1994b, 1995, Rissa 1995, Rissa & Järvinen 1996).
Still, only a fraction (15%-50%, depending on how active Internet users the respondents were) of the users have actually bought something (Nielsen 1995, Rissa & Järvinen 1996, GVU 1996). According to GVU (1996), regional differences exist; Internet shopping is more common in the U.S. than in Europe. But after all, the number of potential customers is vast.
There are not too many scientific publications on Internet shopping because the branch is new and it is developing rapidly. Moreover, quite little research has been conducted on electronic commerce in general.
However, a significant number of surveys have been conducted on the Internet that treat of the profile of the Internet user and the use of the Internet (including Gupta, Pitkow & Recker 1995, GVU 1994-1996, Nielsen 1995, Nielsen 1996, Rissa 1995, Rissa & Järvinen 1996, Taloustutkimus 1996a). The research predicts that Internet commerce will flourish after certain problems, like security issues, have been worked out, and more consumers start using the Internet.
The objective of this study is to compare Internet shopping to conventional in-home shopping.
These issues will be dealt with from the perspective of the consumer. The study consists of an introduction to the Internet as a marketing medium, a literature review on in-home shopping, in-home shoppers and Internet users, and an empirical survey of experienced Internet users.
Issues that will not be dealt with include when a marketer should enter the Internet, what kind of products should be offered, different technical solutions, and the role of Internet service providers.
There is no widely accepted, clear definition of non-store or
in-home shopping. Gehrt, Ingram, and Howe (1991) have proposed
the following definition for non-store retailing.
Non-store retailing consists largely of exchanges transacted via mail and telephone, but also includes exchanges trasacted via party-selling, door-to-door selling, at-work selling and vending machines. It is highly similar to direct markteing, but does not include business-to-business transactions.
In this study, the following terminology will be used:
| Non-store shopping | A way for a consumer to buy without physically attending to a store. |
| In-home shopping | Non-store shopping that happens at home. Does not include shopping from vending machines or at work. |
| Conventional in-home shopping | In-home shopping that does not involve the Internet. Mail-order catalogs, magazine and newspaper advertisements, direct mail, telephone selling, door-to-door selling, party selling, TV shops, etc. |
| Internet shopping | A way to buy on-line without attending to a store. The customer can either buy via the Internet, or see some product information on-line and then order off-line, using telephone, fax or regular mail. |
The Internet is not just one medium. In fact, it is a collection of various different media, and a number of them are useful for marketing purposes.
At first, the Internet was used to transmit mainly textual data. However, in 1993, a rapid development began towards multimedia with the introduction of the WWW. Today, the WWW is the most commonly used medium to distribute public marketing information.
Using the WWW the user can get information from all over the world, including text, pictures, video, voice, computer programs, i.e. any kind of digital data. A certain degree of interactivity between the shop and the customer is also possible. The WWW is mostly a public service.
Electronic product catalogs can be implemented using the WWW. These catalogs are in many ways similar to their paper counterparts. The WWW can be used to send order forms, or to distribute digital products to the customer.
Email is a personal, interactive medium between two or more people. Email is usually text-only, and it is often considered a very private form of communication.
Email can be used as a direct mailing channel - although certain restrictions apply, see Recipient's attitude. Another way is to be in touch with the customer like in conventional telephone selling.
Newsgroups are a version of public bulletin boards. Newsgroups are usually used for textual conversation and announcements among a group of users. The membership of these groups is open for everyone.
If used with taste, newsgroups can be used to distribute targeted marketing information. Most of the groups consist of people who are interested in one particular issue, thus making them an effective segmentation tool.
Mailing lists share the properties of email and newsgroups. They are like clubs. From the marketing point of view, mailing lists are quite similar to newsgroups, except that they are often considered a more private medium than newsgroups.
Telnet and rlogin are methods to use another computer over the Internet. They can be used to implement various complex text based applications. Banks have used them to implement systems where the customer can make basic bank transactions, view the balance etc.
FTP is a method of transferring files between computers. A large number of so called anonymous FTP servers exists where anyone can visit and get various files.
From the marketer's point of view, FTP is useful to distribute sample files, like free demonstration versions of computer programs. The commercial importance of FTP as a separate Internet service has diminished because WWW includes more versatile features to distribute files.
In the future one can expect the coming of various new services. Today, the following media are limited because of hardware and data transfer capabilities. However, systems with more capabilities will most certainly be on the market in the future.
Internet television and video-on-demand
Broadcast Internet television and video-on-demand are services that can have a huge popularity among Internet users in the future. The benefits of broadcasting via the Internet include reduced costs and lack of regulation. Moreover, the Internet makes it possible to rent videos without a physical outlet.
One scenario is that a company could establish its own global TV channel or video-on-demand service with exclusive commercials.
Internet radio
As well as video, the Internet can transmit radio programs. The benefits are similar as with video, but hardware and data transfer requirements are much lower, as well as the costs involved in creating the radio programs.
Internet phone
The Internet can be used to make phone calls all over the world.
The benefit of this is that the cost of an Internet connection
is usually lower than the cost of an international telephone call.
The next few paragraphs will look deeper into the properties of the new media, from a customer's and a marketer's point of view. Three Internet channels (WWW, email and newsgroups) will be compared to a few conventional channels, namely mail-order catalogs, direct mail, TV shopping, telephone shopping and newspaper and magazine ads.
| 1. Type | ||||||||
| Text | ||||||||
| Pictures | ||||||||
| Video | ||||||||
| Voice | ||||||||
| Principal type of medium | ||||||||
| 2. Interactivity | ||||||||
| Interactive | ||||||||
| Vendor and customer in contact simultaneously | ||||||||
| 3. The payer | ||||||||
| Recipient pays | ||||||||
| 4. Must the customer be active? | ||||||||
| Customer gets without requesting |
1. Type
Email and newsgroups are simpler media than the others in the sense that they are normally used only for textual communication. The WWW can be used to distribute text, pictures, video, and voice. In practice, however, hardware capacity limits the transmission of video and voice, and the Web consist mostly of text and pictures, like conventional catalogs, direct mail and ads.
2. Interactivity
The Internet is interactive. Of the conventional media, only telephone, door-to-door and party selling offer interactivity. Utilizing the interactivity is quite easy because email makes it possible to communicate asynchronously, yet personally.
3. The payer
Very often the user pays for the Internet connection. Thus the recipient pays to get some marketing information, as opposed to conventional direct marketing. A number of people think Internet marketing is not desirable because of this fact.
4. Must the customer be active?
Some media can be used to send messages to the customer without request. All the conventional channels can be used this way, although catalogs, for example, are often subscribed to. On the contrary, the WWW needs that the customer be active in getting the needed information.
| 1. Comparison | ||||||||||||
| Easy to compare between competitors | ||||||||||||
| Easy to share experiences with other customers | ||||||||||||
| 2. Product | ||||||||||||
| Extensive assortment | ||||||||||||
| Extensive product information | ||||||||||||
| Samples can be distributed | ||||||||||||
| 3. Time | ||||||||||||
| Immediate delivery | ||||||||||||
| Available 24h/day | ||||||||||||
| Time pressure when buying | ||||||||||||
| 4. Recipient's attitude | ||||||||||||
| Getting the information may be a social event | ||||||||||||
| Browsing for fun | ||||||||||||
| Receiving a marketing message can be particularly annoying | ||||||||||||
1. Comparison
The Internet has a couple of advantages compared with the conventional in-home shopping channels. It's often possible to compare products between competitors. Some companies even list their competitors on their WWW page and encourage comparison.
Newsgroups, in particular, give the opportunity to share experiences with other customers.
The Internet is especially useful for selling digital products, like information, records, video, computer software etc. Samples and even the genuine product can be delivered immediately to the customer.
A Web catalog is like its paper counterpart in many ways. Both can have an extensive assortment and product information. A Web catalog can be even more extensive in these aspects.
3. Time
The Internet is especially beneficial what comes to time factors. The customer can do Internet shopping at any time of day regardless of what time zone the customer is in, like conventional mail order shopping too.
As mentioned, the Internet facilitates rapid delivery too. This is because the time an order spends going from the customer to the vendor is typically a few seconds, the order can be handled automatically and the product - provided it is a digital one - can be delivered right away. However, this makes requirements for the method of payment (see 3.3.3 Perceived risk in Internet shopping today).
Customers can feel time pressure with
various shopping methods. For example, in the case of telephone
selling, the customer has to make a decision while on the phone.
Subhash, Bearden and Teel (1983) found that a hypothetical electronic
shopping system caused the users to feel pressured. This study
is discussed more in paragraph 3.4.2 Shopping medium,
and the effect of time pressure is examined in the empirical survey
of this study.
Getting marketing information may have recreational and social elements. In the case of newsgroups, for example, people can discuss their experiences along with other things, like hobbies. Or one can browse a paper or an Internet catalog for fun without a specific need.
Email and newsgroup ads are similar to TV commercials and telephone sales calls with regard to the reactions of the recipient. These methods can be intrusive and annoying. It is worth a special note that untargeted marketing by email and in newsgroups has a very bad reputation among Internet users, and many people are hostile toward it (Mehta & Sivadas 1995).
| 1. Cost | ||||||||||||||
| Cheap | ||||||||||||||
| Marginal cost per recipient near zero | ||||||||||||||
| 2. Targeting | ||||||||||||||
| Can be targeted to a certain segment level | ||||||||||||||
| Can be targeted to a certain person | ||||||||||||||
| Supports personal offers | ||||||||||||||
| Marketer knows who received the message | ||||||||||||||
1. Cost
Internet marketing is cheap in comparison to conventional channels. One can reach millions of consumers with a reasonable investment. Moreover, most of the costs are fixed, like the cost of creating some WWW pages. Once an initial investment has been made, it is relatively cheap to distribute the same message for even millions of people. This is not the case with sending paper catalogs, for example.
2. Targeting
Internet marketing can be targeted just like traditional direct marketing, even better. One can contact a target segment via a special newsgroup. One can monitor a consumer's behaviour to collect a variety of data for use with personalized campaigns. The gathering of information can be automatic. For example, a WWW store can register which products a customer has been interested in but not yet ordered, and display a customized offer for every visitor.
A number of studies have investigated the in-home shopper. The concept of in-home shopping varies between studies - some studies deal with a certain mode of in-home shopping, like catalog shopping, while others speak of in-home shopping in general - but a number of consistent statements can be found.
Socio-economic group
The in-home shopper belongs to a higher-than-average socio-economic group, measured by education, income and occupational status of the head of the household (Gillett 1970 & 1976, Cunningham & Cunningham 1973, Reynolds 1974, Lumpkin & Hawes 1985). This tendency is especially strong among those families who patronize a number of different in-home shopping channels (Gillett 1976).
There are differences between various in-shopping modes. Peters and Ford (1974, ref. Gillett 1976) found that heavy buyers of cosmetics from door-to-door sellers belong to a lower-than-average socio-economic group. Gillett (1970) found that telephone and direct-mail shoppers had higher family income, education level and occupational class than did catalog shoppers.
Attitudes
What comes to attitudes, the in-home shopper is less conservative than the non-shopper (Darian 1987). The in-home shopper is more willing to take risks and try new things (Reynolds 1974, Darian 1987). The in-home shopper is also more cosmopolitan than the non-shopper (Cunningham & Cunningham 1973).
Age
An interesting factor is age. Reynolds (1974) reported that mail order catalog shoppers tend to be younger than non-shoppers. However, Lumpkin & Hawes (1985) found they are older, and Darian (1987) found that the most likely in-home shopper group is aged 40-49 years. The reason for this can be that those who had learned a new shopping method in the 1960s and 1970s, when they were younger, continued to use it in the 1980s.
Differences between various in-home shopping methods exist. Gillett (1970) found that catalog shoppers were younger than telephone or direct mail shoppers.
A great number of studies have been conducted to find out what kind of people Internet users are. Methodologies vary. The most common method is to conduct the study on a WWW page. The principal disadvange of WWW surveys is that the respondents do not reflect the average Internet user, but, typically, the sample is biased: the respondents are likely to be active WWW users (see Appendix B). Some other surveys (like Nielsen 1995 and 1996), however, rely on conventional probabilistic sampling and conduct the survey using conventional methods, like telephone interviews.
Age and gender
A typical Internet user is a male aged from 20 to 40 (GVU 1994-1996, Nielsen 1995, Rissa & Järvinen 1996). The average age is a few years above 30. One reason for the low average age this is that the Internet started as a university network, and today, a lot of students are introduced to the Internet in universities.
Internet used to be dominated by males. For example, GVU's 1st WWW survey (1994a) found that 95% of the respondents were males. However, recent studies have shown that the percentage of females is rising rapidly. GVU's 5th survey (1996) reports 32% females.
Socio-economic group
As stated above, the Internet started as a university network. Supposedly, this is the reason that Internet users have a higher-than-average education level (Nielsen 1995, Gupta, Pitkow & Recker 1995, GVU 1994-1996), although the level is decreasing.
Internet users also have a high income, even though there are a large number of users with a low income, presumably students (Gupta, Pitkow & Recker 1995, GVU 1994-1996, Nielsen 1995, Rissa & Järvinen 1996). The average income is decreasing.
Experience
About 1/4 to 1/3 of active Internet users have a computer-related work (Gupta, Pitkow & Recker 1995, GVU 1996, Nielsen 1996). The percentage will supposedly decrease.
Measured by income and education, the Internet user is very much like the in-home shopper.
Moreover, it is interesting to note that males under 40 years are very likely to use in-home shopping channels, more likely than females in the same age group (Darian 1987). This leads to a hypothesis that many Internet users could very well be active in-home shoppers too.
Those people who adopt a new innovation first are called innovators, and those who follow early adopters. These groups have very similar characteristics; according to Rogers they form 16% of the market (Rogers & Shoemaker 1971, ref. Wilkie 1994, 337). In developed countries like the United States or Finland, market penetration of the Internet has just exceeded this limit (Nielsen 1996, Taloustutkimus 1996b, Yankelovich Partners 1996). The next adopter group is the early majority, and this group is now coming to the Internet.
Rogers (1962, ref. Kotler 1991, 343) has presented that earlier adopters tend to be younger and have higher social and financial status than later adopters. Earlier adopters also have a different type of mental ability, they utilize more impersonal and cosmopolite information sources, and are in close contact with the origin of new ideas. They also have more cosmopolitan social relationships. Robertson & Kennedy (1968, ref. Urbany & Talarzyk 1983) also stated that innovators are risk takers and socially mobile.
The Internet has experienced a rapid expansion in the 1990s, and it has the properties of an innovation. Not so many years ago, when the Internet was not yet so popular, it was more like a universitity network. This caused the average user to have a rather high level of education the adopters were in close contact with the origin of the new innovation. For example, GVU (1994b) found that 14% of the respondents had a doctoral degree, but after one year, only 4% had it (GVU 1995). Furthermore, the Internet was occupied by technically oriented males, and only the introduction of the WWW has brought the Internet to the masses.
The profile of an earlier adopter is near the profile of the Internet user, at least what comes to age and socio-economic status. What is more, the Internet is helping its users to be more cosmopolitan and socially mobile.
In-home shopping literature presents a number of benefits the in-home shopper is experiencing (Gillett 1970 & 1976, Darian 1987):
It is natural to think all the above benefits may well apply to Internet shopping too.
The Internet offers some unique convenience. GVU (1994b) found 3 convenience related factors that attained a high value in the eyes of consumers: ease of contacting, placing orders, and customization.
Because supply is not so much bound to geographical areas, and because customization is easier with certain products, product assortment will increase. Increased competition, shortened distribution channels and other cost savings may lead to lower prices (Benjamin & Wigand 1995, O'Connor & O'Keefe 1996).
The special features of Internet, like affordable targeted marketing and relative cheapness of marketing efforts, facilitate promoting unique products for various market niches. Moreover, the Internet is very useful for distributing exact product information, like technical data etc.
Literature suggest other benefits too. The Internet facilitates comparison shopping and speeds up the finding of items (Wallace, 1995). An important product category are digital products, which can be distributed on-line very quickly (see Consumer's point of view).
Consumers may also find Internet shopping recreational (Hoffman & Novak 1995). GVU (1994b) found that quality is most important to respondents in choosing a WWW vendor, and that WWW vendors are perceived positively on this characteristic.
In-home shopping has some disadvantages in relation to in-store shopping. A major disadvantage of in-home shopping is the risk perceived by the consumer. Another disadvantage is the lack of the benefits of in-store shopping, namely lack of diversion from routines of everyday life, sensory stimulation, social experiences, and comparison of goods (Darian 1987).
Past research indicates that the most important type of risk is inability to physically inspect the product (Cox & Rich 1964, ref. Darian 1987; Spence, Engel & Blackwell 1970, ref. Darian 1987; Jasper & Ouellette 1994). Other types of risks are inability to make comparisons of product characteristics (Darian 1987), wrong product information and delivery delays (Euromonitor 1987).
Perceived risk is not directly related to shopping behaviour. In-home shoppers are more willing to take risks than non-shoppers (Darian 1987). Risk can also be reduced.
To reduce the perceived risk due to inability to physically inspect the product, the vendor can do a number of things. Product samples can be distributed and money-back guarantees given, as well as a free return postage. Well-known brand names are a good way to ensure quality in the mind of the consumer. One way to promote brands is to use dual channels of distribution, namely offering the product both in-store and in-home so that more consumers can become familiar with the merchandise (Jasper & Ouellette 1994).
The capabilities of the Internet may help to reduce the amount of perceived risk. Certain digital product categories are suitable for on-line trial, like music, video and computer programs. The Internet being a global marketplace, the importance of globally recognized brand names may increase.
GVU (1994b) found that WWW users rate vendor reliability and security of their financial information quite highly. Importantly, these were the characteristics Web vendors are perceived as being at a competitive disadvantage.
Entry barriers for an Internet vendor are very low, so the market can attract many kinds of vendors. And because the Internet is global, vendors can be located anywhere in the world, in a country with less consumer rights, for example.
Another hindrance is the lack of a globally accepted method of payment. A number of electronic methods are being introduced and tested.
The development and adoption of new payment systems is one of the interesting things that can have an impact on Internet commerce in the coming few years.
Before the Internet was popular, there were a number of other electronic marketplaces. A few studies have been published on these electronic shopping channels, some of which were only experimental.
In 1988, Shim and Drake carried out a research (Shim & Drake 1990) on what kind of consumers would be willing to use new electronic shopping channels. At that time the Internet was not widespread, and the then electronic shopping channel was mostly videotex, an example of which is CompuServe The Electronic Mall in the USA. Because electronic shopping was not commonplace, the research sought for potential users of electronic shopping methods.
It turned out that the most likely electronic shoppers were those who had been active in-home shoppers. Potential electronic shoppers were also younger and they often used a computer. It seems that experience with computers and conventional in-home shopping are encouraging people to use new electronic shopping methods, especially young people who are more willing to try new things.
It is disputable if these results can be generalized. Korgaonkar and Smith (1986, ref. Darian 1987) found that the desire for electronic in-home shopping and banking services was not associated with younger age, nor was it associated with higher income or education like in conventional in-home shopping.
Subhash, Bearden and Teel (1983) studied what kind of differences there are between using 1) an electronic product catalog and 2) a conventional paper catalog. A hypothetical electronic, text-only catalog was built for the test.
There was no difference in perceived financial risk and adequacy of product information between the two channels. This was found despite the fact that there were no pictures in the electronic catalog, even though the paper catalog had them. This implies that an electronic catalog may be effective as a marketing medium.
The electronic catalog seemed to increase time pressure on the consumer. The study suggested that because the computer demands an immediate yes/no answer to a buying decision, this leads to perceived time pressure. The consumer could be afraid of landing in impuse buying. - Internet use is often charged based on time spent on-line. This can increase the time pressure perceived when shopping on the Internet.
As stated above, Internet shopping is very much like conventional in-home shopping. One could easily think that Internet shopping would be affected by the same factors as conventional in-home shopping. However, the rapid development of the Internet and the profile of the Internet user lead to assume that there are some other things that explain Internet shopping behaviour too.
I suggest that the following things are especially significant when explaining Internet shopping behaviour (in no special order).
1. Experience in conventional in-home shopping methods
2. Use of the Internet (experience, frequency of use, duration of session)
These suggestions are based on the the above mentioned videotex study (Shim & Drake 1990). It feels natural that active use of in-home shopping and the Internet would imply higher Internet shopping probability, because the shopping method would not carry so high a risk to the experienced user. Moreover, Internet shopping should be more convenient for someone who already spends much time on-line and knows the system.
3. Work in the computer branch
Working in the computer branch will often give better possibilities to be connected to the Internet, thus causing more use of it. They are most certainly experienced computer users, which reduces the risk perceived using new technology. Computer professional will also be more likely to find something to buy on the Internet, because much of the supply is still related to computers.
4. Convenience factors
As mentioned before, the Internet offers some convenience not found elsewhere.
5. Uniqueness of products sold on the Internet
Another benefit of the Internet is the ease of marketing unique products for various niches.
6. Perceived risk
Internet shopping has new types of risks, like risks with method of payment, and risk of not knowing the vendor, which can be located on the other side of the world.
7. Desire to experiment
Because the Internet is a new medium, it may be that the first
transactions one makes are a sort of experiments. People who are
risk takers are more probable to experiment with new things.
Based on the suggestions mentioned above, I have created eight hypotheses that will be studied more thoroughly. These hypotheses compare the Internet shopper to the non-shopper.
H1: Internet shoppers are more active as in-home shoppers
H2: Internet shoppers have more Internet experience
H3: Internet shoppers use the Internet more frequently
H4: Internet shoppers have longer Internet sessions
H5: Internet shoppers are more likely to work in the computer
branch
H6: Internet shoppers are more likely to use the Internet in
their work
H7: Internet shoppers have found unique products on the Internet
H8: Internet shoppers perceive less risk in Internet shopping
than non-shoppers
To study the hypotheses, an empirical WWW survey was run during September and October 1996 in VBShop (http://www.aivosto.com/vbshop.html). VBShop is a WWW site that offers a number of specialized utility programs for computer programmers who program in Visual Basic for Microsoft Windows.
The questionnaire is in Appendix A. During a 3-week period, 387 people from all over the world responded.
It is reasonable to assume that almost all of the respondents are computer programmers because the site is of interest to programmers only. Results of past surveys done with the same method indicate that too (Salste 1996). Past surveys also indicate that most of the respondents are experienced Internet users.
In this survey, 96% of the respondents were male. The average age was 35 years. 79% worked in the computer branch. 46% used the Internet in their work.
Active in-home shoppers
Those who had shopped at home using conventional mail, phone or fax order at least once in the past 3 months were classified as active in-home shoppers. 53% of the respondents belong to this group.
Internet shoppers
In this survey, those who had shopped on the Internet at least once were classified as Internet shoppers.
52% of the respondents belong to this group. This percentage is considerably higher than those in other surveys (Nielsen 1995, Rissa & Järvinen 1996, GVU 1996). This probably reflects the fact that the respondents are Internet and computer experts.
87% of those who had ever shopped on the Internet had shopped there in the past 3 months. This can reflect two things: 1) The Internet is so new that users are just beginning to shop there. 2) The users are happy with Internet shopping and continue to use it. More research is needed in this area.
Active Internet shoppers (work and home)
Those who had shopped on the Internet in the past 3 months for home use were classified as active Internet in-home shoppers. 30% of the respondents belong to this group.
Those who had shopped on the Internet in the past 3 months for home use were classified as active Internet at-work shoppers. 42% of the respondents belong to this group.
The reason for the greater percentage of active work shoppers can be that the respondents were professional programmers. Currently, a large proportion of the Internet market consists of computer related goods.
Hypotheses H1…H8 were statistically tested. Hypotheses that did not pass the 95% significance test were rejected.
H1 was confirmed (p>99.9%). Indeed, Internet shoppers were far more likely to be active in-home shoppers. The result is clear; Internet shopping and conventional in-home shopping behaviour are strongly related.
H2 was confirmed (p>99.9%). It feels quite natural that experienced Internet users would be more likely to have already shopped.
H3 was confirmed (p>95%). It seems that frequency of Internet use is positively related to Internet shopping. Everyday users were significantly more likely to have shopped on the Internet.
There was no correlation between frequency of use and time pressure when using the Internet.
H4 was confirmed (p>99%). There was a significant difference between shoppers and non-shoppers in how long their average Internet session is. One hour seems to be the dividing session length.
There was a slightly negative correlation between session length and time pressure when using the Internet (significance level 94%). Users who had shorter sessions felt slightly more time pressure.
H5 was confirmed (p>95%). Internet shoppers are indeed more likely to work in the computer branch. Next, two additional hypotheses were formed.
H5a: Active Internet home shoppers are more likely to work in the computer branch
This hypothesis did not pass the significance test (90%<p<95%). More research is needed.
H5b: Active Internet work shoppers are more likely to work in the computer branch
This hypothesis was confimed (p>99.9%).
H6 was confirmed (p>95%). Internet shoppers are indeed more likely to use the Internet in their work. As with H5, two additional hypotheses were formed.
H6a: Active Internet home shoppers are more likely to work with the Internet
This hypothesis was rejected. Internet professionals were no more likely to shop for home use.
H6b: Active Internet work shoppers are more likely to work with the Internet
This hypothesis was confimed (p>99.9%).
H7 was confirmed (p>99%). Internet shoppers agreed more with the statement that they had found unique products that interested them.
Non-shoppers had found interesting products too but they had not bought them for some reason. One of the reasons could be a high risk perceived in shopping from the Internet.
Three questions measured the risk perceived in Internet shopping: It's risky to buy without seeing the product first, There is no secure enough way to pay on-line, and It can take a long time before I get the product I ordered.
1. It's risky to buy without seeing the product first
Non-shoppers perceived more risk due to inability to see the product.
This is consistent with other findings in in-home shopping.
2. There is no secure enough way to pay on-line
Non-shoppers were more likely to agree with this statement.
3. It can take a long time before I get the product I ordered
Non-shoppers were more likely to agree with this statement.
Conclusion
Thus it seems that all of the risk factors in this survey contributed negatively to Internet shopping. H8 was confirmed.
The respondents of this survey were computer programmers. This means they are familiar with computers, and relatively familiar with the Internet too. Moreover, almost all of them were males.
Less experienced users may well feel the Internet differently. For example, they may feel a time pressure caused by the computer (as Subhash, Bearden and Teel reported in 1983), thus reducing the likelihood of Internet shopping. According to the survey at hand, time pressure could have a slight positive impact on Internet shopping. However, the source of this time pressure was not detected.
Read more about the limitations in Appendix B: Limitations of WWW surveys.
The final results of the survey were:
| Factor | Impact on Internet shopping |
| In-home shopping experience | + |
| Internet experience | + |
| Activity of Internet use | + |
| Product uniqueness | + |
| Risk due to inability to inspect the product, payment method and slowness of buying | - |
As a conclusion, Internet shopping is very much similar to conventional non-store shopping. Internet shopping is a new method of non-store shopping.
There is a network of factors explaining Internet shopping in Figure 7. The network is rather complicated to include all factors present in this study.
Relationships between factors are marked either negative or positive if applicable, or neither of them if the relationship is more complex. Tested relationships are printed in bold, concluded relationships (not tested) with normal arrows, and relationships that are hypothesized but either not supported by the survey or not surveyed at all are printed with a dotted arrow.
Factors printed in a box with wide borders were included in the analysis of the survey results.
From the network we can conclude, for example, that Internet experience is positively related with frequency of use, which in turn leads to increased probability of Internet shopping.
Factors a marketer can influence
A marketer can influence in a number of factors in the network.
These are price, product assortment, product uniqueness, convenience
factors, perceived slowness, and risk due to method of payment.
The next chapter Implications for marketing describes how
to use these opportunities in marketing.
The results have a number of implications for marketers.
The Internet is a new non-store shopping medium
One of the major findings of the survey is the strong relationship between conventional in-home shopping and Internet shopping. This implies that the Internet is a new alternative for conventional non-store shopping media.
This also implies that active in-home shoppers are a good target segment for Internet marketers. As explained, in-home shoppers belong to the upper socio-economic class. Among other things, they are seeking product assortment and convenience. They have a relatively high education and occupational level, thus they are likely to have an Internet connection too.
Target at active Internet users
It is clear from the results of the survey that active Internet users are more likely Internet shoppers. The number of active users is most probably going to increase because gaining experience increases both use frequency and session duration, and the number of users is increasing rapidly too.
Offer uniqueness
The Internet is a perfect channel for niche products. Rather than fighting for shelf space in a physical store, offer the product in your virtual store. Remember, Internet users have a high income level, they are likely to be attracted by the special qualities of your product.
Target in the near future
New, secure payment methods can be expected in the future. This will reduce the perceived risk in Internet shopping and eliminate one of the inhibitors of Internet commerce. It will also increase the conveniece of this new shopping mode.
As the Internet continues to grow, new consumer segments enter the market. One of the fastest growing user groups are women (GVU 1994-1996).
The Internet is a constantly evolving network, and information about it becomes obsolete fast. Moreover, a lot of questions have not been answered yet.
Are they going to shop again?
Are Internet shoppers happy with their experiences with the shopping method? Are they going to use it again? Do they prefer it to other shopping methods? In which ways?
How about inexperienced users?
The respondents of this survey were experienced computer and Internet users. How do inexperienced users think about Internet shopping? What about people who don't shop at home, is there any chance they could become Internet shoppers?
What is the Internet shopper really like?
Do Internet shoppers belong to a higher-than-average socio-economic class as the literature implies in-home shoppers do? How are they different from non-shoppers demographically? Psychographically? What kind of conveniece do they see with this new shopping method? What are the worst risks?
What is the effect of time pressure?
There are two kinds of time pressure: pressure specific to the Internet, and other time pressure (not examined in this study). What causes time pressure? Is it the computer, is it the cost of connection, is it the time that is "wasted" on-line? According to the survey, users do feel somewhat pressured. Is the impact of time pressure negative - no time to buy even if wanted? According to this WWW survey, the impact may be, however, slightly positive.
What is the effect of new technology?
New applications will almost inevitably emerge. What is the next big step after the WWW? Internet television? Virtual reality? Will the Internet become obsolete with the introduction of a new solution?
The Internet is a collection of new media that have a huge potential to be a popular channel for electronic commerce. Internet shopping resembles conventional in-home shopping in that the customer makes transactions without physically visiting a store.
The Internet provides means to distribute marketing information in ways similar to conventional direct marketing, like catalogs and targeted direct mail. The Internet has other appealing properties too. It is cheap, and it helps personal communication between the seller and the buyer. It is useful for distributing digital products, and it can shorten the time between purchase decision and delivery. It also gives the customer certain extra benefits, like the ability to compare and discuss products.
The average Internet user is quite similar to the average in-home shopper, at least what comes to socio-economic group: both are well educated, have a high income and occupational position. Moreover, the average Internet user is a male aged 20-40 years, which is a very active group at in-home shopping.
In-home shopping has a number of benefits to the customer: it is convenient, it offers product assortment and uniqueness, a geographically larger shopping area, and often a better price too. All these benefits apply also to Internet shopping. The major disadvantage of in-home shopping, namely risk - due to inability to see the product and make comparisons, and to delivery delays - applies to Internet shopping too.
In an empirical WWW survey, the following factors were found to increase the likelihood to shop on the Internet: previous activity in in-home shopping, computer or Internet related work, Internet experience, active Internet use, and product uniqueness. Risk due to inability to inspect the product, payment method and slowness of buying were found to decrease the likelihood to shop.
The conclusion of the study is that Internet shopping is a new
method of non-store shopping.
The purpose of this survey is to find out
Your opinion is equally important have you either done Internet
shopping or not.
1. How long have you used the Internet?
Less than 3 months
3-12 months
1-2 years
More than 2 years
2. How often do you use the Internet?
Less than once a week
Once or twice a week
Several times a week
Every day
3. How long is your typical Internet session?
Less than 15 minutes
15-60 minutes
1-2 hours
More than 2 hours
I'm on-line all day
** Conventional in-home shopping
4. During the last 3 months, have you placed any conventional mail, phone or fax order for home use?
Yes/No/Don't know
** Shopping on the Internet
5. Have you ever ordered or bought something on the Internet?
Answer Yes if you either
A. ordered something by email, WWW or another Internet means, or
B. saw something on the Internet and ordered it by phone, fax
or regular mail, without attending to a store
6. During the last 3 months, have you ordered something on the Internet for home use?
Yes/No/Don't know
7. During the last 3 months, have you ordered something on the Internet for work use?
Yes/No/Don't know
8. Do you agree with the following statements about Internet
shopping?
Scale:
1=disagree completely
2=disagree somewhat
3=neutral
4=agree somewhat
5=agree completely
9. Do you work in the computer branch?
Yes/No
10. Do you work with the Internet?
Yes/No
11. Your age:
12. Your sex: Female/Male
WWW Surveys are limited. The following text is from GVU
(1995).
Basically, the survey suffers two problems: sampling and self-selection.
1. Essentially, when people decide to participate in a survey, they select themselves. This decision may reflect some systematic selecting principle (or judgment) that effects the collected data.
2. The other issue is sampling. There are essentially two types of sampling: random and non-probabilistic. Random selection is intended to ensure equal representation among populations. To accomplish this, steps need to be taken to get respondents in a random manner, e.g., drawing numbers out of a hat. This survey uses non-probabilistic sampling, which does not use randomization techniques to get respondents. This reduces the ability of the gathered data to generalize to the entire user population.
There are other limitations too. GVU (1996) found that over 1/4 of the respondents replied that they have at least once provided false demographic information when registering with Web sites. Males were more likely to provide false information, as well as younger versus older people, both of which were intensely present in the WWW survey of this study.
This suggests that information collected on WWW pages should be interpreted conservatively. Efforts must be taken to explain the purpose of the survey to the respondent, and to tell how the information will be used.
VBShop was a website that evolved into Aivosto Oy. History of Aivosto