Main article

Warning View the most recent version.

Archived Content

Information identified as archived is provided for reference, research or recordkeeping purposes. It is not subject to the Government of Canada Web Standards and has not been altered or updated since it was archived. Please "contact us" to request a format other than those available.

  1. Introduction
  2. Measuring e-commerce in Canada
  3. Trends in Internet shopping
  4. Patterns in Internet shopping
  5. Summary
  6. References

1   Introduction

Before the Internet was launched commercially in 1993, few people outside of the scientific and academic worlds knew anything about this new technology. Commerce has since changed in unimaginable ways and it is now possible to search, purchase and sell just about anything over the Internet (Rowland, 2006). However, some of the expected impacts of electronic commerce (e-commerce), such as the potential detrimental effects on traditional retail, have not yet come to fruition (Sciadas, 2006). It appears that the impact of the Internet on retail shopping depends on the specific product category as well as other factors such as consumer behaviour and concern for online security.

This paper reviews Internet shopping data, trends and patterns in Canada from 2001 to 2007. It begins with a brief description of e-commerce measurement at Statistics Canada, followed by the number and value of Internet orders made from 2001 to 2007. Online purchases by type of product are then examined to illustrate how the Internet has become both a complement to, and a substitute for, traditional retail. The relative concentration of online shopping among certain Canadians is also examined, while the final section identifies the determinants of online shopping during 2007 and the extent to which concern for online credit card security remains an impediment.

2   Measuring e-commerce in Canada

Statistics Canada has been monitoring e-commerce activity in the Canadian economy since 1999, with surveys of both consumer purchases and business sales (see Note to Readers). As depicted in Chart 1, when a new technology diffuses over time, the statistical instruments used to measure its adoption and use must also evolve. In the early stages of an innovation, measurement is focused on the readiness of the market to adopt the new technology. The Household Internet Use Survey (HIUS) was launched by Statistics Canada in 1997 to assess the readiness of households to use the Internet and the possible barriers to their use.

As the adoption and applications of a new technology become more widespread, measurement interest shifts to the types and intensity of uses. In 1999, the HIUS added questions on the use of the Internet for household shopping. With more and more households connected to and using the Internet, the HIUS was replaced in 2005 by the Canadian Internet Use Survey (CIUS), which focused more on individual use, including shopping. When technological adoption approaches saturation the focus shifts again, from adoption and use to the outcomes and impacts on society. Measuring impacts is a challenge for any innovation, but perhaps even more so with commercial applications of the Internet since "the rate of development and innovation on the Web is actually getting faster and faster all the time". 1 

Yet another ongoing consideration in developing measurement tools and methods for Internet use has been international comparability. For policy makers in particular, international benchmarking can be a critical policy gauge. 2  Statistics Canada's Internet use surveys (see Note to Readers), were designed to conform to the Organisation for Economic Co-operation and Development's (OECD) model surveys. Although, international comparability presents its own challenges (for example, see McKeown and Veenhof, 2009), the potential to compare Canadian data with other countries remains a consideration for the Next Internet Use Survey (NIUS). 3 

Note to Readers: Data sources and definitions

Data sources

Beginning in 1999, the Household Internet Use Survey (HIUS), conducted annually from 1997 to 2003, asked if anyone in the household used the Internet from home as part of their buying process, either to research product characteristics or to place orders during the previous 12 months. About 1.8 million households in 1999 had at least one member who used the Internet from home as part of their buying process (Ellison, Earl and Ogg, 2001). Questions were also asked about the number and value of online orders.

Beginning in 2001, these e-commerce questions were extended to include Internet use from any location (home, work, school, library or other).

In 2005, the renamed and redesigned Canadian Internet Use Survey (CIUS) asked individuals aged 18 years and over about their Internet use. All users, from any location, were asked about electronic shopping, including the number and value of orders. In 2007, the target population was expanded to include individuals 16 and 17 years of age. While more than 97% of 16 and 17 year olds used Internet, just 25% reported making online orders accounting for about 2% of the number of orders and 1% of their value.

While the online behaviours of households (HIUS) and individuals (CIUS) should not be compared directly, this study examines the aggregate e-commerce estimates from these surveys.

Business e-commerce

The 1999 Survey of Electronic Commerce and Technology (SECT) estimated $4.4 billion worth of Internet sales by private and public sector enterprises in Canada (Peterson, 2001). By 2007, online sales – business-to-business (B2B) plus business-to-consumer (B2C) – had reached $62.7 billion. The CIUS and SECT estimates differ as the former measures only consumer purchases, including online purchases from international Websites. SECT measures both the B2B and B2C sales of Canadian enterprises, including to individuals and organizations outside of Canada.

E-commerce defined

To date, Canada has used the OECD's narrow definition of e-commerce whereby an Internet transaction "is the sale or purchase of goods or services, whether between businesses, households, individuals, governments, and other public or private organizations, conducted over the Internet". The goods or services are ordered over the Internet, but the payment and the ultimate delivery of the good or service may be conducted on or off-line. This 'narrow' definition excludes orders received or placed by telephone, facsimile or other computer-mediated networks such as EDI.

An OECD Expert Working Group is proposing an update to the definition of e-commerce which would drop the network-based distinction between the broad and narrow definitions. The group suggests that rapid technological development complicates the ability of respondents, businesses in particular, to distinguish between sales carried out through different types of computer-mediated applications. Such a change will not have an impact on e-commerce concepts or data from either the 2009 CIUS or the redesigned 2010 CIUS.

3   Trends in Internet shopping

The examination of Internet shopping in Canada begins in 2001 when 13.4 million online orders were made by over 2.2 million households (Chart 2). By 2007, the Internet was used to make almost 70 million orders by about 8.4 million Canadian individuals aged 16 years and older. 4  While growth in the number of orders has been increasing, it appears to have slowed between 2005 and 2007. The 2009 CIUS will provide another data point to determine whether this declining growth rate signals a trend or is an aberration of sorts. 5 

The value of online orders made by Canadians in 2007 was $12.8 billion (Chart 3). Although increasing, growth in the value of online orders has slowed, particularly between the 2003 to 2005 period and 2005 to 2007 period. In addition to the change from a household survey to one of individuals in 2005, several factors may have influenced e-commerce growth including currency exchange rates, business strategies, and consumer behaviour.

The exchange rate between the Canadian and American dollars can influence the cross-border shopping behaviour of Canadians, both on and off-line. According to the 2007 CIUS, for every $100 spent online, $56 was to Canadian vendors, down from $63 in 2005. 6  Changing business models can also reduce the online shopping options available to consumers. In retailing for example, many companies continue to alter their business plans in an attempt to find the right online functionality. Finally, there has been increased media coverage of data breaches by retailers which may affect online shopping behaviour and consumer confidence in online ordering and payment. In some cases, these concerns may be strong enough to act as a barrier to e-commerce. 7 

The growth in online sales between 2005 and 2007 was driven by a combination of more online consumers and existing Internet shoppers buying more. The number of online consumers aged 18 years and older grew by 18% over the two years. And in 2007, these online consumers reported an average of 8.4 orders worth $1,554, up from 7.2 and $1,150 in 2005, a 16% and 35% increase respectively.

Despite this growth, online sales continue to represent just a fraction of total consumer expenditure on goods and services. To illustrate, total retail sales in Canada were about $412 billion in 2007. The next section examines the pattern of online sales by product category which suggests that the impact of e-commerce has been complementary to traditional retail. This next section will also examine the continued concentration in online sales and some of the factors that influence this shopping behaviour.

4   Patterns in Internet shopping

4.1  Complement and substitute

In a study of the expected outcomes of digital technologies, Sciadas (2006) finds that many have not yet materialized. For instance, e-commerce sales do not seem to justify initial fears of the negative consequences on traditional retail. However, it is apparent that the Internet has changed consumer behaviour fundamentally, perhaps in ways not always anticipated.

The Internet is becoming a substitute for some types of products, particularly those which can be delivered easily, either on or off-line. For instance, almost one-half (45%) of online consumers in Canada reported making travel arrangements over the Internet during 2007 (Table 1) while more than one-third (37%) purchased books or magazines, some of which can be delivered online or inexpensively by courier. Indeed, for travel services and 'other entertainment products', such as concert or sports tickets, the Internet has altered the business model of suppliers as well as the behaviours of consumers.

For other types of products, particularly those which cannot be delivered at the time of purchase, the Internet has become complementary to traditional shopping behaviour (Table 2). In 2007, the Internet was used to browse for consumer electronics (44% of window shoppers) as well as furniture and large appliances (43%). Such browsing was often followed by a trip to a retail location. Of the more than 11 million "window shoppers" in 2007, almost two-thirds (64%) reported making a subsequent purchase directly from a store. Grau (2008) reports that Canadians, while avid online product researchers on par with Americans, are much more likely to make a subsequent purchase in-store rather than online.

An indicative product category is clothing, jewellery and accessories, which ranked fourth in prevalence for both orders and window shopping in 2007 (Tables 1 and 2). It appears that some retailers in this category have moved almost seamlessly from catalogue to Internet-based sales while, for others, it is not always clear whether to extend the Website beyond product information to a transactional capability. Some retailers are altering their business plans in an attempt to find the right online functionality. 8 

Demographics can also explain some of the variations in product categories. In 2007, for example, middle-aged individuals were more likely to purchase furniture and appliances (12% of those aged 30 to 45 years purchased house wares online compared to 6% of those under age 30). Conversely, younger Canadians were more inclined to purchase clothing, jewellery and accessories over the Internet (37% of those under 30 years of age compared to 28% of those aged 30 to 45).

4.2  Concentration

Despite substantial growth in e-commerce over the last few years, not all Canadians have participated equally. In 2007, more than 8.4 million Canadians aged 16 and over (or 32%) made, on average, just over 8 orders valued at $1,520. However, a relatively small group of online consumers in Canada accounted for the bulk of this expenditure. Charts 4 and 5 divide online shoppers into expenditure quartiles, or four equal groups, based on the total value of their orders. 9 

The biggest online spenders, those in the top 25%, racked up more than $1,500 in expenditures during 2007, averaging almost three times the number of orders and more than ten times the total expenditure than other online consumers. Indeed, these top online spenders, just 8% of Canadians and 25% of online shoppers in 2007, accounted for nearly half (46%, Chart 4) of the online orders and more than three-quarters (78%, Chart 5) of their value. These shares have increased from 2005, when the top 25% of online spenders accounted for 44% of orders and 76% of their value (McKeown and Underhill, 2007).

To a large extent, recent growth in online sales was driven by the top online consumers buying more (from 12.8 orders worth $3,056 in 2005 to 15.7 orders worth $4,576 in 2007). Who were these top online spenders? As was the case in 2005, the top online spenders in 2007 had higher levels of household income and education than did other online shoppers. 10  They also have a greater breadth and experience online with more than 90% having used the Internet for at least 5 years and for an average of 15 online activities in 2007. With this level of 'online comfort', 85% of the top spenders reported using a credit or bank card online during 2007 for some or all of their purchases.

What were these top online spenders buying? In 2007, they were more likely to report purchasing all types of products than other online consumers (data not shown). Two types of purchases in particular were prominent among the top spenders: More than three-quarters reported making online travel arrangements and one-quarter reported buying computer hardware. A previous study found that online travel services, because of frequency, and online computer hardware, because of the expense, tend to generate higher online expenditures (McKeown and Underhill, 2007).

4.3  Determinants

This section compares the socio-economic characteristics and online behaviours of those Canadians who used the Internet to make electronic orders in 2007 with those users who did not. Tables 3 and 4 list the variables of interest along with their expected influence on Internet shopping propensity (see Note to Readers: Multivariate logistic model). A multivariate model based on previous work (Noce and McKeown, 2008) is used to disentangle the statistical influence of these variables. The model excludes those Canadians who used the Internet only from locations outside of home. 11 

Note to Readers: Multivariate logistic model

In the model, the dependent variable takes a value of 1 for an Internet user who reported making a purchase online and 0 for those who did not. Some independent variables are entered as continuous and others as discrete predictors (see Appendix A). A control model included those factors such as age, income and education (x1, x2 and x4 to x6) found to be important for Internet use (Noce and McKeown, 2008). The full model added breadth of use (x3), speed of connection (x7) and years online (x8) as well as security concerns (x9).

Age and income (x1 and x2) were entered into the model as continuous control variables. However, the effect of age is not linear since both younger and older Canadians account for a lower proportion of online orders than the average. Household income was positively skewed and transformed logarithmically to base 10. Although logistic regression does not rely on the same distributional assumptions as Ordinary Least Squares, the solution is more stable if the predictors have a multivariate normal distribution (UCLA, 2004). The breadth of use, measured by the number of online activities (x3), had a coefficient of 0.207. In other words, the odds of purchasing online increased by just over 20% with each additional activity reported (e.g. email, online banking, view news or sports, etc.).

Chart 6 presents odds ratios to assess the effect of discrete variables (x4 to x9) on the propensity to shop online. For example, controlling for other variables the odds of an individual with at least some post-secondary education buying online were more than one and a half times greater (1.53) than someone with less education. The effect of a high-speed connection from home is small; those with high speed had slightly higher odds of making an online purchase. A study of Internet use and broadband found that the speed of connection alone does not explain differences in the sophistication of use (Middleton and Ellison, 2008). 12 

Interestingly, men had slightly lower odds of purchasing online compared to women, controlling for other factors. Canadians residing in urban locations also had slightly lower odds in 2007 of making an online purchase than did individuals residing in rural areas and smaller towns, controlling for other factors. Perhaps the lower odds of an urban 'home user' ordering online reflects more options for purchasing at a retail store location in urban areas.

According to Zamaria and Fletcher (2008), Canadians are more worried about the security of financial information online than are residents of the United States (U.S.) and other selected countries. Grau (2008) also finds that concern over the security of credit card payments holds Canadians back from buying more online. In the model, Canadians who were very concerned about online credit card use had only about three-quarters (0.76) the odds of making an online order than those who were either somewhat or not concerned, controlling for other factors (Chart 6).

This finding may also be acting as a proxy for individual 'online comfort.' Previous research suggests that online transactions require a certain 'self-efficacy' or belief in one's ability to accomplish certain tasks online (Underhill and Ladds, 2007). 13  Eastin and LaRose (2000) find that up to two years' online experience may be required to achieve a requisite level of Internet self-efficacy. Indeed, the model finds that the odds of making an online purchase for someone who had been using the Internet for at least two years in 2007 were more than one and a half (1.57) greater than someone with less experience (Chart 6).

5   Summary

Despite substantial growth in online sales from 2001 to 2007, Internet shopping continued to be concentrated along particular product categories and among relatively few Canadians. However, along with increasing product availability, the Internet will continue to have a unique advantage as a channel for commerce, particularly for serving niche markets such as specialty food, rare cars and antiques, books and instructions, and vacation spots (Rowland, 2006). Such markets may be substantive but geographically dispersed.

In 2007, the top 25% of online consumers, just 8% of Canadians, accounted for almost half of the purchases made and more than three-quarters of the value. In the U.S., research indicates online retail sales are beginning to mature and that most new growth is expected from existing online shoppers spending more (JupiterResearch, 2007). The results presented here highlight three factors suggesting that new online consumers should also contribute significantly to growth in e-commerce sales.

First, experience online has a positive effect on the propensity to make an online order. By 2007, 54% of Internet users reported five or more years online, up from 45% in 2005. The 2009 CIUS should find an increasing average number of years online among Canadians, as well as a greater breadth of use and higher odds of making an online order. Online banking and window shopping can be considered as prerequisite behaviours or precursors to placing an online order. In 2007, just under half (43%) of Canadians used the Internet to window shop and for electronic banking and paying bills, up from 37% in 2005.

A second underlying factor is demographic. The observed age differences in online behaviour are not an age effect per se but rather a cohort effect. Almost all Canadians 16 and 17 years (97%) were online during 2007, but they accounted for less than 2% of the number of online orders. 14  However, they are part of the first generation that has grown up with the Internet and will start to earn income over the next few years. These younger users will likely continue to expand their use of the Internet, particularly with the incorporation of e-commerce into social networking and virtual world sites.

Technology is another factor to consider. The Internet is becoming more mobile, allowing for connections through multiple platforms, making online purchasing even more convenient, particularly for digitally delivered services. Such convenience, combined with increasing Website functionality and product availability, should entice many more Canadians, with increasing Internet experience, to begin shopping online. While the short term impact of the economic downturn on e-commerce is less certain, 15  data from the 2009 CIUS will serve to further assess these trends and patterns.

6   References

Associated Press, the (2009). "Home of Web celebrates 20th birthday looking ahead". March 23.

Eastin, M. and LaRose, R. (2000). "Internet self-efficacy and the psychology of the Digital Divide". Journal of Computer-Mediated Communication, 6 (1).

Ellison, J., Earl, L. and Ogg, S. (2001) Internet shopping in Canada. Connectedness Series, 3. Statistics Canada: 56F0004M2001003.

Gault, F. (2006) Measuring knowledge and its economic effects: The role of official statistics, in B. Kahin and D. Foray (eds.), Advancing Knowledge and the Knowledge Economy. Cambridge: The MIT Press.

Grau, J. (2008). CanadaB2C E-Commerce: A Work in Progress. New York: eMarketer.

JupiterResearch (2007). U.S. Online Retail Forecast, 2006 – 2011., January 12.

Ker, A. (2009). Retailers competing for market share: 2008 retail sales. Analysis in Brief, 80. Statistics Canada: 11-621-M2009080.

McKeown, L. and Underhill, C. (2007). Canada's top online spenders: Who are they and what are they buying. Innovation Analysis Bulletin, 9 (1). Statistics Canada: 88-003-X20070019624.

McKeown, L. and B. Veenhof (2009). Internet use: An international and inter-provincial comparison. Innovation Analysis Bulletin, 11 (1). Statistics Canada: 88-003-X200900110816.

Middleton, C. and Ellison, J. (2008). Understanding Internet usage among broadband households: A study of Household Internet Use Survey data. Science, Innovation and Electronic Information Division, Working Paper,3. Statistics Canada: 88F0006X2008003.

Noce, A. and McKeown, L. (2008). A new benchmark for Internet use: A logistic modeling of factors influencing Internet use in Canada, 2005. Government Information Quarterly, 25 (3).

Organisation for Economic Co-operation and Development (2008). The Future of the Internet Economy: A Statistical Profile. OECD Seoul Ministerial on the Future of the Internet Economy.

Organisation for Economic Co-operation and Development (2009). Guide for Measuring the Information Society, 2009. Paris: Working Party on Indicators for the Information Society.

Peterson, G. (2001). Electronic commerce and technology use. Connectedness Series, 4. Statistics Canada: 56F0004M2001005.

Rowland, W. (2006). Spirit of the Web: The age of information from telegraph to Internet. Toronto: Thomas Allen Publishers.

Roy, F. (2007). Cross-border shopping and the loonie: Not what it used to be. Canadian Economic Observer, December 2007. Statistics Canada: 11-010-X200701210464.

Sciadas, G. (2006). Our lives in digital times. Connectedness Series, 14. Statistics Canada: 56F0004M2006014.

Shaw, H. (2009). Canadian Tire kills Web site ordering. Financial Post. January 20, 2009

UCLA (2004). Annotated SPSS output: Logistic regression. UCLA Academic Technology Services:

Underhill, C., and Ladds, C. (2007). Connecting with Canadians: Assessing the use of government on-line. Connectedness Series, 15. Statistics Canada: 56F0004M2007015.

Zamaria, C., and Fletcher, F. (2008). CanadaOnline!The Internet, Media and Emerging Technologies: Uses, Attitudes, Trends and International Comparisons. Toronto: Ryerson University.