Le paysage mondial du commerce électronique

The textbook divides ecommerce into three stages. 1995–2000 was invention: bandwidth was limited, catalogues were simple, venture capital rushed in, and the period ended with the dot-com crash in 2000. 2001–2006 was consolidation: firms moved from technology-driven experiments to business-driven operations, broadband spread, and search advertising and rich media became viable. Since 2007, reinvention has combined social, mobile and local features through smartphones, social networks, apps and local services. The stages matter because they explain why today's tools are bundled the way they are: search advertising belongs to consolidation, while social and mobile commerce belong to reinvention.

With the history in place, the harder question is measurement. Two numbers are often confused: market size and penetration. Size is the value of online retail sales, usually in dollars. Penetration is online retail as a share of total retail. Textbook Table 2-1 gives both for 2020: China US$1,414.3 billion and 24.9%; the United States US$791.7 billion and 14.0%; the United Kingdom US$130.6 billion and 23.3%; Korea US$104.4 billion and 25.9%; Canada US$28.1 billion and 6.2%; Australia US$22.9 billion and 9.4%; Singapore US$3.2 billion and 11.7%.

Read the table slowly. China is by far the largest in money - about 1.8 times the United States and about 13.5 times Korea. But Korea's penetration (25.9%) is slightly higher than China's (24.9%), and the UK is close behind at 23.3%. The United States, despite its size, has a penetration of 14.0%. If you ranked these markets by size, you would get one order; by penetration, another. Neither ranking is wrong, but each answers a different question: how much money moves, versus how deeply online channels have replaced physical retail.

The textbook adds a global check: B2C was only about 18% of global retail in 2022. That single sentence prevents a common mistake - treating the visible platform economy as the whole retail market. More than four-fifths of retail still happened offline. For a seller, this is not a reason to ignore ecommerce; it is a reason to ask which categories and which customers have actually moved online.

Then there is the scope trap. A headline may say "global ecommerce sales exceeded US$5 trillion in 2022", citing eMarketer and including both B2B and B2C. Another may say "global retail ecommerce was about US$5.2 trillion in 2021" (textbook §2.8). The numbers look similar, but the years and the inclusion of B2B differ. A third figure may give one country's B2C online retail. Placing them side by side without labels produces apparent contradictions that are really definitional differences.

China's position is best described in three layers. The first is domestic market size: Table 2-1 makes China the largest B2C online retail market in 2020. The second is platforms and firms: Chinese platforms operate internationally, and their GMV may include sales in many countries. The third is cross-border export capacity, which customs statistics measure differently. Domestic online retail counts purchases inside China; cross-border ecommerce counts transactions between different customs territories. Neither figure can stand in for the other.

For students' home countries, comparable ecommerce-size data are often missing. The honest approach is to use indicators that exist and label what they measure. World Bank data put internet use in 2024 at 35.0% in Kenya, 41.2% in Nigeria and 53.4% in Bangladesh. These figures describe potential access. They do not tell you the online shopping rate, average spending, platform coverage, payment habits or delivery reliability. If no reliable ecommerce-size figure exists, write "no comparable data" and explain the gap; do not enter zero, and do not extrapolate from a neighbouring country.

A practical routine for any figure in this course: first, identify the entity (world, country, platform, firm); second, identify the measure (GMV, revenue, retail sales, penetration, users); third, identify the year; fourth, identify the source and whether it is primary or relayed. Only then compare. This routine is slower than reading headlines, but it prevents the most common error in discussions of the global landscape - comparing numbers that were never measuring the same thing.

Finally, resist the temptation to turn a landscape into a league table. Rankings by size reward large populations and rich economies; rankings by penetration reward small, highly urbanised markets. A country can be small in money and advanced in digital adoption, or large in money and still mostly offline. For strategy, the useful question is not "who is first" but "which measure matters for the decision I am making".