In recent years, e-commerce has experienced massive growth as a result of the COVID pandemic. The share of e-commerce in total US retail sales rose from 11.9% in the first quarter of 2020 to 16.4% in the second quarter of 2020 (source: data from the U.S. Census Bureau).

In lockdown 2020, analysts were excited about the penetration of e-commerce. They spread all kinds of charts with one message: in a few months, e-commerce had made a leap of three to five years.

In e-commerce operations, this demand led to a sudden need for fulfillment capacity. Amazon tried to capitalize on this by expanding its warehouse network and lean management. According to Amazon's annual reports, warehouse space doubled within 2 years. Amazon did not use all of this for pure fulfillment. Amazon has created many transshipment points for the last mile. Amazon uses these to shorten delivery times to customers. The increase in stock levels was greater at Amazon than at any other retailer.

Now people are going back to the office, back to the stores. And for e-commerce, it looks as if much of the growth was only temporary. It means a return to the trend line. Another factor influencing this is the cost of delivery for e-commerce services. Inflation is causing customers to save on delivery costs and prefer to store in a physical store.

With the decline in e-commerce sales in the first quarter of 2022, e-commerce demand has stabilized. It reflects pre-pandemic growth. If the pandemic had not restricted shoppers' choices, e-commerce sales would probably have grown steadily towards current demand.

Benedict Evans adds in his analysis: Is the benchmark for the penetration of e-commerce still correct?

Analysts argued that it does not make sense to look at the share of e-commerce in all retail sales. These include goods such as gasoline or car repairs, which would not be sold online. Statistics usually look at the penetration of "addressable retail". This excludes sales from gas stations, car dealers, car repairs and car parts. In the US statistics, "retail" also excludes bars and restaurants.

Evans argues that "addressable retail" does not make sense. He cites direct to consumer in automotive and other sectors as an example. Tesla sells cars exclusively online. Against this backdrop, the penetration of online induced sales is even greater.