Amazon relocated a portion of its Canadian sourcing from the US to China to evade tariffs.

      Amazon anticipates that package volume in Canada will increase by over 40% between 2026 and 2029, surpassing its growth in the US each year, as indicated by internal documents reviewed by Business Insider.

      A notable detail from a March planning document reveals that Amazon has redirected some Canadian direct-import sourcing from the United States to China, described in the document as a measure “to avoid tariffs,” which suggests a US company circumventing US trade policy by making fewer purchases from America.

      The company was still outlining its expansion plans in late July, following Donald Trump's announcement on July 20 of an additional 50% tariff on certain Canadian imports. These tariffs, which took effect this month, apply to some products that were previously protected under the US-Mexico-Canada Agreement (USMCA).

      Earlier, Amazon had documented its perspective on the risk, noting in March that the tariff impact would be less significant in Canada than in the US because of USMCA coverage. However, it cautioned that “the risk of a Canadian recession is higher than usual due to US trade policy uncertainty.”

      The situation now has reciprocal implications. Canada announced retaliatory tariffs on US imports effective September 8, which include consumer categories such as appliances and electronics, impacting a specific aspect of Amazon's operations.

      This aspect is Remote Fulfillment, a program allowing third-party sellers to list products in the Canadian marketplace while storing inventory in US warehouses, with items only crossing the border once a Canadian customer places an order. The tariff on US consumer goods entering Canada specifically affects these shipments.

      An Amazon spokesperson stated that prices in its Canadian store “have not increased outside of normal fluctuations” and that the company is keeping an eye on the new tariffs for potential effects.

      Since 2010, Amazon claims to have invested over C$65 billion, approximately $47 billion, in Canada and employs more than 46,000 people there.

      The motivation for continuing expansion is competitive rather than macroeconomic. The March document stated, “Amazon Canada faces a critical competitive challenge as major retailers are outpacing our delivery capabilities, putting future growth at risk,” specifically mentioning Walmart, Loblaws, and Best Buy.

      The data supporting this claim is worth examining closely. Amazon provides same-day delivery to about 54.5% of Canadian Prime members, while the document indicates that competitors reach 70% to 85% of Canadian households within two to four hours.

      These statistics utilize different metrics and speeds, preventing a direct comparison, but it is clear enough for Amazon’s planners to have recognized the issue.

      Walmart has since introduced Walmart+ in Canada, offering unlimited same-day delivery, which significantly intensifies competition for Amazon.

      The other key factor is fulfillment capacity, where warehouse automation plays a crucial role. European retailers have been optimizing throughput from existing facilities rather than building new ones, with Decathlon doubling its output across seven sites using robotics instead of expanding infrastructure.

      Amazon’s strategy focuses on proximity. Its projections state that 63% of Canadian shipments will be fulfilled within 160 miles of the customer and 93% within 1,000 miles by 2029, presenting a challenge related to warehouse location rather than just technology.

      An economic reversal detailed in the documents complicates matters. The cost of third-party last-mile delivery in Canada is about half that in the US, diminishing the financial justification for Amazon to manage deliveries internally, with speed becoming a key competitive factor that everyone is investing in.

      This situation has led to a specific decision: Amazon assessed 12 potential conventional delivery stations in Canada, found that they would yield negative five-year returns, and opted for cheaper partner-based models instead.

      These forecasts are preliminary and may change, serving as a meaningful caveat in a document created amid discussions of two tariff rounds rather than mere standard language.

      Historically, tariffs have been more effective at rerouting trade than halting it, as seen with US duties on Chinese electric vehicles, and Amazon’s sourcing change exemplifies this phenomenon.