The Census Bureau dropped Q2 numbers on August 18. U.S. e-commerce grew 12.2% year over year. That is more than twice the 6.7% print for retail as a whole, and it is the second straight quarter of double-digit online growth after a year that sat around 5%.
Online also took a record 17.1% of all U.S. retail spending.
What happened (confirmed)
Seasonally adjusted, not adjusted for prices:
• E-commerce sales: $340.2 billion, up 3.8% from Q1 and 12.2% from Q2 2025
• Total retail: $1,986.5 billion, up 2.9% from Q1 and 6.7% from Q2 2025
• E-commerce share: 17.1% of total retail, up from 17.0% in Q1 and 16.3% a year earlier
Unadjusted, the same quarter is $329.5 billion and 16.4% of retail. Census publishes both. The 17.1% is the seasonally adjusted series operators quote. Neither series is adjusted for inflation.
The year-ago stack is the useful one. E-commerce grew 5.0% in Q2 2025, 5.3% in Q3, 5.9% in Q4, then 10.1% in Q1 2026 and 12.2% in Q2. Two double-digit quarters after four that never cleared 6%.
Source: U.S. Census Bureau, Quarterly Retail E-Commerce Sales, 2nd Quarter 2026 (CB26-133), released August 18, 2026.
What the 17.1% actually measures
Census counts an order as e-commerce if it was placed online. Payment does not have to happen online. The denominator is all retail, including cars, gasoline, and restaurants, categories that barely sell on Amazon or Walmart.
That means 17.1% understates the share of the market you actually compete in. Marketplace Pulse made the same point on August 27. Treat the headline as "online is taking share again," not as your category's true online mix.
The platforms already took it
The market grew 12%. The large platforms grew faster.
• Walmart U.S. eCommerce grew 24%; Walmart U.S. advertising grew 38% (fiscal Q2 ended July 31, reported August 20)
• Amazon Online Stores grew 15% to $70.4 billion; advertising grew 26% to $19.8 billion (Q2 2026 earnings)
• Shopify North America GMV grew 28% (Q2 2026 earnings)
Growth above the market rate comes from somewhere. It comes from stores, and from smaller channels. If your Q4 plan still assumes last year's 5% online world, you are planning against a year that already ended.
Why this matters
A 12% e-commerce print with Amazon, Walmart, and Shopify all running ahead of it is not a "retail is healthy" story. It is a mix story. Shoppers spent more online, and they spent it on the platforms that already have the inventory, the ads, and the two-day promise.
Q4 inbound windows are already in motion. Amazon's Prime Big Deal Days receiving is early September. Walmart's October Event inbound is September 1. If demand is back online at 12% instead of 5%, the cost of showing up late is not a missed promo. It is a missed share of a market that just re-accelerated.
Prices are part of the print. Census does not deflate these numbers. Marketplace Pulse estimates roughly a third of the recent acceleration is bigger tickets, which still leaves real growth near 8%. That is enough. Do not wait for a "real" number. The order volume is already showing up in Seller Central and Seller Center.
Who it affects most
• Brands that cut Q4 inventory or ad budget off 2025's 5% growth year
• Amazon-heavy catalogs with no WFS plan against a Walmart channel growing 24%
• Finance models still treating marketplace GMV as a 2025 run rate plus a holiday spike
• Teams that paused new-ASIN launches because "online slowed"
• Anyone reading 17.1% as a ceiling instead of a lagging share print
What CoreTrex would do this week
1. Rebuild the Q4 unit forecast off Q2, not off 2025. If your category tracked online retail, 12% is the base, not 5%. Then layer the event calendar on top.
2. Check inbound against the dates we already published. Prime Big Deal Days inventory has to land in early September. WFS October Event inventory has a September 1 receive-by. A 12% demand tape does not wait for a late truck.
3. Look at where the extra demand is actually converting. If Amazon units are up and Walmart is flat, you have a listing and ads problem, not a demand problem.
4. Do not confuse inflation for volume. Pull units, not revenue, for the last two quarters. If units are up, buy the inventory. If revenue is up and units are not, you have a price and fee problem.
5. Put ad budget on in-stock SKUs on both marketplaces. A 26% Amazon ads print and a 38% Walmart ads print mean the auction is already pricing in the demand you just read about.
Our take
2025 was a pause, not a new normal. Census just confirmed it.
The operators who spent last year talking themselves into a mature, 5% e-commerce market are about to walk into Q4 with last year's inventory and this year's demand. The platforms already took the share. The only question is whether your SKUs are on the receipt.
Q4 forecast still sitting on 2025 math? CoreTrex will rebuild the unit plan off the new demand tape, pick the inbound that can still make September, and put ads only on listings that are actually in stock. Talk to a strategist.