The Census Bureau’s November 19 release puts two measures side by side: U.S. retail e-commerce sales were 10.4% higher in the third quarter of 2026 than a year earlier, while total retail sales were up 5.6%. Yet e-commerce accounted for 17.9% of retail sales on a seasonally adjusted basis, compared with 17.1% a year earlier. The faster growth is meaningful, but the share did not jump by 10.4 percentage points. One number is a growth rate for a component; the other is that component’s portion of a larger total. Keeping those questions separate helps put online shopping headlines in proportion. U.S. Census Bureau

The release date and reference quarter are separate

The release was published November 19, 2026, and describes the third quarter, July through September. Its $347.3 billion estimate is seasonally adjusted and expressed in current dollars: Census adjusts for seasonal variation, but not for price changes. The release date tells readers when the estimate became available; it does not mean the spending occurred in November. U.S. Census Bureau

For a year-over-year comparison, the bureau said adjusted e-commerce sales rose 10.4% from Q3 2025, while total retail sales rose 5.6%. For a quarter-to-quarter comparison, e-commerce rose 1.6% and total retail sales 1.1%. These are different windows. A useful habit is to name the period and comparison aloud before drawing a conclusion: “Q3 compared with Q3 last year,” rather than simply “online sales are up.”

A growth rate is not a share of the market

The seasonally adjusted e-commerce share was 17.9% in Q3 2026, against 17.1% in Q3 2025. That is an increase of 0.8 percentage point in the share. It is not a 10.4 percentage-point increase. The 10.4% figure describes how the dollar value of e-commerce changed relative to its own value a year earlier; the share describes e-commerce dollars divided by total retail dollars.

A small round-number example makes the distinction clear. Suppose online sales start at $100 out of $1,000 in total sales, a 10% share. If online sales grow 10% to $110 while the total grows 5% to $1,050, the new share is about 10.5%, not 20%. Online sales grew faster, so their slice expanded, but the denominator grew too. The Census figures follow this same arithmetic, though the actual estimates use a much larger, survey-based total.

Seasonally adjusted and unadjusted shares answer different needs

The Q3 release reports both adjusted and not-adjusted estimates. The adjusted share was 17.9%; the not-adjusted share was 17.2%. The not-adjusted e-commerce dollar estimate was $335.7 billion, and the adjusted estimate was $347.3 billion. The gap is not a contradiction or a change in what counts as online. Seasonal adjustment smooths recurring patterns in the series; the two presentations should not be mixed within one comparison. U.S. Census Bureau

The quarter illustrates why that matters: the unadjusted e-commerce share was 18.9% in Q4 2025 and 16.5% in Q3 2025. Holiday timing can make raw quarters look quite different. For near-term comparisons, seasonally adjusted values can help control recurring calendar patterns. For a question about the actual dollars recorded during a specific quarter, the unadjusted series may be useful. State which series you use and compare like with like.

These are nominal sales estimates, not a count of orders

The Census release says the estimates are not adjusted for price changes. A larger dollar total could reflect some combination of more transactions, more expensive items, or a shift toward higher-priced goods. The headline alone cannot separate those possibilities, so 10.4% growth should not be described as 10.4% more packages, units, or inflation-adjusted consumption. It is a nominal sales estimate. U.S. Census Bureau

“E-commerce” is also a defined statistical category. Census includes sales where an order is placed, or its price and terms are negotiated, through an online system; payment itself need not happen online. The estimate comes from the retail survey sample, and the release notes that the latest quarter is preliminary and subject to revision. It also notes a historical revision using corrections and results from the 2023 and 2024 Annual Integrated Economic Survey. That revision history is another reason to treat the series as measured and updated data, not a fixed ledger. U.S. Census Bureau

What the measure can and cannot tell us

Census describes e-commerce estimates as coming from the same sample used for the Monthly Retail Trade Survey. The quarter’s estimate combines weighted reported or imputed sales, and the survey covers employer firms; its published definitions exclude some activities that may feel digital to consumers, such as online travel services and financial brokers. It is a retail measure, not a census of every online transaction in the economy. U.S. Census Bureau U.S. Census Bureau

The careful takeaway is narrow but informative: online retail sales grew faster than total retail sales over the year through Q3, and their adjusted share rose modestly. That supports saying e-commerce gained share in this measure. It does not establish why, tell whether households bought more physical goods after inflation, or show that every retailer benefited. Those questions require additional data. When reading the next release, check the reference quarter, adjustment basis, price treatment, comparison period, and revision status before turning a percentage into a story about consumer behavior.

THE EVIDENCE

Sources & further reading

  1. Quarterly Retail E-Commerce Sales, Third Quarter 2026 ↗U.S. Census Bureau
  2. Monthly Retail Trade Survey Methodology and Technical Documentation ↗U.S. Census Bureau