E-commerce is growing. Profit is a separate question.

What the latest releases actually say

The U.S. e-commerce market deserves attention. It also deserves a careful reading. The Census Bureau’s August 18 release estimated second-quarter 2026 retail e-commerce sales at $340.2 billion, seasonally adjusted, with 12.2% growth from a year earlier and a 17.1% share of total retail sales. These are figures from that release, not a fresh October measurement. They were not adjusted for price changes. [1]

There is an essential footnote: Census now warns that the August report no longer contains its most up-to-date estimates following revisions on September 28. It says revised e-commerce figures will appear with the third-quarter release scheduled for November 19. Treat those August numbers as a dated snapshot, not an unrevised dashboard. [4][3]

A newer signal comes from the October 15 advance retail release: September retail and food services sales increased 0.8% from the previous month and 6.3% from September 2025. That is a broader spending measure, not an e-commerce-only growth rate, and it also excludes adjustment for price changes. It adds context; it cannot fill the gap in the online-sales series. [2]

A bigger market does not describe your margin

Sales estimates answer a market-size question. They do not tell an individual merchant how much is left after buying inventory, winning the order, delivering it, and handling a return. Nor does nominal sales growth establish how much more merchandise changed hands: spending can change with prices as well as purchasing activity.

The useful business question is narrower: which orders make the next order easier to earn? A campaign can increase revenue while leaving too little contribution to fund customer service, inventory, or the next campaign. That is an operating possibility to test in your own numbers, not a conclusion about every seller in this market.

Build a scorecard that survives the headline

Start with contribution per order using a consistent definition: revenue after discounts and expected refunds, less the product and variable costs you choose to attribute to that order. Make the definition visible so a seemingly better month is not just a change in what you counted.

Then compare cohorts of customers acquired in the same period. Look at repeat purchasing, refund patterns, and how much contribution those customers generate after acquisition costs. Separate an introductory discount from a sustainable full-price relationship. These are suggested management checks, not universal benchmarks.

Finally, stress-test an otherwise attractive offer. Ask what happens if shipping costs more than expected, return rates rise, or repeat purchases arrive later. A store with several brands should be able to explain the economics of each one; an aggregate revenue total can hide a weak offer behind a stronger one.