Amazon Paid Units Grew 17 Percent and Sellers Carried 61 Percent of Them: The Marketplace Is the Store

Amazon sold 17 percent more paid units worldwide in the second quarter of 2026 than it did a year earlier, and third-party sellers accounted for 61 percent of those units. Those two figures, both published by Amazon in its own quarterly release on 30 July 2026, are the ones an operator actually needs from this earnings cycle. They say that the demand Amazon is manufacturing with record delivery speeds and AI shopping assistants is being filled, in the majority, with inventory Amazon does not own. The seller is not a guest on the platform. The seller is the platform’s stock. Every fee change, every handling-time rule and every fulfillment policy reads differently once you accept that, and Amazon’s own disclosure is what forces the point.

Two percentages that keep getting confused

This quarter produced two numbers in the mid-to-high teens that describe completely different things, and they are already being swapped for each other in coverage.

• North America segment sales grew 16 percent year over year, to $116.2 billion. That is dollars, in one geography, including retail, seller services, advertising and subscriptions.
• Worldwide paid units grew 17 percent year over year. That is units, everywhere, and it is the closest thing Amazon publishes to a measure of how many things actually got bought.

The rest of the top line: net sales rose 20 percent to $200.6 billion, from $167.7 billion a year earlier. International segment sales grew 15 percent to $42.2 billion. AWS grew 37 percent to $42.2 billion. Operating income was $27.5 billion, up 43 percent from $19.2 billion.

Net income was $62.6 billion, or $5.75 per diluted share, a 245 percent increase. Set that number aside when you think about the store. Amazon states that the quarter includes $53.4 billion of non-operating pre-tax other income, primarily from its investments in Anthropic. It is a mark on an investment, not a signal about retail demand.

The unit growth is the fastest in the window Amazon discloses

Amazon publishes six quarters of unit metrics in its supplemental business metrics. Worldwide paid-unit growth over those six quarters runs 8 percent, 12 percent, 11 percent, 12 percent, 15 percent, then 17 percent. The trend has been up for three straight quarters and Q2 2026 is the strongest of the set.

That matters because of how Amazon’s chief executive described the previous quarter. Speaking in May, Andy Jassy called the first quarter’s 15 percent paid-unit growth “the highest since the tail end of covid lockdowns”, as reported by EcommerceBytes. The quarter just reported beat it by two points.

The definition is worth knowing. Amazon’s units are physical and digital units sold net of returns and cancellations, by Amazon and by sellers in its stores, plus Amazon-owned items sold in other stores. They exclude AWS, certain subscriptions, rental businesses, advertising businesses and gift cards, and both the unit-growth and seller-mix figures exclude Whole Foods Market. It is a clean measure of transaction volume, which is why it beats the revenue lines as a seller’s benchmark.

61 percent is a share, and it moved

The seller unit mix is the figure worth writing on the wall, and it deserves to be read precisely rather than triumphantly. Across the same six disclosed quarters, third-party sellers accounted for 61 percent, 62 percent, 62 percent, 61 percent, 60 percent, then 61 percent of worldwide paid units.

Two things are true at once. Sellers carry the clear majority of everything sold in Amazon’s stores, and have done so in every quarter Amazon discloses. And the mix is one point lower than the same quarter last year, when it stood at 62 percent, having recovered a point from the 60 percent low in Q1 2026. Over twelve months the mix tilted marginally toward Amazon’s own retail, inside a pool of units that grew 17 percent.

So the honest framing is not “sellers are winning more of Amazon”. It is that sellers hold a structural majority of the store’s volume while Amazon’s first-party retail holds its own at the margin. For anyone planning around fees, the majority is the part that matters: a one-point move in mix does not change who supplies most of the shelf.

Where the money lands in the P&L

Amazon reports revenue by line, and the lines make the seller’s position concrete. All figures are for Q2 2026, with year-over-year growth:

• Third-party seller services: $46.78 billion, up 16 percent from $40.35 billion. Amazon defines this line as commissions plus any related fulfillment and shipping fees, plus other third-party seller services. This is what the marketplace charges its suppliers.
• Online stores: $70.43 billion, up 15 percent. Amazon’s own retail sales, recorded gross.
• Advertising services: $19.81 billion, up 26 percent. The fastest-growing line in the store, well ahead of units.
• Subscription services: $13.73 billion, up 12 percent.
• Worldwide shipping costs: $27.87 billion, up 19 percent.

Line those growth rates up and the operator’s problem is visible without any modelling. Units grew 17 percent, seller services revenue 16 percent, advertising 26 percent, and shipping costs 19 percent, which is faster than both online stores at 15 percent and seller services at 16 percent.

That last relationship is the one to keep. The cost of physically moving goods is growing faster than the revenue Amazon books from moving other people’s goods. You do not need a leaked memo to understand why fulfillment pricing, dimensional rules, placement fees and handling-time requirements keep moving. Meanwhile the fastest-growing line of all is the one you bid into voluntarily.

Why “the marketplace is the store” is a practical position, not a slogan

Look at what Amazon chose to highlight from the quarter. It delivered more than 40 percent more items same-day or overnight in the first six months of 2026 than in the same period last year. It merged Rufus and Alexa+ into Alexa for Shopping, and says active users came close to doubling while interactions rose more than 5x year over year. It added millions of products to selection, including over 700,000 from named brands. Its 30-minute delivery service, Amazon Now, grew gross sales and units sold more than 80 percent quarter over quarter and served more than 60 percent more customers.

Every one of those is a demand-side investment. And 61 percent of the units that demand converts into are supplied by third parties. Amazon is building the fastest, most conversational storefront in retail and stocking the majority of it with your inventory. That is the whole strategic asymmetry, and it cuts both ways.

It cuts your way in that Amazon cannot hit a 17 percent unit growth rate without sellers. Selection, coverage, long-tail availability and the speed promise all depend on third-party stock sitting in the network. That is real leverage and it is structural, not seasonal.

It cuts against you in that being the majority of the inventory makes you the majority of the cost base. When shipping costs grow 19 percent, the pool being asked to absorb that is mostly sellers. A fee change is not Amazon adjusting the rent on a tenant. It is Amazon repricing its own cost of goods, which happens to be your margin. Read fee announcements as inventory economics rather than as landlord behaviour and they stop being surprising.

What the figures do not say

Discipline on the boundaries of this data matters as much as the data.

1. 61 percent is units, not money. It is not GMV, not revenue and not profit share. Amazon does not publish seller GMV in this release, so any dollar figure attached to that 61 percent is somebody’s estimate, not a disclosure.
2. The 17 percent is worldwide and all-category. Your category may be growing at half that or twice it. Treat it as a benchmark to measure yourself against, not a forecast to plan inventory on.
3. The mix figures exclude Whole Foods Market and units exclude gift cards, AWS, rentals and certain subscriptions.
4. Guidance signals a slower headline quarter. Amazon expects Q3 2026 net sales of $197.0 billion to $202.0 billion, growth of 9 percent to 12 percent, and says that excluding the impact of Prime Day in both 2025 and 2026 the year-over-year growth rate would be nearly 400 basis points higher. Operating income is guided to $22.5 billion to $26.5 billion against $17.4 billion a year ago. If you see a slower Q3 growth print, check the Prime Day timing before concluding demand cooled.

The one operator takeaway

Benchmark your own unit growth against 17 percent, not against your own last year. If your units grew 8 percent in a store whose units grew 17 percent, you lost share inside a growing market, and flat-to-up revenue hid it from you. That single comparison is worth more than the rest of the earnings release, and it is available to you before any fee announcement lands.

The rest follows from position. Sellers supply 61 percent of what Amazon sells. That is not a courtesy extended to guests, it is the operating model. Price, argue and plan accordingly.

Sources

• Amazon.com, Inc., Amazon.com Announces Second Quarter Results, 30 July 2026 (earnings release, including Supplemental Financial Information and Business Metrics and Certain Definitions)
• Amazon.com, Inc. via Business Wire, Amazon.com Announces Second Quarter Results
• EcommerceBytes, Ina Steiner, Amazon North American Sales Grow 16% in Second Quarter 2026
• Digital Commerce 360, Amazon sales and quarterly results