The Watson Weekly is sponsored by Avalara — the agentic AI platform automating global tax and compliance for leading eCommerce brands with simple plans, transparent pricing, and complete coverage across Shopify, Stripe, WooCommerce, BigCommerce, and more.

UPS and PayPal reported on the same morning and told opposite stories about the same thing.

UPS moved 3.3% fewer packages a day in the US and collected 9.3% more for each one. Domestic revenue rose 6% to $14.93B, domestic operating profit rose 21% to $1.19B, margin reached 8%. Small and mid-sized businesses went from 32.0% of domestic volume to 34.5%. That is the trade Carol Tomé has been promising since she started walking away from Amazon volume, and it finally arrived in one clean quarter. The market ignored it. UPS closed down about 6%, because GAAP EPS was $0.71 against $1.76 adjusted, the gap being $891M of after-tax separation costs from the Driver Choice Program, and because Q3 domestic revenue is guided flat with volume down mid single digits.

PayPal did the reverse. Total payment volume grew 10% to $486.4B. Revenue grew 5% to $8.68B. Run those against each other and the take rate fell from 1.87% to 1.79%. Transaction margin dropped to 44.9% from 46.4%. GAAP operating income fell 5%, non-GAAP fell 8%, and the $1.38 that beat consensus was still down 1% year over year. Active accounts ended at 439M, up 0.3%, and 200,000 lower than March. PayPal is processing more and keeping less of it, which is what happens when Braintree grows mid-teens and branded checkout is guided to low single digits.

Shopify's GMV hit $115.6B, up 32%, 30% in constant currency. Revenue grew faster than volume at 34%, which is precisely the thing PayPal cannot do. Operating income was $488M against $291M a year ago. On the call the company put payments at 68% of GMV, B2B GMV up 76%, and said agentic traffic and orders tripled year over year, with catalog-fed AI searches converting at roughly twice the rate of scraped data. Those last figures are Shopify's own and unaudited, and they are the ones worth arguing about.

Amazon cleared $200B in a quarter for the first time. AWS grew 36.7% to $42.2B on $16.6B of segment operating income. Advertising grew 26%. AWS and the entire International segment both landed at exactly $42.2B, so the cloud business is now the size of everything Amazon sells outside North America.

Neither Amazon nor Shopify led with this one. Amazon's $62.6B of net income includes $53.4B of non-operating pre-tax other income, primarily marks on its Anthropic stake. Shopify's $1.50B includes $1.06B of after-tax equity investment gains. Two of the four biggest bottom lines this cycle were mostly not the business.

Amazon's trailing twelve month free cash flow is now an outflow of $7.6B against an $18.2B inflow a year ago, on $66.1B more spent on property and equipment. The paper gain and the cash burn are the same bet. Only one of them has settled.

Peloton earned $63.2M in FY26, its first profitable year, and the stock fell roughly 13% on the news.

FY26, reported August 6:

The profit is genuine and worth crediting. Q4 operating expenses fell 12%, subscription gross margin reached 73.6%, and net leverage improved from 1.2x to 0.3x. Total debt is still $1.3B, sitting against $1.2B of cash. For a company that spent four years as a case study in overbuilding, that is real work.

What the guidance says is that the engine behind it has mostly run its course. FY27 revenue is guided to $2.3 to $2.4 billion, off 3.9% at the midpoint, which management tied to lapping the price increases it put through last fall. Q1 subscriptions are guided down 9.8% year over year.

Cost reduction and pricing are one-time levers. Pull them again and the base you're pulling against is smaller.

Stern called FY26 the year Peloton grew up. The balance sheet supports him. The subscriber file is guided down again in FY27, and nothing in the outlook says what stops it.