Amazon's Q3 in five numbers: what the quarter says about AI shopping and CPG

Earnings releases are written for investors, so the parts that matter to a consumer brand tend to sit in the middle of a long list of bullet points. Amazon's third-quarter 2026 release, published October 29, is a good example. The headline numbers were net sales up 12% to $201.9 billion and AWS growing 39%. The numbers a snack, beauty or cleaning brand should care about are further down.

We pulled out five of them, checked each one against the release or a report of the call, and wrote down what we think each means for a CPG team. Where Amazon gives a comparison without a baseline, we say so. Where a number is a quote from the call rather than the release, we name the publication that reported it.

1. Alexa for Shopping: 400 million customers, interactions up 570%

The release says Amazon "continued to drive rapid adoption of Alexa for Shopping, its agentic shopping assistant, which was used by 400 million+ customers in the last 12 months, with customer interactions up 570% year-over-year."

Two cautions before reading too much into that. First, "used by" over 12 months is a wide net. It includes anyone who asked the assistant a single question, and for part of that period the assistant was still called Rufus. Amazon folded Rufus into Alexa for Shopping in May. Second, Amazon does not define a "customer interaction." A shopper who asks three follow-up questions about one product may count as three. A 570% rise means the assistant is now part of how a lot of people shop on Amazon. It does not tell you how many orders it decided.

Even with those cautions, the direction is clear. The shopping assistant is now a mainstream Amazon surface, not a test. When we wrote our Alexa for Shopping playbook last week, the case rested on Amazon's launch claims. The quarter adds scale.

2. Groceries and essentials: about half of units

The second number is the one most CPG teams will underline. Amazon said it "continued growing its global Grocery business, with fresh groceries and everyday essentials growing about 3x faster than the rest of the store and representing about half of all units sold in Q3."

Half of units is not half of revenue. A bottle of dish soap and a television each count as one unit. Still, by count of things shipped, Amazon is now as much a store for consumables as for anything else, and that half is growing faster than the rest.

Amazon also said the number of items ordered through Amazon Now, its ultra-fast delivery service, is up 10x year over year, and that Amazon Now is available in over 450 towns and cities. Fast delivery of small baskets is a grocery and household behavior. It points the same way.

For a brand, the practical meaning is that the shopper asking Amazon's assistant "what's a good dishwasher pod for hard water" or "reorder my usual oat milk" is now a large and growing share of Amazon's traffic in your aisle.

3. Advertising: $19.6 billion, and a timing dip

Advertising is where many CPG brands feel Amazon most directly, through Sponsored Products and Sponsored Brands budgets. Here is advertising services revenue for the last six quarters, as Amazon reports it.

Show the data

Quarter	Revenue, $ millions
Q2 2025	15694
Q3 2025	17703
Q4 2025	21317
Q1 2026	17243
Q2 2026	19809
Q3 2026	19637

Source: Amazon, Q3 2026 results, segment information. Prime Day fell in Q3 in 2025 and mostly in Q2 in 2026, which shifts ad revenue between those quarters.

The raw revenue looks flat from Q2 to Q3. The growth rate tells a sharper story.

Show the data

Quarter	Growth excluding FX
Q2 2025	22%
Q3 2025	22%
Q4 2025	22%
Q1 2026	22%
Q2 2026	26%
Q3 2026	12%

Source: Amazon, Q3 2026 results, segment information. Amazon attributes the Q3 2026 slowdown mainly to Prime Day timing and expects high-teens ad growth in Q4 2026.

Amazon explains the drop directly in its guidance section: advertising revenue grew 12% excluding currency "impacted by timing anomalies primarily related to Prime Day." It reminds readers that Prime Day happened entirely in Q3 in 2025 but mostly in Q2 in 2026, and that total third-quarter growth would have been about 400 basis points higher without the shift. It expects advertising growth in the high teens in Q4.

So we would not read the 12% as a sign that brands are pulling back. Q2 and Q3 together look like the same steady growth, split differently.

The more interesting ad news is product, not revenue. Amazon introduced Amazon Ads Agent, which it describes as bringing "Amazon's full offering under a single, AI-first platform," and launched Full-Funnel Campaigns in it for all US advertisers. Advertisers "provide initial inputs, then Amazon's AI plans, executes, and continuously optimizes the rest." For a small brand, that means more of your Amazon ad decisions may be made by Amazon's software. The inputs you give it, starting with accurate product data, matter more.

4. The store itself: online sales growth after the Prime Day shift

Online stores, Amazon's first-party retail line, grew 2% excluding currency in Q3 after 15% in Q2. Third-party seller services grew 10%. Both carry the same Prime Day distortion, so the useful comparison is across the two quarters, not within one.

Table: Amazon retail and advertising lines, Q2 and Q3 2026 ($ millions, growth excluding FX)

Line	Q2 2026	Q2 growth	Q3 2026	Q3 growth
Online stores	70,432	15%	68,470	2%
Third-party seller services	46,780	16%	46,492	10%
Advertising services	19,809	26%	19,637	12%
Subscription services	13,730	12%	14,006	12%

Source: Amazon Q3 2026 results, segment information.

Subscription services, which is mostly Prime, grew a steady 12% both quarters. Prime members are the shoppers most likely to use Subscribe & Save and the assistant, so that steady line matters more to replenishment brands than the volatile retail ones.

5. 50% more per order, by Amazon's own comparison

The fifth number comes from the call rather than the release. Fortune reported that Andy Jassy told analysts US customers who use Amazon's shopping agent "spend 50% more per order than those who do not use it."

This is the number most likely to be repeated in sales decks, so it is worth being careful with. It compares users with non-users. People who choose to talk to a shopping assistant may already be heavier, more engaged Amazon shoppers. The figure does not show that the assistant causes larger baskets. It does show that the shoppers using it are valuable ones, which is reason enough to care what it says about your products.

In the same answer, according to Fortune, Jassy said outside shopping agents such as Meta's Muse are "still not great, nor more convenient than coming to a retailer directly." Amazon blocked Muse from its store in September. Read together, the 570% and the "not great" line describe one strategy: keep the shopping conversation on Amazon, inside Amazon's assistant.

What changed this week, not just this quarter

The same day as the results, Amazon announced more than 100 new Alexa for Shopping capabilities, available to all US customers. Several touch replenishment directly. Shoppers can ask it to skip the next Subscribe & Save delivery, pause a subscription or change a delivery to a different frequency. It can build lists, manage gift registries and update delivery instructions.

TechCrunch's write-up frames these as chores the assistant now handles. For a brand that relies on Subscribe & Save, they are more than chores. Skipping or slowing a delivery used to take several taps in the app. Now it takes one sentence. We look at what that means for replenishment brands in a separate playbook.

What we would do with these numbers

None of this calls for a new strategy. It calls for moving Amazon's assistant up the list of places you check.

1. Ask the assistant about your category, monthly. Use the shopper's words: "best unscented laundry detergent for sensitive skin," not your product name. Write down which brands it names and what it says about yours. Tip 25 covers what to do when it gets something wrong.
2. Treat essentials listings as the assistant's source. If about half of Amazon's units are groceries and essentials, the assistant is answering a lot of questions about products like yours. Bullets, A+ content and the Q&A section are the facts it can read. Make sure size, count, ingredients and use are in plain text, not only in images.
3. Write for replenishment. State how long a pack lasts at normal use. Tip 40 walks through it.
4. Keep other retailers consistent. Assistants compare. If your Walmart and Target listings say something different, the cheapest fix is making them match. Tip 30 explains why other retailer listings matter.
5. Watch what you hand to Ads Agent. If Amazon's AI will plan more of your campaigns, check the product data and creative it starts from.

A brand cannot see inside Amazon's assistant, but it can see its answers. Some teams do this by hand with a fixed list of questions. Others use tools; XENA Intelligence runs shopper questions across AI assistants, and its SearchPanel product scores how each product shows up in the answers. Either way, the habit matters more than the method.

The bottom line

Amazon's quarter confirms two trends at once. Consumables are now about half of what Amazon ships, and more of the shopping on Amazon happens in a conversation with its assistant. For CPG brands, the two meet in the same place: the assistant answering questions about everyday products and, starting this week, adjusting the subscriptions that keep those products coming. The assistant reads your listing, so your listing is where the work starts.