Alexa for Shopping's 100+ account actions: lists, billing and delivery for CPG

On October 29, 2026 Amazon published a straightforward claim: Alexa for Shopping can now handle more than 100 common shopping actions by conversation for all US customers. The examples are not science fiction. Create a shopping list. Add items to a baby registry. Skip a Subscribe & Save delivery. Change delivery instructions on a Palo Alto address. Ask where a refund is.

We already wrote a Subscribe & Save playbook the week that update landed, and an earlier feature-by-feature Alexa playbook for listing hygiene. This post is the rest of the October 29 wave: lists, registries, delivery instructions, membership questions and billing. Those actions change how shoppers organise purchases around your product, not only how often a subscription arrives.

The numbers Amazon put on the table

Amazon's post states three adoption figures worth keeping together:

Claim	Figure
Customers who used Alexa for Shopping in the last 12 months	Over 400 million
Growth in customers using agentic features since the start of the year	More than 8x
Where the new actions are available	All US Amazon customers (Shopping app and desktop)

Source: Amazon, October 29, 2026.

Retail Dive on November 5 repeated the 400 million figure and added Andy Jassy's Q3 line: US customers using Alexa for Shopping spend 50% more per order than those who do not. That is earnings commentary, not a public study design, but it is the number Amazon's CEO chose to highlight.

Show the data

Metric	Value
Customers in last 12 months (millions)	400
Agentic-feature user growth since year start (multiple)	8

Source: Amazon, Alexa for Shopping adds 100+ new capabilities, Oct 29, 2026. 400 million is 'over 400 million' in Amazon's wording. Agentic growth is 'more than eight times.' Chart uses the floor figures Amazon stated.

For a CPG brand selling on Amazon, the implication is not "voice shopping is niche." It is that a large installed base can now rearrange lists, delivery notes and billing questions without opening the old account menus.

What the 100+ actions cover (from Amazon's list)

Amazon grouped the new actions into five buckets. We keep their wording close so you can map each one to listing work.

1. Lists and registries. Create, add to and edit lists. Move items between lists. Examples Amazon gives: grocery lists, baby registries, holiday wish lists, wedding registries. Sample prompts: "Create a new shopping list," "Add all to my baby registry," "Show me what's left on my wedding registry."
2. Subscribe & Save. Skip the next delivery, pause a subscription, change delivery to every three months.
3. Prime and other memberships. "What are my Prime benefits?" "How much have I saved with Prime this year?" "When does my Audible subscription renew?"
4. Delivery instructions. Change or add delivery instructions on an address.
5. Payments and billing. Unrecognised charges, gift card balance, refund status.

The Subscribe & Save row is familiar. The list and registry row is the one most CPG teams under-prepare for. A protein bar, diaper, formula, sunscreen or spice set that never appears in registry-friendly language loses the moment when a shopper says "add that to the baby list" and the assistant has to decide what "that" is.

Playbook: lists and registries

Treat list and registry behaviour as a discovery surface, not a afterthought.

Write pack copy that survives a list context. A list item often shows a short title. If your Amazon title leads with a promo phrase and buries count, flavor and who it is for, the list looks like clutter. Lead with product, count and use.

Mark gift and registry intent in bullets. For baby, holiday and household staples, say what is in the box in plain units (12 cans, 30 sachets, 2-pack of 500 ml). Registry shoppers and agents both need that.

Keep variants distinct. When someone says "add the vanilla one to my list," the assistant needs clean flavor and size attributes. Duplicate or overlapping child ASINs are how the wrong item lands on a wedding list.

Align the brand site. Amazon says Alexa for Shopping draws on catalog detail, reviews and information across the web. If your site calls a pack "family size" and the listing calls it "32 count" with no household guidance, list suggestions get fuzzier.

Playbook: delivery instructions and constraints

Delivery instruction prompts sound like logistics, not merchandising. They still touch the brand.

• If your product is temperature sensitive, hazmat-adjacent, or needs adult signature, say so in the listing bullets and in A+ content the assistant can read. A shopper who tells Alexa "leave it at the side door" after buying ice cream or probiotics is not your fault, but unclear storage copy makes disappointed reviews more likely.
• For heavy multi-packs, state weight and whether the box is porch-friendly. Agents that manage delivery notes will not invent your constraints.
• Match any "ships in X" claim across the detail page and your own site. Contradictions are the sort of thing conversational assistants surface when a shopper asks why a box is late.

Playbook: billing, refunds and trust copy

Amazon's examples include "I have an unrecognized charge on my account" and "Where is my refund?" Those flows are Amazon's to run. Your job is not to invent a parallel refund bot. Your job is to avoid giving the assistant mushy fuel.

• Refund-relevant claims (seal guarantee, freshness dating, "contact us if unsatisfied") should match what your brand site and package say.
• Avoid listing promises you cannot honour off-Amazon. When a shopper asks the assistant about a charge, inconsistent warranty language becomes a support ticket.
• Keep ingredient and allergen facts stable. Billing chats often sit next to "what did I buy" questions.

Playbook: Prime benefit questions near your category

Shoppers will ask Alexa what Prime has saved them this year. You cannot control that answer. You can control whether your Subscribe & Save and Prime-eligible offers are described accurately so the assistant does not misstate a discount when comparing options.

• State the Subscribe & Save discount and interval options in the listing the way Amazon shows them, not as a custom marketing fraction.
• If a multipack is the unit that makes Prime shipping sensible, say how long the multipack lasts for a typical household (and label any worked example as illustrative if you invent a household size).

A one-week test plan

Day 1: From a US account, open Alexa for Shopping and create a list. Add your top three ASINs by voice or chat. Note which title string appears.

Day 2: Ask it to move one ASIN to a second list or a registry-style list. Note failures on variant names.

Day 3: For a Subscribe & Save ASIN, run skip, pause and "every three months" prompts. Confirm the listing still explains days of supply.

Day 4: Ask a delivery-instruction prompt on a recent order of your product (or a comparable SKU). Check whether your page warned about storage.

Day 5: Ask a refund-status style question on a test order if you can do so safely. Compare the assistant's product description with your bullets.

Day 6: Compare the same three ASINs on your brand site. Fix any name, count or claim drift.

Day 7: Log the failures, assign owners (copy, catalog, site), and retest after changes.

XENA Intelligence's work on how assistants describe products still applies on Amazon's own surface: the facts have to be consistent before the conversation starts. SearchPanel will not replace walking through Alexa for Shopping with your own ASINs, but it is a useful parallel check on how other engines talk about the same SKUs.

What not to duplicate from earlier playbooks

If you already ran the Subscribe & Save cadence work from our October 30 post, do not redo it from scratch. Keep that cadence table, then add list/registry and delivery/billing checks on top. If you never ran the May-era Alexa listing audit, start with the original playbook for attributes and Q&A, then return here for the October 29 account actions.

Amazon is turning account chores into sentences. Brands that only optimise for the buy button will miss the sentences that decide whether your product stays on the list, arrives on the porch correctly, or survives the "what is this charge" moment.