Amazon is replacing Rufus as a visible shopping chatbot with a broader assistant called Alexa for Shopping. Rufus remains part of the underlying technology, but the customer experience now stretches across Amazon’s app, website and selected Echo Show devices.
The change is more than a rebrand. Alexa for Shopping can answer product questions, compare alternatives, retrieve purchase history, track prices and support scheduled or conditional buying actions. It combines Amazon’s product knowledge with the personal context held by Alexa+.
Amazon says Rufus helped more than 300 million customers research, compare and buy products in 2025. Monthly active users reportedly rose more than 115%, while engagement increased almost 400% year over year.
Shopping assistance moves into the main search journey
Customers no longer need to open a separate chatbot window. They can ask shopping questions through Amazon’s principal search bar, including requests for recommendations or details about earlier purchases.
The assistant can place products side by side, generate summaries on product pages and provide category overviews in search. For larger purchases, it can build guides combining features, prices and reviews from Amazon and the wider web.
This integration makes conversational discovery feel less like an experimental feature and more like a new interface for the store. Instead of translating an intention into keywords and filters, a customer can describe the situation the product must solve.
Amazon is rolling out Alexa for Shopping to signed-in US customers without requiring Prime, an Echo device or the Alexa app. Echo Show 15 and 21 owners with Alexa+ also receive a full-store experience using voice, touch or both.
Memory turns recommendations into delegated actions
Alexa for Shopping uses Amazon browsing, purchases, shopping history and Alexa interactions to personalise its responses. Customers can manage details such as family members, pets, interests and dietary requirements.
The assistant can monitor an item’s price for up to a year and expose its price history. It can also schedule recurring purchases, create birthday reminders and suggest gifts. Conditional instructions allow an item to be added to the cart when it reaches a target price and has not been bought recently.
Amazon’s Shop Direct can surface products from other online stores. For eligible products, the Buy for Me agent may complete a transaction with the customer’s primary address and payment method.
That progression—from advice to a prepared cart and then an automated purchase—changes the nature of consent. A useful shopping agent needs to distinguish a reminder, a recommendation and permission to spend money.
Trust depends on transparent commercial boundaries
Amazon operates the marketplace, sells its own products, earns fees from third-party sellers and now provides the agent influencing the choice. Customers need to understand how recommendations are ranked and whether commercial relationships affect the result.
Personalisation also needs careful controls. Purchase history may reveal health needs, family circumstances or sensitive interests. Users should be able to see which data shaped a recommendation, remove outdated assumptions and limit information shared across Alexa and shopping surfaces.
Automated purchases should begin with narrow value and quantity limits. The assistant must show the item, merchant, total cost, delivery conditions and return policy before an instruction becomes standing authority. High-value or unusual purchases deserve fresh confirmation.
Price tracking creates another responsibility. A lower price is not necessarily a better deal if the seller, product configuration or terms have changed. The agent should explain the comparison rather than simply announcing that a threshold was met.
Amazon’s investment in AI infrastructure is considerable, with the company linking a sharp increase in property and equipment spending to AI. Alexa for Shopping provides a direct route for that investment to influence revenue-generating customer behaviour.
The product could make complex shopping faster and routine restocking nearly invisible. Its long-term advantage will not depend only on how well it finds products. It will depend on whether customers can predict when the assistant is advising, when it is acting and whose interests it serves at each step.
