Amazon looks likely to take Alexa for Shopping, the agentic assistant that replaced Rufus in the United States in May, into at least three international marketplaces by the end of the first quarter of 2027, and to open a conversational shopping assistant in at least one marketplace where Rufus has never been available (Mexico, Brazil or Australia). The prediction does not rest on a press release. It rests on a burst of 14 job postings across Seattle, Palo Alto, London and Bengaluru between August 19 and September 18, 2026, one of which is titled “Principal PMT, Rufus International” and describes making “Worldwide from Day 1” the default for every marketplace. Read against the Alexa+ device rollout (twelve countries after the India launch on September 16) and the two-speed copy Amazon used for Prime Big Deal Days on September 15, the pattern suggests the geography gap between the US assistant and everyone else is about to close.
In short
- The prediction: by March 31, 2027, Alexa for Shopping (the brand plus at least one automation feature such as Price Alerts, Auto-Buy or Scheduled Actions) is likely live in three or more non-US marketplaces, with the UK and Canada the base case for first launch, and a conversational shopping assistant under either name is likely to open in at least one of Mexico, Brazil or Australia.
- Signal 1 (hiring): 14 distinct Rufus and Alexa for Shopping postings on amazon.jobs in the 30 days to September 18, including a Principal Product Manager for “Rufus International”, a London engineering manager for a team that will “greatly expand Rufus’s availability to new cohorts of customers”, and a senior corporate counsel for agentic commerce deals.
- Signal 2 (footprint): Alexa+ reached India in Early Access on September 16 and Australia on August 5, its eleventh and twelfth countries, with “more than 10 additional countries in 2027” promised; Rufus itself serves nine marketplaces, and the merged Alexa for Shopping product serves one.
- Signal 3 (calendar): Amazon’s US announcement for Prime Big Deal Days (October 6 to 7) sells Deal Alerts, Auto-Buy and 365-day price history under the Alexa for Shopping name; the European announcement, published the same day, still calls the assistant Rufus and lists none of the automation features.
- The counter-case: Rufus took between two and fourteen months to follow its US launch into each new marketplace, EU consumer and payments rules make Auto-Buy harder to ship, and a bipartisan Senate letter on September 17 has just put Alexa for Shopping in front of the FTC.
Why this matters now
Amazon has spent 2026 turning a chatbot into an agent. Rufus launched to a small US group in February 2024, reached all US customers that July, and then spread to the UK, Germany, France, Italy, Spain, Canada, India and Japan over the following fourteen months. On May 13, 2026, the company retired the Rufus name in the US and launched Alexa for Shopping, a product it describes as Rufus’s product expertise merged with the personalization of Alexa+. The new assistant can track prices for a year, set alerts, schedule repeat purchases and, with Auto-Buy, complete a purchase without the customer touching the app.
That last step matters because it changes what an international launch involves. Answering a question about headphones needs a catalog, a language model and some guardrails. Buying the headphones on the customer’s behalf when the price drops needs stored payment credentials, consent flows, refund logic and a legal position in every jurisdiction where it ships.
The US version has all of that; the eight other Rufus marketplaces do not. Amazon’s own July 30 earnings commentary claimed more than 350 million customers had used the assistant in the trailing twelve months, with active users nearly doubling and interactions up more than five times year over year, which is exactly the kind of engagement data that tends to unlock the next tranche of investment.
The question this piece tries to answer is not whether Amazon wants the agentic version of its assistant in more countries. Every signal says it does. The question is whether the observable inputs (who Amazon is hiring, what the device team has already shipped, and how the company is talking to shoppers in different markets this quarter) support a specific window.
We think they support the first quarter of 2027, and we lay out what would have to happen for that to be wrong. This is a different question from the one we asked in July, when the hiring pattern pointed to Amazon opening its catalog to external AI agents; this time the signals concern Amazon’s own assistant, and where it is allowed to act.
Signal 1: a 30-day hiring wave with an international product owner at its center
Between August 19 and September 18, 2026, amazon.jobs listed 14 distinct roles whose title or opening paragraph names Rufus, Alexa for Shopping or conversational shopping, according to the public search endpoint on the site. Seven are in Seattle, three in Palo Alto, three in London and one in Bengaluru. That is not an enormous number for a company of Amazon’s size, and we would not build an argument on the count alone. The argument comes from what the postings say.
The anchor is a Principal Product Manager, Technical, for “Rufus International”, posted in Seattle on August 21. The listing says the role “will define how we build Rufus for the world”, owning “the product vision and multi-year technical strategy for global expansion, making ‘Worldwide from Day 1’ the default across every marketplace and surface”. It notes that Rufus is “already serving customers across Europe, Japan, India, and Canada” and asks the hire to “dramatically reduce the time, cost, and complexity of bringing Rufus innovations to customers globally”, spanning “global feature parity, localization, skills and tool integrations, and platform evolution”. A company that plans to keep its agent US-only does not hire a principal to make worldwide parity the default.
The London postings are the second tell. A Software Development Manager for “Alexa for shopping (Rufus)”, posted August 19, describes a team “building new software systems and customer experiences that will greatly expand Rufus’s availability to new cohorts of customers”, building “agentic AI systems powered by LLMs”, and calls the team “strategic” with “visibility with leadership at the highest levels”. On September 8, London added a Senior UX Design Program Manager for Alexa for Shopping, whose brief covers the experience “from the Amazon Shopping app and website to Echo Show”, and a Software Development Engineer in Test for the “Rufus Quality Engineering team” tasked with enabling continuous deployment across Rufus services. Amazon has kept the Rufus science and product core in Seattle and Palo Alto since 2024; standing up a design program office and a release-quality function in London is what a second launch region looks like from the outside.
Three more postings fill in the operating model. Two Applied Science Manager roles in Palo Alto (September 1 and September 9) describe teams working on “LLM and/or VLM post-training and alignment for new shopping experiences” and name “multi-agent systems” as core expertise, which is the science stack behind Auto-Buy and Scheduled Actions rather than question answering. In Bengaluru, an operations team that manages annotation for training and evaluation says its remit runs “across shopping features, devices, and locales”.
And on August 31, Amazon posted a Senior Corporate Counsel, Agentic Commerce Experiences, in Seattle, to “independently draft, structure, and negotiate complex commercial and transaction agreements” and “manage multiple deals at once” for “a team building agentic shopping experiences”. Lawyers who close deals are hired when there are deals to close: local payment partners, device retailers, content and data licensors.
Signal 2: Alexa+ has already gone where Rufus has not
The second signal is the device side of the same product. Alexa+, the generative assistant Amazon announced in February 2025 and rolled out to US customers in early 2026, has moved through international markets faster than Rufus ever did. Mexico opened Early Access on January 22, 2026, followed by the UK on March 19, Italy on April 15, Spain on April 23, Germany on May 7, France on May 26, full availability in Canada on June 11, Brazil on June 18, Australia on August 5 and India on September 16, alongside Austria.
The India launch, in Hindi and English, is free during Early Access and free for Prime members afterward, with a price of INR 2,000 a month for non-Prime customers, according to Amazon’s announcement and reporting by TechCrunch. Amazon’s own launch post says it “will also expand Alexa+ to more than 10 additional countries in 2027”.
Put the two maps side by side and the gap is the story. Alexa+ is live or in Early Access in twelve countries. Rufus is live in nine marketplaces. Alexa for Shopping, the product that fuses them, is live in one.
Four Alexa+ countries (Mexico, Brazil, Australia and Austria) have no Rufus at all, which means a customer there can ask the device to reorder groceries but cannot ask the shopping app to compare two coffee makers. One Rufus marketplace (Japan) has no Alexa+. The eight markets where both are live (the UK, Canada, Germany, France, Italy, Spain, India and the US) are the obvious candidates for the merged product, and the US already has it.
This matters for timing because Alexa for Shopping is not a new model waiting to be trained; it is a packaging decision layered on infrastructure that has already cleared the hard parts of localization. Alexa+ in the UK, Germany and India already handles local language, local partner integrations (the India version connects to Swiggy, Zomato’s District and MakeMyTrip) and local payment rails for tasks like grocery ordering. The India post explicitly describes “ordering groceries from Amazon Now” as a supported task. Once a device can transact on a customer’s behalf in a marketplace, the remaining work to let the shopping app do the same under the Alexa for Shopping name is mostly product surface, consent design and legal review, which is the work the London and Seattle postings describe.
Signal 3: Prime Big Deal Days is being marketed at two speeds
The third signal is the most recent and the most specific. On September 15, Amazon announced that Prime Big Deal Days will run October 6 to 7 in 22 countries. The US announcement on aboutamazon.com is, in effect, an advertisement for Alexa for Shopping: it invites customers to “create Deal Alerts to track deals for specific product needs”, to “set a target price on any specific item and get notified when the price drops, or turn on Auto-Buy”, to ask the assistant to “see up to 365 days of pricing insights”, and to ask “are any items on my list on deal right now?”. The word Rufus does not appear.
The European announcement on aboutamazon.eu, published the same day for the same event, is built differently. It tells shoppers that “Amazon’s AI shopping assistant, Rufus, helps customers find gifts for their hardest-to-shop-for loved ones, plan gluten-free holiday menus, or even source the perfect Christmas tree for a small space”, and that Lens Live “shares insights and answers from Rufus”. There is no Deal Alert, no Auto-Buy, no price-target language and no Alexa for Shopping brand. The two documents describe the same 48-hour event with two different assistants, and only one of them can buy things.
Why does that count as a forward-looking signal rather than a description of the status quo? Because deal events are where Amazon has historically introduced assistant features to international customers, and because the European copy is now visibly a version behind. The company that wrote “Worldwide from Day 1” into a job description on August 21 published a marketing gap three weeks later that its own product team is being hired to close.
The most plausible reading is that the automation features were not ready for a September 15 announcement in Europe, and that Black Friday (November 27) and the first quarter of 2027 are the next two windows. The October event gives outside observers a clean checkpoint: if the UK, German or Canadian storefronts carry Alexa for Shopping branding or price-target features by October 6, the timeline in this piece is running early; if they still say Rufus, the base case holds.
Signals matrix
| Signal | Date | Source type | What it shows | Weight |
|---|---|---|---|---|
| Principal PMT, Rufus International (Seattle) | Aug 21, 2026 | amazon.jobs listing | Multi-year owner for “Worldwide from Day 1” parity across marketplaces | High |
| SDM, UX program manager, SDET (London) | Aug 19 to Sep 8, 2026 | amazon.jobs listings | Second-region engineering, design and release function; “expand availability to new cohorts” | High |
| Applied Science Managers (Palo Alto) | Sep 1 and Sep 9, 2026 | amazon.jobs listings | Post-training and multi-agent science for “new shopping experiences” | Medium |
| Senior Corporate Counsel, Agentic Commerce Experiences | Aug 31, 2026 | amazon.jobs listing | Deal-making lawyer for agentic shopping partnerships | Medium |
| Alexa+ India Early Access, Australia launch | Sep 16 and Aug 5, 2026 | Amazon announcements | Device assistant now in 12 countries, 10 more promised for 2027 | High |
| Prime Big Deal Days copy split (US vs EU) | Sep 15, 2026 | Amazon announcements | Automation features marketed in the US only; Europe still on “Rufus” | Medium |
| Q2 2026 engagement disclosure | Jul 30, 2026 | Earnings commentary | 350m users in 12 months, active users nearly 2x, interactions 5x | Context |
What the pattern suggests
Taken together, the three signals describe a company that has finished the science phase of its shopping agent, has proved the demand case in the US, has already solved localization on the device side in a dozen countries, and is now hiring the product, design, quality and legal functions that a multi-market launch needs. None of those facts alone would justify a dated prediction. The combination does, because each one removes a different reason the launch could stall: the international PMT removes the “nobody owns it” problem, the London team removes the “no second-region capacity” problem, Alexa+ removes the “localization from scratch” problem, and the Q2 numbers remove the “does anyone use it” problem.
The precedent that governs timing is Rufus itself. Amazon took the assistant from full US availability in July 2024 to the UK in September 2024 (about two months), to Germany, France, Italy and Spain in beta by November 2024 (four months), and to Japan in beta by January 2025 and full availability in September 2025 (fourteen months). Alexa+ followed a slower curve, roughly eleven months from US Early Access to Mexico and nineteen months to India, but that curve included building an entirely new assistant.
Alexa for Shopping is a merger of two products that already exist in the target markets, so the relevant precedent is the faster one. Applying a two-to-nine-month lag to the May 13 US launch gives a window from July 2026 to February 2027. The July end of that window has already passed without a launch, which is itself informative: it suggests the block was not engineering but the automation features, and it narrows the base case to the back half of the window.
The specific forecast, then, is this. By March 31, 2027, Alexa for Shopping is likely live under that name, with at least one of Price Alerts, Auto-Buy or Scheduled Actions, in three or more marketplaces outside the US. The UK and Canada are the base case for first launch because both share English, both already have Alexa+ at full availability, and both had Rufus before Germany.
India is the likely third, with the Alexa+ Early Access that began September 16 as the enabling step and Amazon’s stated 20% growth in Indian smart-home device sales as the commercial motive. A conversational shopping assistant is also likely to open, under either the Rufus or Alexa for Shopping name, in at least one of Mexico, Brazil or Australia in the same window, because the “new cohorts of customers” language in the London posting reads most naturally as new marketplaces rather than new segments in existing ones. We put the first leg at roughly 65% and the second at roughly 55%.
Prior precedents: how fast Amazon has exported AI shopping features
| Feature | US launch | First international market | Lag | Footprint at the 18-month mark |
|---|---|---|---|---|
| Rufus (conversational assistant) | Feb 2024 beta, Jul 2024 full | UK, Sep 2024 | 2 months from full US | 9 marketplaces |
| Alexa+ (device assistant) | Early Access from Feb 2025 | Mexico, Jan 2026 | 11 months | 12 countries, 10+ promised for 2027 |
| Amazon Lens (visual search) | 2024 to 2025 in stages | Progressive | Not disclosed | 21 countries as of Q2 2026 |
| Alexa for Shopping (merged agent) | May 13, 2026 | None yet | 4 months and counting | 1 market |
Wider context: the agent race is now a distribution race
Amazon is not the only company with a shopping agent, and the competitive clock is part of why the international gap is unlikely to persist. OpenAI’s checkout inside ChatGPT, Google’s push through AI Mode and Gemini, and the Universal Commerce Protocol that Amazon itself joined in April 2026 have all turned agent-mediated shopping into a channel with real order volume. In the US, Amazon’s answer has been to make its own assistant the most capable one on its own storefront. Outside the US, that answer currently does not exist in agentic form, which leaves European, Canadian and Indian shoppers with a Rufus that can compare products while third-party agents increasingly offer to buy them.
There is also a merchant-facing reason for Amazon to move. The company’s July 2 posting for a Principal PMT for “Partner Enablement” on Alexa for Shopping describes a platform (internally abbreviated AFS) on which teams including “Your Orders, Payments, Amazon Business, Health, Grocery, and Softlines” can “independently build, evaluate, launch, and operate” their own assistant experiences. That is a platform model, and platform models are expensive if they serve one country.
The internal economics push toward parity as much as the competitive ones do. For sellers and brands, the practical consequence is that the ranking and content rules that already govern how Amazon surfaces products in the US, including the conversational layer, will likely apply in the UK, Germany and India within two quarters rather than two years.
Finally, the device business gives Amazon a reason to prefer markets where it sells hardware. India’s Alexa+ launch post cited 20% year-over-year growth in smart-home device sales there. Australia’s Alexa+ arrival was described locally as fast-tracked from a planned 2027 date.
Mexico and Brazil got Alexa+ before Germany. Those choices say that Amazon’s international assistant roadmap is being driven by device attach and Prime penetration, not by the order in which Rufus first shipped, and that is why we expect at least one Alexa+-only country to receive a shopping assistant in the window even though it never had Rufus.
Implications for sellers, brands and platforms
For marketplace sellers in the UK, Germany, France, Italy, Spain, Canada, India and Japan, the most concrete implication is that price-target automation is coming to their listings. Auto-Buy and Price Alerts turn a customer’s stated willingness to pay into a standing order that executes on the seller’s next price move. In the US this has already changed the meaning of a temporary price cut: a discount that would once have generated a spike of impulse buys now also triggers a queue of pre-committed purchases, and a seller who cuts price by ten dollars to win the Buy Box can find a cohort of Auto-Buy orders clearing at the new price within minutes. Sellers should expect the same mechanics in the UK and Canada first and should model their promotional calendars accordingly.
For brands, the assistant’s international arrival raises the stakes of product content. Alexa for Shopping synthesizes catalog data, reviews and web information into recommendations, and its Deal Alerts run on category-level intent (“alert me to deals on cozy sweaters”) rather than on brand searches. Brands that have invested in the structured attributes and comparison-ready copy that AI assistants read will benefit disproportionately when the agent starts recommending in a new language; brands that rely on sponsored placement in the search grid will find that the conversational surface does not carry the same inventory. Our guide to preparing a store for AI shopping agents lays out the specific fields and content patterns that determine inclusion.
For competing platforms and retailers, the timing forecast is a planning input. Walmart’s Sparky, Shopify’s agentic tooling and the regional marketplaces in Europe and India all compete for the same shoppers. If Amazon’s agent reaches the UK and Germany by early 2027, the window in which a rival can claim “the only assistant that can actually buy for you” in those markets is measured in months. Retailers with their own assistants should treat Prime Big Deal Days and Black Friday as the two dates on which Amazon is most likely to reveal international automation features, and they should have a response ready for the day the UK storefront switches its assistant’s name.
For investors, the international rollout is a lens on Amazon’s stated priority of raising the share of purchases that start in conversation. The Q2 disclosure of 350 million users and five-fold interaction growth was a US-heavy number. Expanding the merged assistant to markets that already hold roughly a third of Amazon’s retail revenue is the most direct way to keep that growth rate from decelerating in 2027, and the hiring pattern says the company knows it.
Caveats: what could go wrong
The first and largest risk is regulatory, and it comes from two directions. In Europe, Auto-Buy is a merchant-initiated transaction executed without the customer present, which interacts with strong customer authentication rules under the revised Payment Services Directive, with the Consumer Rights Directive’s requirements around explicit order confirmation, and with the AI Act’s transparency obligations for systems that make decisions on a person’s behalf. None of those rules bans an automated purchase, but each adds design work and legal review that the US launch did not need.
The London postings are consistent with that work being under way; they are not proof that it is finished. If Amazon decides to launch the international product without Auto-Buy, the prediction as written (brand plus at least one automation feature) can still be met through Price Alerts or Scheduled Actions, but the more cautious reading is that Europe gets a thinner product and a later date.
The second direction is the US itself. On September 17, Senators Tammy Baldwin and Rick Scott asked the FTC to investigate whether Alexa for Shopping and Walmart’s Sparky “obscure and suppress” information about Made in USA claims, citing a July report from Columbia Law School’s Center for Law and the Economy. That letter lands in the same month as a $20 billion FTC advertising suit against Amazon and an expanded $2.5 billion Prime settlement.
A company facing three simultaneous regulatory fronts over its shopping experience may reasonably decide that the safest quarter to expand an assistant that makes purchase decisions is not this one. That would push launches into the second quarter of 2027 without changing the direction.
The third risk is that we are over-reading the hiring. A Principal PMT for “Rufus International” with a “multi-year” mandate could be building platform foundations for 2028 rather than shipping in 2027; the job text speaks of reducing the “time, cost, and complexity” of global launches, which is what a company writes when those launches have been slow. The London team’s “new cohorts” could mean new customer segments in existing markets, such as business or grocery customers, rather than new countries.
And the science-manager postings could reflect attrition replacement rather than growth. We think the combination with the Alexa+ map and the Prime Big Deal Days copy makes the geographic reading more likely, but the hiring signal alone is compatible with a slower schedule.
The fourth risk is the Alexa+ pricing model. Alexa for Shopping in the US is free to all signed-in customers, while Alexa+ abroad is free for Prime members but priced for others (INR 2,000 a month in India). If Amazon decides the merged assistant should inherit the Prime gate internationally, launches become entangled with Prime pricing decisions in each market, which have their own calendar. A Prime-gated launch would still count under our definition, but it would likely reduce the pace and the number of markets.
Scenarios to March 31, 2027
| Scenario | What happens | Observable by | Our probability |
|---|---|---|---|
| Base case | Alexa for Shopping brand plus at least one automation feature in the UK, Canada and one of India or Germany; a shopping assistant opens in one of Mexico, Brazil or Australia | Black Friday 2026 (first market), Q4 2026 results in early Feb 2027, March 31, 2027 (final check) | About 55% |
| Early | UK or Canadian storefronts carry Alexa for Shopping branding or price-target features before Prime Big Deal Days | October 6, 2026 | About 15% |
| Thin launch | Brand changes in three or more markets but automation features stay US-only through Q1 2027 (prediction fails on the feature leg) | March 31, 2027 | About 15% |
| Regulatory pause | Rufus remains the international product through Q1 2027; first international Alexa for Shopping launch slips to Q2 2027 or later | Q1 2027 results, late April 2027 | About 15% |
FAQ
What exactly is the prediction, and when can it be checked?
Two legs. First, that by March 31, 2027, Alexa for Shopping is live under that name with at least one of Price Alerts, Auto-Buy or Scheduled Actions in three or more Amazon marketplaces outside the US. Second, that by the same date a conversational shopping assistant, under either the Rufus or Alexa for Shopping name, is available in at least one of Mexico, Brazil or Australia.
Both can be checked by opening the relevant storefront and by Amazon’s launch announcements. Early checkpoints are Prime Big Deal Days on October 6 to 7, Black Friday on November 27, and the Q4 2026 results in early February 2027.
Isn’t this just a rebrand? Why does the name matter?
The name is a proxy for the feature set. In the US, Amazon attached the Alexa for Shopping name to the version of the assistant that can act (track prices, alert, schedule and buy), while keeping Rufus as the name for the version that answers questions. That is why the prediction requires the brand plus an automation feature rather than either alone. A pure rename with no automation would be a thin launch and would not count.
How many job postings are we talking about, and how were they counted?
Fourteen distinct listings posted between August 19 and September 18, 2026, found by querying the public search endpoint on amazon.jobs for “Rufus”, “Alexa Shopping”, “Alexa for Shopping” and “conversational shopping” and de-duplicating by listing ID, keeping only postings whose title or opening paragraph names the assistant. Seven are in Seattle, three in Palo Alto, three in London and one in Bengaluru. Older listings from June and July, including the Partner Enablement and agentic-identity roles, were used as context rather than counted.
Why would the UK and Canada go first rather than Germany or India?
Language and readiness. Both markets share English with the US build, both have Alexa+ at full availability (Canada since June 11, the UK since its March Early Access), and both received Rufus before the continental European markets in 2024. India is a strong third because Alexa+ Early Access opened there on September 16 and Amazon has cited 20% growth in device sales, but its Early Access status means the merged product likely follows rather than leads.
Could Amazon launch in Mexico, Brazil or Australia without ever calling it Rufus?
Yes, and that is the more likely path. Those markets have Alexa+ but never had Rufus, so Amazon has no legacy brand to retire there. The second leg of the prediction therefore accepts either name. The signal that supports it is the London posting’s language about expanding availability “to new cohorts of customers”, which reads most naturally as new marketplaces, combined with Amazon’s demonstrated preference for launching assistants where device sales are growing.
What is the strongest argument that this does not happen by March 2027?
Regulation on both sides of the Atlantic. Auto-Buy in Europe must clear strong customer authentication and explicit order-confirmation rules, and the AI Act adds transparency obligations. In the US, a bipartisan Senate letter on September 17 asked the FTC to examine how Alexa for Shopping handles Made in USA claims, in the same month as a $20 billion FTC advertising suit and a $2.5 billion Prime settlement.
A company under that much scrutiny could reasonably delay expanding an assistant that makes purchase decisions. We rate a slip into Q2 2027 or later at roughly 15%.
Does the Q2 earnings data actually support international demand, or is it a US number?
Amazon did not split the 350 million figure by country, and the fastest growth in interactions is likely US-weighted because the automation features that drive repeat use exist only there. That is precisely the argument for expansion: the markets with Rufus but no automation are where the engagement multiplier has not yet been applied. The number should be read as evidence that the product works, not as evidence that international customers are already using the agentic version.
What should a non-US seller do before the launch?
Three things. Audit listing content for the structured attributes and comparison-ready copy that conversational assistants read, because Deal Alerts run on category intent rather than brand searches. Model the effect of price-target automation on promotional pricing, since a price cut can now trigger pre-committed Auto-Buy orders as well as impulse purchases. And watch the UK and Canadian storefronts around October 6 and November 27 for a name change, which is the earliest public sign that the timeline is running ahead of this forecast.
How would you score this prediction as wrong?
If on March 31, 2027, fewer than three non-US marketplaces carry the Alexa for Shopping name with at least one automation feature, the first leg fails. If none of Mexico, Brazil or Australia has a conversational shopping assistant by the same date, the second leg fails. Partial credit applies if the brand ships without automation (the thin-launch scenario) or if exactly two markets launch. A launch that is Prime-gated still counts.