On the go? Listen to this roundup in audio format.
Welcome back to the Retail Roundup, your monthly recap of the biggest developments shaping retail media, marketplaces, and eCommerce strategy.
We're recapping May 2026, and if you thought the industry was entering a quiet summer lull, Amazon spent the last four weeks proving everyone completely wrong.
From a sweeping AI rebrand to a full-scale logistics business being sold to the world, this month delivered some genuinely interesting moves. Below is a structured breakdown of the five biggest stories and what each one means for your business.
Here are the highlights:
- Goodbye, Rufus! Amazon Merges Its AI Shopping Experience Under the Alexa Brand
- Amazon Now Launches 30-Minute Delivery in Major U.S. Markets
- Dynamic TV Creative Comes to Prime Video Ads
- Amazon Opens Its Supply Chain Infrastructure to the World
- LinkedIn Audience Signals Land Inside Amazon DSP
Let's dive in.
1. Goodbye, Rufus! Hello, Alexa for Shopping
This one arrived in waves. First, Amazon ran a test merging Rufus directly into the main search bar, surfacing AI-powered results at the very top of the search experience. Then, almost before the industry could fully process that, Amazon made it official: Rufus and Alexa+ are being unified into a single cross-device entity called "Alexa for Shopping."
Technically speaking, Alexa for Shopping is powered by Cosmo (similar to Rufus), Amazon's underlying AI engine, and combines three things: Amazon's deep product catalog knowledge, in-depth information from across the web (not just Amazon's own marketplace), and a user's personal preferences, shopping history, and conversations from both Amazon.com and Alexa devices.
That cross-web research capability is meaningful. Alexa for Shopping now works as a full-fledged AI research engine for commerce, comparable to ChatGPT, Gemini, or Perplexity, just pointed squarely at purchase decisions.
Alexa for shopping capabilities:
- Cross-Device Persistent Memory: Remembers desktop actions on Echo devices.
- Contextual Cart Prompts: Surfaces verbal Echo conversations on mobile apps.
- Unified Search Integration: Processes natural queries directly inside search bars.
- Side-by-Side Product Comparison: Evaluates multiple search results simultaneously.
- Top-of-Search AI Overviews: Displays high-level product category summaries instantly.
- Extended Price History: Tracks product costs up to one full year.
- Conditional Scheduled Purchases: Triggers orders based on price drop rules.
- Conversational Cart Building: Reorders past items via simple voice prompts.
- Preference Transparency Control: Users can audibly review what Alexa knows.
- Custom Cross-Web Guides: Compiles deep product research from around the web. Alexa now functions as a full-fledged AI search and research engine for commerce, just like ChatGPT, Gemini, or Perplexity
No, Search is Not Dead
The "search is dead" hot takes are already circulating. But Amazon isn't replacing keyword search; it's making both modalities available from the same search bar. A shopper can still type "dry dog food for senior dogs" the old-fashioned way. But they can also ask, "I have a border collie, what's a great food for her coat?" and get a personalized answer. Same bar, both options, different customer types served.
What is changing is the concept of a fixed ranking position. Several individuals could ask the same question and get four different results, each shaped by their history and preferences. The concrete notion of "I rank #1 or #5 for this keyword" is going to get murky quickly.
The Double-Sided Strategy (And the UCP)
Behind the scenes, Amazon is also staffing a 40-engineer team to integrate its commerce infrastructure into third-party agentic platforms via the Universal Commerce Protocol (UCP). This is Amazon stopping its losing battle against AI bots scraping its data, and instead building official APIs on its own terms.
The logic is clean. If you stay on Amazon, Alexa for Shopping keeps you captive inside the ecosystem. If you leave to research on ChatGPT or Perplexity, those external AIs still need to route the transaction through Amazon's UCP, and Amazon remains the merchant of record.
The walled garden has a one-way gate now. Information can flow out to external LLMs. External AIs cannot reach back in. And execution is ultimately why this holds. You can have a great conversation on ChatGPT, but ChatGPT doesn't own warehouses or delivery trucks. Amazon does. When it's time to actually get the product to a doorstep, the catalog and the network that win are Amazon's.
What This Means for Brands: Operationally, Nothing Has Changed Yet
From a Rufus optimization standpoint, nothing is different. The same knowledge graphs, the same back-end metadata, the same principle of making your product content speak clearly to use cases and USPs. That all carries over.
Amazon is essentially positioning Alexa for Shopping as the best-in-class shopping assistant - the same way Claude Cowork has become the reference point for workplace productivity AI. That's the category they're trying to own.
Pros
- Both traditional and conversational search are available from the same bar. Amazon is expanding the shopping experience, not disrupting it.
- The AI shopping assistant positions Amazon's platform as the highest-intent discovery environment available.
- Brands with strong, structured PDP content are well-positioned to show up across both modes.
Cons
- Traditional keyword ranking metrics become less reliable as personalization scales up across the user base.
- Brands whose content doesn't address specific use cases, customer contexts, or questions will lose visibility in AI-generated results faster than in traditional search.
💡 Tips
- From an optimization standpoint, keep doing what you're doing. The knowledge graphs and metadata signals powering Alexa for Shopping are the same as Rufus. There's no new playbook, just more urgency to execute the existing one well.
- Think in use cases, not just keywords. Alexa for Shopping personalizes results based on context. Make sure your content addresses who uses your product, when, and why, not just what it is.
- Reframe how you think about ranking. Position-based KPIs will become increasingly unreliable. Start building measurement frameworks around share-of-voice, content quality scores, and recommendation frequency instead.
- Don't place a single bet. Amazon, ChatGPT, Perplexity - shoppers may start their journey anywhere. Get your product data clean and consistent across sources so you show up wherever the recommendation happens.
2. Amazon Now Launches 30-Minute Delivery
Amazon officially launched Amazon Now, an ultra-fast delivery service promising thousands of fresh groceries and household essentials delivered in 30 minutes or less, available 24/7.
The service is already live in Atlanta, Dallas-Fort Worth, Philadelphia, Seattle, Austin, Denver, and other major markets, with plans to scale to tens of millions of customers before the end of the year.
How It Works
Rather than shipping from massive regional fulfillment centers, Amazon Now operates through a network of smaller, localized micro-fulfillment hubs positioned close to dense urban areas. When a qualifying order is placed, an Amazon associate picks it, and a local driver delivers it, often before the hour is up.
On the app, eligible customers see a "30-Minute Delivery" banner on their homepage, and qualifying items are flagged in search results. Selection is intentionally focused on high-urgency categories: fresh produce, dairy, health, baby, pet care, and personal care.
The costs:
- Prime members: $3.99 delivery fee; $1.99 small-order surcharge on orders under $15
- Non-Prime members: $13.99 delivery fee; $3.99 small-order surcharge on orders under $15
Pros
- Opens a genuinely new demand channel for high-velocity grocery and essential categories.
- Geo-targeted advertising can directly influence Amazon Now eligibility by spiking local velocity.
- 24/7 availability captures purchase intent at moments traditional retail can't.
Cons
- Eligibility is entirely algorithmic - no direct lever for brands to pull. The selection isn't based on what you submit - it's based on what the data says.
- Amazon's AI analyzes localized purchase history by ZIP code, identifies the highest-velocity items people buy when they're in a rush, and automatically pulls qualifying products into the micro-hub network.
- A single stockout event in a market removes the product immediately, putting high operational pressure on inventory management.
- Focused on select categories - not a fit for most non-perishable, non-urgent products.
💡 Tips
- Automated Inbound Placement: Let Amazon distribute your inventory across regional nodes by enrolling in Amazon's split-shipment and inventory placement features. The more locations your stock is positioned at, the higher the chance a local hub picks you.
- Flawless In-Stock Rate: A single out-of-stock event in a specific market is enough to get dropped from Amazon Now. The algorithm replaces you with a competitor instantly to protect the 30-minute promise.
- Local Sales Velocity: Use geo-targeted Amazon DSP campaigns to spike velocity in specific launch cities. The higher your local sales rate, the more likely Amazon's system is to pull that SKU into the nearby hub.
- Run geo-targeted DSP campaigns in Amazon Now launch cities. Spiking local velocity is the most direct signal brands can send to get the algorithm's attention.
- Think time-of-day. Grocery and personal care ads surfaced during pre-dinner windows or weekend morning hours aligned with the urgency Amazon Now is built to capture.
3. Dynamic TV Creative Comes to Prime Video
Amazon Ads just announced Dynamic TV Creative for Prime Video - an AI-powered layer that takes a single uploaded base creative and dynamically customizes headlines, CTAs, images, and product displays at the moment of impression for individual viewer segments.
Several individuals could be watching the same show on Prime Video tonight and see four completely different versions of the same ad - each one built from Amazon's trillions of first-party shopping signals and tailored to what each viewer's browsing and purchase history suggests about their likelihood to act.
How It Works
The operational lift is minimal. Brands upload one base creative, optionally provide headline variations (or let Amazon auto-generate them), and toggle "Enable Dynamic TV Creative" in campaign setup. From there, Amazon's intelligence layer handles the customization in real time.
Targeting elements that can be dynamically adapted include:
- CTAs (e.g., "Add to Cart," "Visit Brand Store," "Check Out the Reviews")
- Headlines
- Images and product display modules
The feature is currently endemic only, meaning it's available for brands advertising products sold on Amazon. Phase 2 of the beta is planned for July 2026 and will expand to include multi-ASIN carousels.
Why This Matters
The traditional ceiling on CTV creative personalization has been the number of versions a brand pre-produces. Four versions mean four audience experiences. Dynamic TV Creative effectively removes that ceiling - hundreds of distinct audience segments can receive distinct creative executions, all from a single upload.
For brands running upper-funnel streaming budgets, this is a meaningful step toward closing the loop between awareness investment and measurable purchase behavior. The expectation is more efficient media performance as creative relevance improves, but as with anything in early beta, the standard approach applies: test, verify, and let the data confirm the lift.
One nuance worth flagging: Dynamic CTAs mean you're partially ceding control over which conversion outcome Amazon prioritizes for a given viewer. That's a tradeoff worth watching closely once reporting becomes available.
Pros
- Dramatically reduces the creative decision-making burden on brand and agency teams.
- Scales personalization well beyond what traditional A/B testing allows.
- Aligns ad experience more closely with individual shopper behavior, which should improve performance efficiency over time.
Cons
- Currently in beta and limited to endemic advertisers.
- Reporting on dynamic variant performance is not yet available, limiting early optimization visibility.
- Ceding CTA control to Amazon's optimization requires a degree of trust that some teams will need time to develop.
💡 Tips
- Get into the beta now. Early access means earlier learning. The brands building familiarity with Dynamic TV Creative before Phase 2 launches in July will have a head start on multi-ASIN formats.
- Build measurement expectation-setting into your stakeholder conversations. Full performance reporting isn't available yet, but that's not a reason to wait. Frame early tests as paid learning on a format that's likely to become standard.
4. Amazon Opens Its Supply Chain Infrastructure to the World
Amazon has officially launched Amazon Supply Chain Services (ASCS) - a move that takes the full global logistics machine Amazon built for itself (ocean freight, air cargo, bulk warehousing, last-mile parcel delivery) and opens it to any business, even those that don't sell a single item on Amazon.com.
If that framing sounds familiar, it should. This is the AWS playbook: build massive internal infrastructure to solve your own problem, perfect it at scale, then sell it to the world.
The early adopter case studies make the scope clear:
- Procter & Gamble is using Amazon's freight services to transport raw materials to production facilities and move finished goods through its distribution network.
- 3M is using Amazon's logistics to move products from manufacturing sites to distribution centers worldwide.
- Lands' End is using Amazon's unified inventory pool to fulfill orders across multiple sales channels.
- American Eagle Outfitters is using Amazon's parcel network to deliver orders from their own website directly to customers.
What Amazon Is Actually Trying to Achieve
There are four things happening here simultaneously:
- Monetizing Excess Warehouse Capacity: Amazon overbuilt its physical footprint during the pandemic boom. By opening up the network to non-Amazon commerce, they turn a massive fixed-cost burden into a high-margin, recurring B2B revenue engine.
- Moving Upstream into the Global Supply Chain: Amazon is no longer just trying to dominate the "last mile" delivery to a consumer's doorstep. By handling raw materials for companies like P&G and 3M, Amazon is inserting itself into the entire global trade pipeline from factory floor to end-user.
- The Ultimate Data Play: If a brand runs its entire omnichannel business through ASCS, Amazon gains absolute visibility into global commerce data. They will see what is selling on Shopify, TikTok Shop, SHEIN, and Walmart, where raw materials are sourced, and what the exact fulfillment margins are.
- Driving Down Internal Costs via Network Effects: The more volume that flows through Amazon’s ships, planes, and hubs, the lower the unit cost becomes. Non-Amazon businesses are effectively funding the efficiency of Amazon's own retail network.
The "Easy Button" for Logistics (Pros)
- Radical Simplification: Instead of juggling five different contracts (ocean freight forwarders, customs brokers, drayage companies, bulk 3PLs, and parcel carriers like FedEx/UPS), brands can manage everything via a single, automated console.
- Unified Inventory Pooling: This is a massive cash-flow win. Instead of fragmenting stock ("1,000 units allocated for Amazon FBA, 500 for Shopify, 500 for wholesale"), brands like Lands' End use one unified inventory pool. Amazon’s advanced forecasting places inventory closer to localized demand across all channels.
- Unrivaled Last-Mile Speed: Brands gain access to Amazon's 7-day-a-week parcel shipping with predictable 2-to-5-day delivery speeds, bypassing traditional carrier bottlenecks during peak seasons.
The "Deal with the Devil" risks (Cons)
- The Platform Risk Nightmare: If Amazon controls your factory-to-doorstep supply chain, you are entirely at their mercy. If you have a marketplace suspension, a billing dispute, or Amazon suddenly raises fulfillment fees, your entire business freezes, including your independent website and wholesale channels.
- Handing Over the Competitive Playbook: You are giving your biggest competitor an intimate look at your supply chain data. If Amazon sees a specific product exploding on Shopify via their logistics network, their private-label division has the perfect blueprint to copy it.
💡 Tips
- Run a parallel logistics track. Maintain at least one independent fulfillment relationship outside of Amazon. Even if ASCS handles the majority of your volume, you need an exit ramp.
- Know what data you're giving up. If your highest-velocity SKUs' full supply chain is visible to Amazon, you're essentially sharing a product roadmap with your most capable competitor. Weigh that against the operational savings.
5. LinkedIn Audience Signals Are Now Inside Amazon DSP
This one is a continuation of a pattern we've been tracking for months: Amazon DSP is steadily consolidating media buying that used to require five different platforms into one.
Amazon Ads and LinkedIn just announced a major collaboration that brings LinkedIn's first-party professional audience data into premium streaming TV environments, via Microsoft Monetize, the SSP of LinkedIn's parent company, Microsoft, and a preferred Amazon DSP partner.
What This Actually Means (And What It Doesn’t)
To be clear on the mechanics: this is not a way to run ads on LinkedIn through Amazon DSP. It is a way to use LinkedIn's audience signal data to target people watching streaming TV.
The targeting dimensions LinkedIn makes available include job title, industry, and seniority, drawn from LinkedIn's billion-plus member base. Combined with Amazon's existing first-party purchase and browsing data, DSP advertisers can now build campaigns that reach, say, VPs of Supply Chain at mid-size CPG companies while they're watching Thursday Night Football on Prime Video.
The Broader DSP Story
This partnership follows the Netflix + Amazon DSP integration we covered in the March edition. That move unlocked Amazon's first-party purchase signals layered onto Netflix inventory. This one unlocks LinkedIn's professional identity signals layered onto premium streaming TV.
The common thread: Amazon DSP is systematically building toward a single platform where a media buyer can reach any high-value audience, in any environment, measured through one unified system. The Trade Desk doesn't have Amazon's retail data. LinkedIn's native ad platform doesn't have streaming TV inventory at this scale. That combination is genuinely difficult to replicate.
Pros
- Opens a powerful upper-funnel video channel for B2B marketers who have historically struggled to find precision targeting in TV environments.
- Simplifies media buying by eliminating the need to manage LinkedIn targeting and streaming TV buying through separate platforms.
- Measurement consolidates into Amazon's system, which supports more accurate attribution.
Cons
- Just because the audience can be reached doesn't mean it will convert. LinkedIn audience performance in non-B2B, non-endemic contexts is unproven.
- CTV CPMs are premium - this isn't a low-budget testing environment.
💡 Tips
- Match the product to the cohort before you activate. A garlic press in front of VPs in the technology sector is not a strategy. Start with a clear hypothesis: what is the professional audience signal that actually predicts intent for your product?
- Non-endemic B2B services are the natural early adopters. SaaS platforms, business credit cards, professional services, and enterprise tools are the categories most immediately suited for this.
- Test before you scale. LinkedIn's brand name carries weight in a media plan, but performance in this targeting context is unproven. Run a controlled test before committing meaningful budget.
Final Thoughts
May 2026 had one theme running through all five headlines: Amazon playing a longer game than most brands realize.
Alexa for Shopping isn't just a rebrand; it's a positioning move to own AI-assisted commerce across every platform where shopping might start. Amazon Now isn't just fast delivery; it's a direct counter to Walmart's store-as-warehouse advantage. ASCS isn't just a logistics service; it's a global data acquisition strategy at supply chain scale. Dynamic TV Creative and the LinkedIn DSP partnership aren't just ad products; they're proof that Amazon's advertising stack is closing in on true full-funnel capability.
The throughline for brands is consistent: the gap between those treating content, advertising, operations, and data as separate functions and those treating them as one connected system is widening. Fast.
Subscribe for More!
We publish a new Retail Roundup every month. Stay in touch on LinkedIn to keep track of the most important changes in the Retail media landscape.
Give It a Listen
You can tune in for the full interview with Armin Alispahic and Pat Petriello on the Ecommerce Braintrust hosted by Julie Spear and Jordan Ripley.
This show gives you access to the world's best brains when it comes to building momentum online for established consumer brands. Join in and listen to discussions with expert guests about e-commerce strategies, trends, and innovations.
Read more from this author: