On January 27, 2026, Amazon closed every Amazon Fresh and Amazon Go store in the country. The trade press called it a retreat. The data says otherwise: same-day perishable delivery grew roughly 40x in the year around the closures, and Amazon is now spending $4 billion to reach 13,000 new zip codes. Here's what actually changed, and what it means for the next 90 days.
Because they never earned the shelf space they occupied, and Amazon said so itself, years in advance.
Physical grocery runs on 1 to 3% net margins. Amazon loaded its Fresh stores with Just Walk Out technology, Dash Carts, and custom fixtures, none of which gave a shopper a reason to drive past the Kroger they'd been using for years. Amazon booked a $720 million impairment charge against Fresh and Go store assets in Q4 2022, tied to property, equipment, and operating leases. That's not a speed bump. That's a write-off on a thesis that wasn't working, disclosed three years before the stores actually closed.
CEO Andy Jassy said as much directly: "We won't expand unless we see that type of resonance. We're not just going to be undisciplined." Amazon's own retail leadership put it plainer still: the stores hadn't gotten the fundamentals right. What closed on January 27, 2026, 57 Fresh stores and 15 Go stores, per the internal memo CNBC obtained, wasn't a snap decision. It was the last step of a plan Amazon had been telegraphing since the impairment charge four years earlier.
Into the channels growing faster than the stores ever did.
Amazon's grocery business generates more than $150 billion in gross sales and serves over 150 million customers a year, a figure that spans delivery, Whole Foods, and shelf-stable grocery sold through Amazon.com. None of that volume evaporated when the stores closed. It concentrated into three channels already growing faster than physical retail could.
| Where the Volume Moved | What Changed | What It Means for Your Brand |
|---|---|---|
| Same-day perishables | Expanded from ~1,000 to 2,300+ cities; monthly active perishables customers up 50%+ since January | The unified cart is the moat: a customer adding groceries to a same-day order shops twice as often and adds triple the items |
| Whole Foods (physical) | Now Amazon's only physical grocery brand; 550+ stores, 100+ more planned, Daily Shop format doubling to 10 locations | The most direct on-ramp to Amazon's physical footprint for natural, organic, and specialty brands |
| Shelf-stable Amazon.com | $100B+ in gross grocery sales, compounding through Subscribe & Save | The quiet, undisrupted growth channel for pantry, snacks, and household staples |
| Rural delivery (new) | $4B investment, 200+ new stations, 13,000 new zip codes, 4,000 communities gaining same-day access | Uncontested territory: no incumbent with a five-year review and ranking head start |
The through-line across all four is the unified cart. A customer ordering laundry detergent adds strawberries. A customer buying diapers adds frozen meals. Grocery becomes the cross-sell layer for the entire Amazon ecosystem, a structural advantage no standalone grocer can replicate, regardless of how good their own delivery gets.
Three things, and all three are already reshaping where revenue shows up.
Product discovery is now entirely digital, and Amazon's AI shopping layer just changed underneath it. Rufus, used by 300 million customers in 2025, credited with nearly $12 billion in incremental annualized sales, and making shoppers 60% more likely to complete a purchase, was retired on May 13, 2026 and folded into a single, unified assistant called Alexa for Shopping. A shopper asking it "what's a good high-protein snack for kids" gets a recommendation built from conversational content, dietary callouts, and review sentiment, not whoever bid highest on "protein snack." Listings still written for keyword-matching are optimizing for a system that no longer exists in isolation.
Subscribe & Save moved from nice-to-have to survival strategy. Top-performing grocery brands drive 20 to 35% of total Amazon revenue through S&S, and every active subscription is a competitor permanently locked out of that purchase cycle. Brands running tiered incentive structures, not flat discounts, see the strongest conversion and the lowest skip rates.
Advertising budgets that haven't shifted are bleeding money against a channel that rewards it more than almost any other. Grocery converts at 30 to 35% on Amazon, among the highest of any category, which makes allocation matter more here than almost anywhere else. Brands capturing displaced demand right now run roughly 60% on conversion (Sponsored Products, high-intent terms), 20% on consideration (Sponsored Brands, category targeting), and 20% on awareness (DSP, Streaming TV). Kraft Heinz documented a 40% sales lift on Crystal Light, Mio Energy, and Kool-Aid through in-store digital advertising at Whole Foods alone: that's not a test result, that's a working format.
Brands capturing the volume moving off the shelves run roughly 60% on conversion, 20% on consideration, and 20% on awareness, not last year's spend pattern.
Amazon telegraphed this reorganization in writing, on the record, for more than a year before a single store closed.
The organizational record is public and dated. Tony Hoggett, then head of Worldwide Grocery, departed in November 2024. Whole Foods CEO Jason Buechel was appointed to a newly created VP role over all of Worldwide Grocery Stores, Whole Foods, Fresh, and Go, unified under one executive for the first time, in January 2025. Amazon merged the Go and Fresh corporate teams that March. By June 2025, a fully restructured grocery leadership board was named, with Whole Foods' SVP of Merchandising given oversight of Fresh and Go merchandising specifically, the concrete answer to whether vendor-facing leadership had actually merged, more than a year before this research. Whole Foods corporate employees moved onto Amazon's own benefits and compensation structure starting that December, with full integration confirmed for December 2026. The January 27, 2026 store closures were the visible last step of a sequence that had been running, on the record, for fourteen months.
Amazon's grocery reorganization ran on the public record for fourteen months before a single store closed.
A Marketplace Assessment is a senior operator's read on your same-day eligibility, AI-discovery readiness, Subscribe & Save performance, and ad allocation, tailored to your portfolio and current Amazon maturity.
Get Your Free Amazon Grocery Strategy Assessment