Amazon and Walmart Race to Become Merchants’ Operating System

The retail landscape used to have four primary levers: share of shoppers, share of sellers, share of advertising dollars and share of wallet.

However, artificial intelligence (AI) is creating the fifth and sixth levers: share of decision-making and share of intent infrastructure.

Headlines this week across the retail marketplace made that shift visible. Amazon’s upgraded Seller Assistant can continuously monitor a merchant’s business and recommend or execute actions within seller-defined limits. Its new Multichannel Selling tool brings orders, inventory and catalogs from Walmart, eBay, TikTok Shop and Shopify into Seller Central. And Amazon Ads Agent can now plan, manage and optimize advertising through natural-language instructions.

Meanwhile, Walmart’s Scintilla platform is evolving from retail analytics toward what Walmart calls “commerce intelligence,” using first-party data and AI to identify performance changes and opportunities, and connect those insights to actions such as replenishing inventory or launching Walmart Connect campaigns. The mega-retailer also announced Thursday (Oct. 1) a new featured called “Shop to Light” where customers can use the Walmart app to ask for help finding an item and the store system in response causes the product location to light up while a small LED on the shelf label flashes to help the customer spot it.

The retail platform is becoming the intelligence layer surrounding a transaction. If merchants begin operating their businesses through retailer-owned intelligence layers, who captures the software spend that previously went elsewhere?

Read more: Amazon Wants to Run Merchants’ Businesses Beyond Amazon 

The Retailer Is Becoming the Merchant’s Operating System

The PYMNTS Intelligence report “Why Main Street’s Digital Survivors Are Pulling Ahead: Four Years of Small Business Data“ found in May that among businesses using marketplaces, 48% reported increased sales through that channel between January 2022 and February 2026. The share of SMBs that sold through their own websites was 57%, underscoring how little of a business may be captured by looking at one marketplace alone.

After all, the merchant technology stack has traditionally been fragmented by design. A seller might use one platform for eCommerce, another for marketplace management, an enterprise resource planning (ERP) tool for inventory, a logistics provider for fulfillment, specialized software for advertising and analytics, and agencies or employees to coordinate everything between them.

Amazon’s announcements, however, are designed to remove those boundaries. Amazon already possesses information about demand, conversion, advertising, inventory, fulfillment, pricing and customer behavior. Multichannel Selling extends the view beyond Amazon by allowing sellers to manage Walmart, eBay, TikTok Shop and Shopify activity from Seller Central. Amazon said merchants using it act on its recommendations more than 90% of the time, while more than 230,000 sellers use the tool monthly.

Amazon is building software capable of helping a merchant decide what to stock, where to sell, how to advertise and how to fulfill an order, while simultaneously building software that helps a consumer decide what to buy.

Read more: Amazon Builds Its Own Answer to Walmart’s 5,000 Stores 

Walmart is pursuing the same opportunity from a different starting point. Scintilla is designed around Walmart’s first-party commerce data and connects intelligence to execution. Walmart’s stated vision includes identifying an item at risk of going out of stock and moving the supplier directly into replenishment, or translating an insight into a Walmart Connect media campaign.

Amazon and Walmart don’t necessarily need to replace ERP providers, commerce platforms, retail analytics companies, marketplace-management software or advertising technology companies outright. Not if they can become the interface through which merchants decide what those systems should do.

Logistics, as PYMNTS CEO Karen Webster argued in her January trends piece, is the story of retail in an agentic world. The PYMNTS Intelligence report “Will the 2026 Shopping Season Go Agentic?“ found that roughly 132 million people have made a retail purchase with AI assistance, while 22% now begin online retail research in an AI tool.

The buyer and seller aren’t autonomous, of course. Humans continue setting objectives, permissions and constraints. But more of the tactical decisions between those objectives and the eventual transaction can be delegated to software.

For all PYMNTS digital transformation and B2B coverage, subscribe to the daily digital transformation and B2B newsletters.