Why Walmart, Amazon and Alibaba Will Control 50% of Retail Sales within 20 Years

Amazon, Walmart and Alibaba are assembling the physical capabilities and data-driven interaction services to provide product discovery, transaction, and fulfillment for everything a consumer needs.

Originally published in Internet Retailer.

The recent news about Amazon’s intention to acquire Whole Foods generated plenty of analysis. And Walmart’s tit-for-tat reaction by trying to ban its partners from Amazon Web Service has done much the same. Nonetheless, I was surprised to find a number of gaps in the coverage.

The Whole Foods announcement, in particular, spawned speculations about Amazon’s ambition in grocery. Their ambitions (as with those of Walmart and Alibaba) are much larger.

Amazon is building a new operating system of commerce. Walmart is doing the same, as is Alibaba. I see them assembling the physical capabilities and data-driven interaction services to provide product discovery, transaction, and fulfillment for everything a consumer needs. The goal is not dominance of grocery or custom apparel or any particular product category, but rather to create the total customer satisfaction delivery system. From a business perspective, the goal is to maximize share of wallet among Prime customers (and their corollaries at Walmart and Alibaba).

My analogy: Amazon is building the WinTel-based PC of retail while most other retailers are trying to market their Commodore 64 or TRS-80.


In platform-based businesses where scale and network effects are decisive, there is only room for one, two, or some few number of basic operating systems.


At its core, retail is distribution. The relentless competition among retailers means the relentless pursuit of efficiency. I believe the leaders of these companies recognize that efficiency in retail (i.e., distribution) can only be sustainably achieved through well-implemented scale.

Scale enables operational efficiencies as well as commercial advantages through stronger negotiating positions. Walmart exerted its leverage to transform the inbound supply chain. Amazon exerted its leverage to transform the product discovery, price comparison, purchase, and last-mile delivery experiences.

The key thing brands need to realize is that in platform-based businesses where scale and network effects are decisive, there is only room for one, two, or some few number of basic operating systems. There is room globally for Windows, iOS, and niche computer operating systems. There is room nationally for Fedex, UPS, and USPS. There is room globally for Amazon, Walmart, and who else?

Within six months, all Whole Foods stores are activated on Prime Now, and online grocery will become “mainstream” in urban areas of the U.S. by Christmas. (By comparison, online grocery is already relatively mainstream in South Korea and urban China.)  And within one year there will be some 450 “Amazon Fresh” stores across the U.S., each with a section that is Amazon Go-powered exclusively for Prime members. They’ll be like airline club lounges for Prime members. Each of these will also be a free returns center for Amazon merchandise.

Looking further ahead, within three years we’ll see that 20% or more of some grocery categories (Dog food, paper goods, diapers) will be purchased online; Amazon Flex will become an Uber competitor that shuttles Prime members from their homes to the store (as well as other places)—perhaps trips are free if you purchase over $100 during the trip; and Prime members will tap their phones when they check out at (or into) Whole Foods/Amazon Fresh.

The additional data that Amazon now acquires on these Prime members will enable Amazon to determine what products could be more efficiently and conveniently delivered directly to the consumer’s home, or which ones make most sense for Subscribe and Save. (With the extra data, Alexa could suggest alternatives to your regular brand purchases: “Would you like me to reorder your regular dog food, or do you want to try something different?”)

Within 5 to 10 years the scale of these changes becomes more apparent. Consumers will purchase as much from Amazon as from Walmart (which today has about 11% of retail sales excluding auto, gasoline, restaurants and travel and  approximately 22% of grocery sales). Consumers won’t see as many Amazon stores as Walmart stores, but the share of mind and share of purchase will be comparable. Amazon and Walmart each may build (or buy?) a network of “Starbucks-like” locations to better distribute and collect merchandise, provide services, and showroom product.

USPS, FedEx, and UPS will probably contract with Amazon so that Amazon can do the last-mile delivery to certain consumer neighborhoods. New houses and apartments will be built with special equipment to handle daily Amazon and Walmart deliveries and pickups. And “Amazon Go” will become an externalized retail platform used by small and local retailers similar to how Amazon’s Marketplace, FBA, and Pay services became externalized services.

In 10-20 years, we’ll see somewhere between 50-75% of all retail transactions (excluding gasoline, restaurants, and travel) flow through the Amazon, Alibaba, and Walmart operating platforms. About 25-50% will flow through remaining local convenience stores, farmers’ markets, specialized boutiques, brand-owned and operated stores, and the remainder through “big box” and nationalized chains that still manage to hang on with tailored value propositions. This is not crazy. Walmart already has about 22% of such retail sales, and Amazon in the U.S. already has approximately 50% of e-commerce sales. In China, Alibaba accounts for over 75% of e-commerce sales. This will take 10-20 years.

So, what should brands do?

The important realization is that Amazon, Alibaba, and Walmart are not retailers in the traditional sense. They are platforms that are eliminating traditional retailers. In a nutshell, brands need to become expert operators of these new platforms, which requires them to develop new skills that look a lot like those of a retailer—being a manufacturer and brand manager is no longer enough.

Join the conversation