Amazon Shut the Door, Walmart Opened It: AI Shopping and a $74.4 Billion Ad Market
If shoppers hand purchases to AI agents, retail media may stop being a fight over the first screen. The contest could turn on who the AI picks and who completes the order.
The 20-second read
1. What happens to ads when the search box disappears?
Search for wireless earbuds on Amazon and ads appear between the product listings. Brands pay for those slots, and Amazon turns shoppers' attention into revenue.
Now suppose a consumer tells an AI: "Buy me wireless earbuds under $100 with good call quality and fast delivery." If the AI handles everything from comparison to checkout, the person may never see the search results, or the ads inside them.
| Metric | Figure |
|---|---|
| 2026 US retail media ad market forecast | $74.4 billion |
| Forecast combined share of Amazon and Walmart | 88% |
In the market as Emarketer estimates it, the two companies' combined share comes to about $65.5 billion. That is not the sum of their actual ad revenue. It is a calculation that applies market share to the market's size. The point is that a huge business built on selling shoppers' attention now faces a change in how people buy.
2. Amazon closes the door, Walmart opens it
In September, Amazon blocked Meta's shopping agent, Muse. Amazon said Muse appeared to have accessed its site without authorization and to have collected and stored customer credentials. The choice amounts to a refusal to hand customer information and the shopping experience to outside services.
Walmart went the other way. On Oct. 21, it unveiled infrastructure that lets outside shopping agents link to customer accounts and buy products. Wherever consumers talk to an AI, the path to buying Walmart's goods stays open.
Amazon: control the touchpoint. It blocks outside agents and puts weight on protecting customer relationships, data and its own shopping experience.
Walmart: connect outward. It allows agent integration and aims to capture sales wherever an AI starts a purchase.
The side that opens the door does not necessarily win. Amazon has its own AI shopping service. If customers keep shopping inside Amazon's AI, it may be able to protect both ad revenue and purchase data.
3. What the AI reads matters more than the ad
For ads aimed at people, striking images and copy matter. An AI, by contrast, has to find products that fit the customer's criteria. Whether it can accurately read product details, price, inventory and delivery terms could weigh more heavily on the purchase decision.
The flow looks like this:
- Before: search, ad exposure, customer click
- Change: customer gives the AI conditions, AI compares products
- New contest: AI recommendation, checkout, fast delivery
Platform power shifts along the way. Today, the shopping site decides what to show customers first. Going forward, the AI that talks with the customer may narrow the candidate products first. Retailers would then compete on product data that AI selects well and on the ability to fulfill real orders.
Morgan Stanley researchers have suggested that by 2030, AI agents could handle 10% to 20% of US e-commerce spending. As that share grows, the roles of the company that finds products and the company that delivers them may separate.
4. Will ad revenue move rather than vanish?
Traditional advertising places products in the search results customers see. If AI shopping grows, ad and fee models tied to the AI's recommendation process or to actual purchase results could become more important.
That creates new stakeholders. Walmart has a chance to receive purchase demand sent by outside AI. Meta and other AI service companies have a chance to control the front end of purchase decisions. Players strong in inventory and delivery may also gain power by turning AI recommendations into real transactions.
But that does not mean Amazon's ad business is about to collapse. Amazon executives have said outside agents have not had a meaningful effect on the current ad business. Placing ads inside its own AI shopping service could also reshape its revenue model.
- No. Even if search ads shown to people shrink as a share of the total, ads and fees tied to AI recommendations or purchase results can emerge. What changes is where ads appear and who gets charged.
- It has the advantage of being able to capture the purchase demand that outside AI secures. On the other hand, the first point of conversation with the customer may remain with the AI service. Expanding sales and holding control of the customer relationship are separate issues.
- Whether Amazon can protect its advertising business inside its own AI while also keeping its competitiveness in inventory, payments and delivery in an environment where outside AI starts the orders. Ad revenue growth and access policy need to be watched together.
What to watch
Insight Times Editorial Desk
