Amazon quietly tests paying publishers per click as agentic AI era takes shape

Amazon is quietly piloting an advertising model that could reshape how products get discovered. Its Native Commerce Advertising program breaks with the classic affiliate model by offering publishers 20 to 60 cents per click generated by qualified traffic it sends to Amazon, whether or not the visitor buys.
Unlike the long-running associates program, where publishers earn only on completed purchases, the new model addresses a long-standing complaint: content creators drove massive traffic but earned nothing unless visitors converted immediately. Major media companies including the published report and Vox Media are already participating.
For sellers the implications cut several ways. Publishers now have stronger incentives for showcasing Amazon items in their content, potentially lifting discovery traffic. Amazon gains richer click-to-purchase journey data that could improve its ad targeting. And by paying for clicks, Amazon is competing directly with Google and other ad networks for publisher inventory. Expansion beyond the pilot is expected, with niche bloggers, YouTubers, podcasters and newsletter creators likely beneficiaries, though a click-based model rather than a commission one means the marketplace will tighten traffic-quality controls.
A companion analysis making the rounds frames a still larger shift: e-commerce software is entering an agentic era. The first era, traditional SaaS, gave sellers comprehensive but reactive platforms. The second brought composable, API-first building blocks. The third, now arriving, replaces both with AI agents that learn at inference time, combining a language-model brain, advanced reasoning, retrieval databases of retail knowledge, persistent memory and tools that act across business systems.
The practical picture for sellers: agents that automatically adapt to Amazon API changes, fix listing compliance issues before the seller notices, monitor competitors and adjust strategy in real time, predict inventory needs and pre-prepare reorders, and recommend new product opportunities from market gaps. Because learnings are shared across an agent platform's customers, a fix discovered for one seller benefits all.
The battle lines are already visible. Amazon blocks rival agents like Google's shopping bot while rolling out its own, including the Rufus assistant and a Buy for Me feature that purchases across platforms, building a defensive moat around the customer relationship. Analysts warn that decisions made over the coming year will decide if sellers control their digital destiny or become passive fulfillment nodes in someone else's AI ecosystem.
In other items: a new AI-powered image tool for Amazon listings entered beta, promising data-driven main images, lifestyle shots and infographics in minutes; review-generation services built on package-insert QR codes continue to grow within Amazon's terms of service; and a podcast episode traced how a 52-year-old actress turned homemade face cream into a multimillion-dollar skincare brand that landed on the Inc. 500. Trivia making the rounds: ChatGPT now fields 2.5 billion prompts a day, still a fraction of Google's roughly 13.7 billion daily searches.



