The Maze: Social commerce is still growing. The threat is that AI commerce may grow on a steeper curve. U.S. social-commerce sales reach $102.68 billion in 2026 and $152.50 billion in 2029. Under EMARKETER's aggressive adoption case, AI-platform-driven ecommerce reaches $225.21 billion by 2029—$72.71 billion above social. This is not a forecast of social shopping dying. It is a forecast of the shopping gateway moving from feeds and creators toward assistants that can compress discovery, evaluation, and purchase into one conversation.
The established channel keeps expanding—and still loses relative ground. The forecast puts U.S. social-commerce sales above $100 billion in 2026, then lifts them 48.5% to $152.50 billion by 2029. That is healthy growth, roughly 14% a year. Yet the aggressive AI path starts near zero, catches social around 2028, and finishes 47.7% larger. Mature channels do not need to shrink to become strategically less important. They only need a faster gateway to capture the next increment of demand.
The middle case is the more useful warning. Three AI trajectories fan out from the same small base. Aggressive adoption clears social commerce by 2029. The base case ends just below it. Conservative adoption remains far smaller. The conclusion therefore depends on behavior, retailer integration, and whether assistants become a habitual starting point. The winning scenario is not preordained. But even the middle path puts AI beside a channel that required years of feed design, creator tooling, checkout integration, and advertising infrastructure to build.
Both totals use broad attribution, so the comparison is about gateways. In the analyst's methodology clarification, AI-driven sales include purchases completed inside a general assistant, originating from an attributable assistant link, involving a retailer's embedded shopping assistant, or routed through a specialist shopping agent. Social commerce includes native transactions such as TikTok Shop and Facebook Marketplace, plus purchases completed on a retailer site after a social ad click. These are expansive channel definitions, not mutually exclusive consumer identities.
Social may keep influence while AI captures decision control. Feeds and creators are good at creating desire. Assistants are designed to narrow options, compare trade-offs, and recommend the next action. That creates a new commercial split: social can remain the spark while AI becomes the filter. Attribution then gets awkward. The network, creator, publisher, assistant, retailer, and payment layer can all claim a piece of the same journey. Whoever owns the decision interface will have more leverage over product visibility, referral economics, and the data used to prove influence.
Why it matters: Ecommerce operators should stop treating AI commerce as another traffic source in the analytics menu. If assistants become the shopping gateway, product data must be machine-readable, recommendations must be measurable, and commercial agreements must reward the content that shaped the decision. Social platforms still own attention; retailers still own assortment and fulfilment. But assistants can sit between them and the buyer. That middle position is where comparison, attribution, and eventually margin control accumulate.
