20% of 2025 holiday sales driven by AI agents: rethinking discovery marketing

A new State of the Industry report, sponsored by Swap and based on a survey of 80 brand, retailer, and agency respondents, signals a transformative shift toward agentic commerce. While the report is promotional, the underlying trends it highlights are substantiated by broader industry data: according to Salesforce, AI and agents drove 20% of all retail sales during the 2025 holiday season, translating into $262 billion in spending. This is not a futuristic prediction; it is a present-day market reality that demands attention from e-commerce players. The core premise is that AI agents are no longer just tools for product discovery but are evolving into full-funnel guides that can evaluate options, facilitate decisions, and even complete purchases. This marks a profound departure from traditional e-commerce architectures, which rely on static product pages and manual navigation. Instead, agentic storefronts—conversational, context-aware interfaces—promise to shepherd shoppers from initial query to final checkout with minimal friction. Juan Pellerano-Rendon, CMO of Swap, articulates the key differentiator: these agents understand context, ask clarifying questions, and remember past interactions, thereby changing the economics of customer acquisition and retention. For brands, this means that while showing up in AI-generated search results (the discovery moment) is crucial, the real competitive advantage lies in the post-click experience. A static landing page that fails to engage can squander the opportunity, whereas an agent that continues the conversation can dramatically lift conversion rates. The report found that 90% of respondents are already using AI to improve the shopping experience, but most are likely at earlier stages—such as automated product recommendations or chatbots. The leap to a fully agentic storefront, where the AI owns the entire journey, is just beginning. This evolution challenges the existing ad-tech and marketing stack; brands may need to rethink attribution models, as last-click credit becomes blurred when an agent guides across multiple touchpoints. Data privacy and consumer trust will also be critical hurdles: an agent that remembers shopping history raises the specter of tracking debates. However, if executed well, the potential for hyper-personalization could lead to higher average order values, lower return rates, and stronger brand loyalty. The fact that Swap is sponsoring and leading this conversation suggests that a new software category is emerging, with early movers racing to establish standards. For retailers, the operational implications include overhauling website infrastructure, integrating with AI platforms, and upskilling teams. Agencies, meanwhile, must help brands optimize for AI-mediated journeys, a shift that echoes the SEO revolution of the early 2000s but with far greater complexity. As AI models continue to advance—capable of reasoning, generating rich media, and even negotiating—the agentic storefront could become the dominant interface for digital commerce. Forward-looking companies will treat this not as an incremental feature but as a strategic imperative.