AI Agents Poised to Reshape E-commerce and Advertising

Nicholas Grus, Director of Research at ARK, and Varshika Prasanna, Research Associate, delve into the transformative potential of the AI consumer operating system. They argue that the shift from a 'query to answer' model to a 'query to action' model, spurred by AI agents, will fundamentally alter e-commerce and advertising.

The Agentic Era: A New Consumer Operating System

Grus outlines the evolution of consumer interaction with technology, moving from command-based interfaces to the web, then mobile, and now into the 'agentic era' initiated by the launch of ChatGPT. This new era is characterized by AI agents that can take action based on user queries, marking a significant departure from the traditional question-and-answer format.

The adoption rate of AI chatbots is remarkably fast, with penetration nearing 20% within three years, significantly outpacing the internet's adoption rate, which took seven years to reach a similar penetration level. This rapid adoption highlights the growing influence of AI in consumer behavior.

Transforming E-commerce with AI purchasing agents

Prasanna explains that 'agentic commerce' is becoming a reality in 2025, with AI purchasing agents playing a crucial role. These agents can compress the consumer shopping journey, from discovery and engagement to decision-making and purchase, by offering a high degree of personalization. This personalization is key, as 95% of the consumer journey often occurs before a purchase is even made.

The time it takes to complete a transaction is also shrinking dramatically. While a pre-internet purchase might have taken an hour, agentic transactions are predicted to occur in as little as 90 seconds. This acceleration is made possible by agentic protocols that allow retailers to connect their back-end systems, providing agents with the necessary context and data.

Key protocols enabling this shift include Anthropic's Model Context Protocol (MCP) for data access, and OpenAI and Stripe's Agentic Commerce Protocol (ACP), along with Google and Shopify's Universal Commerce Protocol (UCP) for transaction settlement. These protocols aim to unify the commerce stack, a significant improvement over the fragmented connections of the internet era.

The Economic Impact of AI Agents

The research forecasts a substantial economic impact, with AI agents expected to capture a larger share of digital transactions. This share is projected to grow from approximately 2% in 2025 to 25% by 2030, representing over $8 trillion in global online consumption. This growth is driven by consumers delegating more commerce activities to AI.

AI Search and Advertising Opportunities

Grus highlights advertising as another significant opportunity, particularly within the search domain. With search being a $350 billion market dominated by players like Google, AI is poised to disrupt this space by shifting search advertisements to the agentic world. AI search traffic is predicted to grow from 10% to 65% of global search traffic within five years.

While advertisers are historically slow to adopt new platforms, the growth of AI search and its advertising potential is undeniable. Companies like OpenAI are already exploring advertising within chatbot ecosystems, providing an early test case for this thesis. The monetization of AI search is expected to grow substantially, even with a built-in lag for advertiser adoption.

Monetization Strategies: Direct vs. Indirect

Currently, AI monetization is largely subscription-based. However, the research suggests that indirect monetization, through a take rate on commerce and advertising, presents a much larger opportunity. Drawing parallels with social media platforms, where indirect monetization drives the majority of revenue, AI platforms are expected to follow a similar path.

The combined market for AI-mediated lead generation, advertising, and commerce is projected to grow from $20 billion today to $900 billion by 2030. This growth is fueled by the new consumer operating system powered by AI agents, fundamentally changing how humans interact with computers and conduct commerce.