E-commerce is on the verge of undergoing one of its most profound transformations since the advent of the smartphone and search. The key to this change is Agentic Commerce: a new way of shopping online where users will utilise artificial intelligence (AI) agents in the purchasing process.
We are starting from a context of change where the traditional web traffic landscape is changing forever. The use of generative AI (Google AI Overviews, Google AI Mode, ChatGPT, Claude and Perplexity) is transforming the way we search and discover things online, replacing traditional search engines.
Gartner predicts that search engine traffic will fall by 25% in 2026 and by 50% in 2028, thereby reducing organic website traffic. Traffic from LLMs will grow exponentially until it surpasses organic traffic (traditional search) in 2028, according to a study by SemRush.
How can we capture the traffic generated by LLMs?
The emergence of these new channels for searching for products and services makes it necessary to adjust e-commerce marketing strategies to be able to channel LLM traffic towards our website. To do this, the website must be given authority in all areas related to our commercial offering by designing a Generative Engine Optimisation (GEO) strategy.
The GEO strategy aims to be visible to LLMs in response to user queries and consists of two types of actions:
- On-site actions: optimising the structure, content and technical aspects of the website to make it AI-friendly.
- Actions on external pagesto our website in order to establish a presence on current AI sources: blogs, comparison sites, forums, reviews and other websites.
How does conversational technology affect e-commerce?
We are facing a new paradigm in which conversational models (chat) take precedence for the consideration, search, selection and payment of goods and services. The emergence of conversational models is not merely a change in UX, but rather represents a new transactional model based on natural language. Users engage in natural conversations to describe their needs using everyday language. This moves away from the use of specific keywords required by traditional search.
In practice, a user who opts for generative AI to satisfy their need for a specific good or service could previously discover and compare options within the LLM, make a purchase decision and go to the specific site of the chosen product to complete the purchase. But now a new model has emerged to improve this experience: agentic commerce.
What is Agentic Commerce and how does it work?
Agentic shopping, or Agentic Commerce, as mentioned at the start, is based on the use and delegation of the purchasing process to artificial intelligence (AI) agents. In the traditional model, the user searches, compares and buys. In Agentic Commerce, an AI agent does all this for you.
This new way of shopping redefines the traditional online purchasing funnel: it is no longer just a user journey through our website, but becomes an algorithmic decision regarding our product. Artificial Intelligence not only recommends products, but acts as an agent with the autonomy to make purchasing decisions on the user’s behalf.
The key difference lies in the shift from manual searching to delegation:
- Traditional Commerce: the user searches for “Padded jacket”, compares prices, reads reviews and buys.
- Agentic Commerce: the user tells the AI: “Buy me the best padded jacket for the cold for my trip to New York on a budget of €150”. The AI analyses data, searches for offers, checks stock and completes the transaction.
How can you adapt your online business to Agentic Commerce?
Agentic Commerce is not a passing fad; it is an evolution of e-commerce driven by AI where the user is no longer the primary customer. We must treat AI agents as the “primary customer” and adapt e-commerce to become their preferred choice by making optimisations to our online shop:
- Reorganise the catalogue: continuously review attributes, images, descriptions and reviews, considering how the AI interprets them and not just how a human user perceives them.
- Intent-driven content: content that responds clearly and contextually to the questions (prompts) that buyers ask in natural language, thinking from a conversational perspective.
- Establish connections with external AI agents: invest in APIs and connectivity to display inventory, prices, delivery times and terms transparently.
- Build proprietary brand agents capable of negotiating, responding to complex queries and keeping the value proposition alive in a machine-to-machine interaction environment.
Now is the time to start preparing to integrate into a world where AI agents can be our best customers or, at worst, our competitors.
