Glossary:Agentic Commerce
9 min.
What is Agentic Commerce?
Agentic Commerce refers to the approach where AI agents independently make purchasing decisions on behalf of consumers — without the human having to manually control every step. The agent searches offers, compares prices, checks availability, and completes transactions. All of this operates based on preferences and context that the user has set beforehand.
The model reverses the classic shopping experience: Instead of a human accessing a website, navigating, filtering, and purchasing, they delegate these tasks to an AI agent that completes them faster and more consistently. For businesses, this represents a fundamental shift — their website must now not only convince human visitors but also be reliably read and evaluated by AI systems.
40 million AI-supported sessions have already been processed by branchly — a sign of how quickly agent-based interactions are establishing themselves in the market (Source: branchly, 2026).
How does Agentic Commerce work?
An AI agent in the context of Agentic Commerce operates in several phases, some of which occur in parallel:
Understanding the task (Intent Parsing): The agent receives an order — either explicitly formulated ("Order the cheapest version of this product under €100") or derived from stored preferences. It breaks complex goals into specific sub-tasks.
Research and evaluation: The agent searches sources — product catalogs, price comparison sites, review platforms, and its own purchase histories. It evaluates results based on defined criteria: price, quality, delivery time, compatibility.
Making a decision: Based on the collected data, the agent selects the best option. In case of uncertainties or high risk, it requests human confirmation. For routine transactions, it operates autonomously.
Executing the transaction: The agent completes the purchase — including payment processing, address entry, and order confirmation. It communicates directly with the merchant's APIs.
Learning and adapting: Each transaction refines the user's model. Returns, changes, and feedback flow into future decisions.
For retailers looking to attract AI agents as customers, the technical quality of their digital presence is crucial: structured product data, clean APIs, clear availability information, and machine-readable content. branchly addresses this point — as an AI layer that makes website content accessible to both human visitors and AI agents.
Market size and growth
The numbers behind Agentic Commerce are impressive — and they are moving quickly.
McKinsey estimates the global market potential at $3 to $5 trillion by 2030 (Source: McKinsey, October 2025 — The Agentic Commerce Opportunity). This is not just about direct revenue, but a complete shift in how purchasing decisions are prepared, made, and executed.
Morgan Stanley estimates the share of US e-commerce by 2030 at $190 to $385 billion — which corresponds to 10 to 20% of the total market. Notably, 23% of Americans already report using AI systems while shopping (Source: Morgan Stanley, December 2025 — Agentic Commerce Market Impact Outlook).
The growth data at the traffic level confirms this. Adobe Digital Insights recorded a year-over-year growth of 4,700% in AI-generated traffic on retail websites. Visitors coming via AI channels exhibit significantly better behavior: +84% in revenue per visit, +32% in time spent, -27% in bounce rate (Source: Adobe Digital Insights, August 2025 — Generative AI-powered Shopping).
Salesforce adds to the picture: 39% of consumers already use AI for product searches. During the last holiday season, $260 billion in revenue was attributed to AI-driven purchasing decisions (Source: Salesforce, July 2025 — Consumer Shopping AI Trends).
These numbers make one thing clear: Agentic Commerce is not a future scenario. It is happening right now.
Agentic Commerce vs. classic online shop
Feature | Classic Online Shop | Agentic Commerce |
|---|---|---|
Decision-maker | Human visitor navigates manually | AI agent acts on behalf of the user |
Page structure | Optimized for human readability and visual persuasion | Optimized for machine readability (structured data, APIs) |
Purchase process | Multiple clicks, forms, manual decisions | Autonomous execution based on predefined preferences |
Personalization | Rule-based, segment-based, A/B tested | Individual per user, context-dependent, learning |
Comparison research | User manually switches between tabs and pages | Agent searches multiple sources in parallel within seconds |
Language barrier | Website language dictates experience | Agent interacts regardless of website language |
Touchpoint | Visitor comes to the website | Agent accesses API or structured data — without visible visit |
Success measurement | Pageviews, Bounce Rate, Session Duration | Transaction quality, agent compatibility, API availability |
For medium-sized enterprises, this shift means: Traffic from AI-driven channels behaves fundamentally differently than classic search engine traffic. branchly helps cater to both visitor types — through an AI layer that converts both human visitors and provides machine-readable structures.
Why Agentic Commerce is relevant for European companies
AI agents as a new customer group
An AI agent does not book a hotel because the homepage looks nice. It evaluates availability, price, review score, and cancellation terms — and does so in seconds. Websites that do not provide this information in a machine-readable way will simply be bypassed in agent-based purchasing processes.
This is a new form of visibility that goes beyond classic SEO: It is no longer just about ranking in Google search results but about being recognized and selected by AI agents as a trusted source.
The Insight Advantage
branchly installations across over 11 million users show a clear pattern: Websites with deeply integrated AI layer generate behavioral data that no other tool provides. This data becomes even more valuable in an Agentic Commerce world — as it reveals what questions AI agents ask, which information they prioritize, and where product pages lose clarity.
GDPR and EU AI Act
European companies face a dual challenge: They must seize the opportunities of Agentic Commerce without violating data protection requirements. AI agents operating on behalf of EU citizens are subject to the same privacy standards as direct users. branchly operates on Microsoft Azure in European data centers — GDPR-compliant, EU AI Act-compliant, and accessible according to WCAG. This means no compromises between technological leadership and regulatory security.
How branchly enables Agentic Commerce
branchly is built as an AI layer that integrates natively into a website — not as a superficial widget, but as a deeply woven part of the digital experience. This positions the platform particularly well for a world where AI agents become regular "visitors".
Advisor — AI-supported product advice
The Advisor module combines conversational AI with product catalog integration. It understands complex queries ("I'm looking for a laptop for graphic design, budget €1,200, must be lightweight"), filters the catalog based on criteria, and returns structured recommendations. For Agentic Commerce, this means: Even if an AI agent searches on behalf of the user, the Advisor delivers consistent, machine-readable responses — not just a visual interface.
Chatbot — generative AI chatbot
The branchly chatbot module is not a rule-based bot with decision trees. It accesses all website content, product catalogs, and connected data sources to generate contextually relevant responses. The interaction rates speak for themselves: As a floating widget, branchly implementations reach 5 to 10% of visitors (compared to an industry average of 0.5 to 1%). When embedded directly into a page, the rates rise to 45 to 50% (Source: branchly customer data, 2026). A generative AI chatbot that achieves these rates produces the amount of data needed for Agentic Commerce optimizations.
Navigator — guided purchasing processes
The Navigator module guides users through multi-step processes — configurations, booking flows, comparison logics. This structure is directly usable in an Agentic Commerce world: AI agents can interpret Navigator flows as machine-readable decision logic and derive the necessary steps for a transaction. branchly supports 101 languages natively — without separate translation work.
Agentic Commerce in practice: Typical scenarios
E-Commerce
A user configures their AI agent: "Order coffee that I like monthly, below €25, with sustainability label, delivery by Friday." The agent evaluates offers from multiple shops — price, rating, delivery conditions, certifications. It selects, purchases, and confirms. For the shop, this is an invisible visit without page views, but with a real transaction.
Shops that provide structured product data, clean checkout APIs, and clear certification information are chosen more often. branchly supports e-commerce teams in building these structures while simultaneously converting human visitors with the Advisor module.
Tourism
A traveler delegates the booking search: "Find a city hotel in Vienna for two nights in early May, under €150 per night, breakfast included, cancellable." The agent compares Booking.com, the hotel website, and regional providers. Websites that provide structured availability and pricing data while also showing clear cancellation policies win the booking order.
branchly customers like Tourismus Regensburg benefit from this doubly: The generative AI chatbot serves human visitors in 101 languages, while the structured content data in the background forms the basis for agent-based booking requests.
Financial services
A user commissions their AI agent: "Compare the three cheapest checking accounts without a minimum deposit for me and open one." The agent checks conditions, fee structures, and contract details on various bank websites — and only acts where the information is clear, complete, and machine-readable.
For financial companies, this means: Product information must not only be documented for compliance checks but also be evaluable by AI agents. branchly implementations in this segment combine GDPR-compliant data operation with structured, agent-friendly content — without compromising regulatory security.
Related terms
AI Chatbot
AI Product Advisory (Advisor)
Conversational AI
AI Search
RAG (Retrieval-Augmented Generation)
Agentic RAG
Frequently Asked Questions
What is Agentic Commerce in simple terms?
Agentic Commerce means that an AI agent shops for you — without you having to carry out each step yourself. You define your preferences (budget, quality criteria, delivery terms) once, and the agent searches, compares, and buys on your behalf. This ranges from the monthly ordering of recurring products to complex travel bookings.
How does Agentic Commerce differ from traditional e-commerce?
In traditional e-commerce, a person visits a website, navigates manually, and makes one purchasing decision after another. In Agentic Commerce, an AI agent takes these steps autonomously. The key difference: The agent evaluates offers in parallel, acts faster, and is not influenced by visual design decisions. For retailers, this changes what constitutes a "good" product page.
What technologies are behind Agentic Commerce?
Agentic Commerce builds on several layers of technology: Large Language Models (LLMs) for language understanding and decision logic, Retrieval-Augmented Generation (RAG) for access to external product data, Agentic-RAG architectures for multi-step research processes, and open APIs for transaction execution. These are complemented by structured data formats (Schema.org, JSON-LD) that make product information machine-readable.
How large is the market for Agentic Commerce?
Very large and growing. McKinsey estimates the global potential at $3 to $5 trillion by 2030 (McKinsey, October 2025). In U.S. e-commerce alone, Morgan Stanley expects a share of $190 to $385 billion by 2030, which would correspond to 10 to 20% of the total market (Morgan Stanley, December 2025). Even today, 23% of Americans use AI when shopping.
What must retailers provide technically to be ready for Agentic Commerce?
Three things are crucial: First, structured product data (Schema.org markup for price, availability, features), so that AI agents can reliably read product information. Second, clean APIs or machine-readable checkout processes, allowing agents to execute transactions. Third, clear, consistent content without contradictions — agents compare dozens of sources and choose the most reliable. branchly helps build these structures as an integrated AI layer.
How does Agentic Commerce change conversion optimization?
Classic CRO (Conversion Rate Optimization) optimizes for human perception: colors, copy, social proof, trust signals. In Agentic Commerce, these factors do not matter for the agent. Instead, the focus is on: data completeness, API stability, price transparency, and structured information. branchly provides behavioral data from human interactions that help optimize both dimensions simultaneously.
Is Agentic Commerce GDPR compliant?
That depends on the implementation. AI agents acting on behalf of EU citizens process personal data — purchase histories, preferences, payment data. All parties involved (users, agent providers, retailers) must act in compliance with GDPR. For retailers, this means: privacy notices must cover agent-based interactions, and data processing agreements (DPAs) must be concluded with AI platforms. branchly is GDPR compliant and EU-AI-Act-ready — with Microsoft Azure EU hosting.
Which industries benefit the most from Agentic Commerce?
The use in areas with a high transaction frequency and comparable products is growing particularly quickly: e-commerce (especially consumer goods and electronics), travel and tourism (flights, hotels, activities), financial products (accounts, insurance, investments), and B2B purchasing (recurring orders, supplier selection). Common to all: Purchase decisions can be made based on clear criteria that an AI agent can represent.
What does Agentic Commerce mean for SEO?
Search engine optimization remains relevant but changes its character. Instead of just ranking for human users on page 1, content must also be recognized as a reliable source by AI agents. This means: more structured data, more precise facts with sources, clearer product information, and a technical infrastructure that can be read by agents. Adobe Digital Insights observes that AI-driven traffic already generates 84% higher revenue per visit than traditional traffic (Adobe, August 2025).
What does an Agentic-Commerce-capable AI solution cost for businesses?
branchly offers an entry point starting at €499/month (Starter plan with 1,000 sessions). This plan already includes the modules relevant for an agent-friendly website infrastructure: structured content preparation, generative AI chatbot, advisor, and navigator. For medium to large websites with higher interaction volume, the Pro plan starts at €1,799/month with 5,000 sessions. Enterprise solutions are configured individually.
