Introduction
Artificial intelligence is steadily transforming digital commerce, but the next shift may be far more radical than personalisation engines or recommendation systems.
In its report “The Agentic Commerce Opportunity: How AI Agents Are Ushering in a New Era for Consumers and Merchants,” McKinsey argues that the future of commerce will increasingly be mediated by autonomous AI agents that can search, evaluate, negotiate, and purchase products on behalf of users.
This emerging paradigm—often referred to as agentic commerce—represents a structural change in how transactions occur online. Instead of humans manually navigating websites and comparing options, AI agents will handle complex decision-making and purchasing workflows autonomously.
For engineers, product leaders, and digital strategists, this shift raises profound questions about platform architecture, data infrastructure, APIs, and trust systems.
In this article, we will explore the core insights from McKinsey’s research and discuss what they mean for organisations preparing for the next generation of digital commerce.
What Is Agentic Commerce?
Agentic commerce refers to a new model of digital commerce in which AI agents act on behalf of consumers or businesses to complete purchasing decisions and transactions.
These agents can:
Interpret user intent
Search across multiple marketplaces
Compare products and pricing
Negotiate or optimise offers
Execute transactions automatically
Unlike traditional e-commerce systems that rely on manual interaction, agentic commerce delegates parts of the decision-making process to intelligent software agents.
According to McKinsey, this shift could fundamentally reshape the customer journey, replacing traditional browsing experiences with intent-driven transactions mediated by AI systems.
In practical terms, this means that instead of a customer visiting a retailer’s website to buy a product, they might simply instruct an AI assistant:
“Find me the best running shoes for marathon training under £150 and order them before the weekend.”
The AI agent then performs the entire process—from discovery to checkout.
The Economic Potential of Agentic Commerce
McKinsey’s analysis suggests that the impact of this shift could be enormous.
By 2030, AI agents could influence between $3 trillion and $5 trillion in global commerce transactions, with up to $1 trillion of US retail spending alone mediated by AI systems.
These numbers illustrate that agentic commerce is not simply a technological curiosity—it represents a potential restructuring of digital markets.
Three economic dynamics underpin this transformation:
1. Reduced Transaction Friction
AI agents can instantly analyse large volumes of product data, eliminating the time consumers spend browsing, comparing, and evaluating options.
2. Hyper-personalised purchasing decisions
Agents can optimise purchases based on user preferences, budgets, past behaviour, and contextual signals.
3. Continuous optimisation
Unlike human shoppers, agents can constantly monitor markets and update decisions in real time.
From a systems perspective, this means commerce shifts from episodic shopping journeys to continuous optimisation processes.
From E-Commerce to Intent-Driven Commerce
One of McKinsey’s most important insights is that agentic commerce changes the fundamental architecture of the customer journey.
Traditional e-commerce follows a familiar funnel:
Discovery
Product comparison
Decision
Checkout
In an agentic world, this funnel collapses into a single step: intent.
The AI agent becomes responsible for the entire workflow.
This transformation has major implications for how brands compete:
Visibility in search may matter less than machine-readable product data
Brand experience may shift from visual design to algorithmic trust signals
Price comparison could become fully automated
In other words, the “storefront” of the future may not be a website—it may be an API endpoint consumed by autonomous agents.
The Automation Curve of Agentic Commerce
McKinsey also frames the evolution of agentic commerce through an automation curve, describing increasing levels of AI delegation.
At the lowest level, AI tools simply assist users.
Examples include:
recommendation engines
search assistants
product comparison tools
At higher levels of automation, agents begin to act on behalf of users, performing tasks such as:
subscription management
inventory replenishment
travel booking
procurement workflows
At the most advanced level, AI agents operate with full decision authority within defined constraints, making purchases independently.
The progression mirrors earlier automation trends seen in finance, logistics, and manufacturing: systems gradually move from decision support to autonomous execution.
Why Infrastructure Matters: The Engineering Challenge
While much of the discussion around AI focuses on models and prompts, McKinsey emphasises that the real challenge lies in building the infrastructure that enables agentic commerce to operate at scale.
This infrastructure includes several key components.
Structured Product Data
AI agents rely on machine-readable product information to evaluate options.
This means merchants must ensure that:
product attributes are structured
metadata is complete
information is standardised
Unstructured content—such as marketing copy or product images—may become less relevant than structured data that algorithms can reason about.
Agent-Ready APIs
Commerce platforms will need APIs that allow AI agents to:
query product catalogues
compare pricing
access inventory data
execute transactions
In this model, APIs effectively become the primary interface for commerce systems, replacing traditional user interfaces in many contexts.
Identity and Trust Layers
If AI agents can make purchases autonomously, platforms must implement robust systems for:
authentication
permissions
spending limits
fraud prevention
Trust frameworks will be critical to ensuring that agents operate safely and responsibly.
Payments and Transaction Protocols
Agent-initiated payments require new capabilities in payment infrastructure.
These include:
tokenised payment credentials
machine-to-machine authorisation
continuous risk monitoring
Without these mechanisms, autonomous commerce systems cannot operate securely.
Business Model Disruption
One of the most striking implications of agentic commerce is how it could disrupt traditional digital marketing and retail strategies.
Historically, brands have competed for attention through:
search rankings
advertising
website design
brand storytelling
In an agentic world, AI agents become the gatekeepers of demand.
This means merchants must optimise not only for human perception but also for algorithmic decision-making systems.
Examples of new optimisation strategies may include:
structured product attributes
real-time pricing APIs
transparent product data
reliable fulfilment signals
In effect, algorithmic reputation could become more important than traditional marketing.
Trust, Risk, and Governance
Despite its potential, agentic commerce also introduces significant challenges.
One of the biggest is trust.
Consumers must trust that AI agents will:
represent their interests
avoid biased recommendations
protect their data
make safe purchasing decisions
Meanwhile, merchants must trust that agents interacting with their systems are legitimate and not malicious.
To address these concerns, organisations will likely need new governance frameworks covering:
explainability of AI decisions
auditing of agent behaviour
regulatory compliance
dispute resolution mechanisms
These governance challenges echo broader debates in Responsible AI, where transparency and accountability are critical to adoption.
Strategic Implications for Businesses
For companies operating in digital commerce, McKinsey’s report highlights several strategic priorities.
1. Prepare Product Data for Machine Consumption
Product catalogues must evolve from marketing assets into structured data products.
2. Build Agent-Compatible Platforms
Systems should expose APIs that allow AI agents to interact with commerce infrastructure.
3. Experiment with Agent-First Experiences
Companies should test how AI assistants integrate with their purchasing workflows.
4. Invest in Trust and Governance
Strong governance frameworks will be essential to maintain user confidence in autonomous systems.
Early movers that prepare their infrastructure today may capture disproportionate advantages as agentic commerce grows.
Why This Matters for Engineers
Although the conversation around AI often focuses on models like GPT-style systems, the agentic commerce paradigm highlights a deeper reality:
The future of AI-driven systems will be defined by architecture, data infrastructure, and interoperability.
Engineers building commerce platforms must therefore think beyond traditional web experiences and consider:
machine-to-machine interactions
structured data ecosystems
agent-oriented APIs
autonomous transaction workflows
In other words, AI agents may soon become the primary “users” of many digital systems.
Designing for that future requires a new engineering mindset.
Conclusion
McKinsey’s report on agentic commerce paints a compelling picture of how AI agents could reshape digital markets over the coming decade.
The core idea is simple but powerful: commerce will increasingly be mediated by intelligent software agents acting on behalf of humans.
This transformation could unlock trillions of dollars in economic value while fundamentally changing how consumers discover products, how merchants compete, and how digital platforms are built.
For engineers and digital leaders, the challenge is clear.
Preparing for the agentic era will require building systems that are:
machine-readable
API-driven
secure and trustworthy
designed for autonomous interaction
Organisations that embrace this shift early may define the next generation of digital commerce platforms.
Those that do not risk becoming invisible in a marketplace where the most important customer is no longer human—but an AI agent.
FAQs
1. What is agentic commerce?
Agentic commerce is a form of digital commerce where autonomous AI agents perform tasks such as product discovery, comparison, and purchasing on behalf of users or organisations.
2. How big is the agentic commerce opportunity?
McKinsey estimates that by 2030, AI agents could influence $3 trillion to $5 trillion in global commerce transactions, with significant adoption in retail and digital services.
3. How does agentic commerce differ from traditional e-commerce?
Traditional e-commerce requires human interaction at key decision points. Agentic commerce delegates parts of the purchasing process to AI agents that can analyse options and complete transactions autonomously.
4. What infrastructure is required for agentic commerce?
Key infrastructure components include:
structured product data
agent-compatible APIs
secure identity and permission systems
automated payment protocols
governance and auditing frameworks
5. What should businesses do today to prepare?
Organisations should focus on:
structuring product data
building API-driven commerce platforms
testing AI-powered purchasing workflows
implementing responsible AI governance frameworks
