You’re browsing for a new TV on your phone during your commute. The specs look good and the price seems fair.

But when it comes time to hit buy, you close the tab and wait until you’re home on your laptop.

Gen Z consumers may laugh but it’s typical Millennial behavior. But why?

Because the bigger screen feels more reliable, checkout feels safer, and you want one last look at the retailer’s site before committing.

Agentic commerce is challenging that.

Instead of switching devices or even visiting the retailer’s website, intelligent AI agents will do the heavy lifting: researching products, comparing options, evaluating delivery times, and in some cases, completing the purchase on your behalf.

Agentic commerce promises a seamless path from discovery to purchase, finding exactly what you want and paying for it in a single flow.  

We’re already seeing early examples:  

• Perplexity partnered with PayPal to build Buy with Pro, which lets users compare and purchase products directly within a chat.  

• Amazon’s Buy for Me completes purchases from external retailers without leaving the Amazon app. 

• Google’s new AI Mode shopping feature surfaces tailored results, tracks prices, and prompts you to buy at the right moment. 

• Both Visa and Mastercard have unveiled AI agents to securely browse, select, and pay on behalf of consumers.

Core features of agentic commerce

Autonomy 
The agent can act, not just suggest. For example, the agent chooses among several TV options and makes the purchase with your approval, or even without needing you to click “buy” each time. 

Context awareness 
You can input your preferences, budget, past behavior, calendar, and so on. Also, agents have access to real-time data like availability and delivery times. 

Multi-step workflows handled 
The agent can manage the entire purchase journey, from searching and comparing products, applying preferences or constraints, scheduling delivery, and even handling returns. 

Visibility and readiness 
For agentic commerce to work at scale, the backend (catalogs, inventory, logistics) must be machine-friendly, meaning clean structured data, APIs, and real-time inventory.

How retailers are using AI agents

Businesses are implementing AI to automate certain work processes to optimize daily operations. AI agents go a step further, taking on complex reasoning and multi-step problem-solving, acting more like an autonomous teammate.

Some of the common areas where AI agents are applied include:

Merchandizing 
AI agents can analyze sales data, customer behavior, and competitor pricing to build smarter promotion strategies. Instead of blanket discounts, they identify which products, like underperforming items, would benefit most from tailored promotions.

Inventory management 
Agents can monitor real-time sales trends, forecast demand based on factors like the season, and automatically trigger reorders when stock runs low. By taking into account things like supplier lead times and shipping delays, they help businesses reduce overstocking and prevent stockouts.

Customer service 
Beyond chatbots, AI agents can handle complex customer inquiries without human escalation as well as process returns. When a case does require a human, the agent can summarize the interaction and suggest next steps, saving time and improving customer satisfaction.

B2B commerce 
In procurement, AI agents act as virtual assistants that can automatically reorder supplies at pre-negotiated prices. Not only that, but they can also track delivery statuses and flag opportunities for cost savings. This helps to cut down on the manual workload and keep supply chains moving along.

How AI agents could change digital commerce 

The potential of agentic commerce is, like many technologies, disruptive and transformative.

Many view AI agents as the future of digital commerce, breaking up the monopoly of search engines when it comes to user intent, and become the default way of searching and purchasing online.

There are signals of long-term shifts in consumer behavior: ecommerce-related Google searches have declined 10% year on year. Nearly 1-in-6 consumers are using generative AI tools for product and service recommendations. And 71% want it integrated into their shopping experiences.

With consumer behavior evolving, here are some major shifts we could see in digital commerce: 

Search & discovery redefined

Discovery won’t always be via storefront search or browsing. It may happen through AI agents you prompt (“suggest me a 55-inch TV under £800, low environmental impact, good reviews”). Instead of starting with a website or app, discovery could begin with a simple prompt to an AI agent, which then compares products across different retailers for you. 

“Zero UI”

Shoppers may interact with AI entirely through voice or text prompts. Smart agents may even handle routine purchases automatically, such as reordering products, making manual clicks optional or rare.

Machines buying with brands discovered downstream

If agents decide, brands that are not “agent ready” (poor data, poor fulfillment, lack of machine-readable metadata) could be invisible. Product data, stock info, return and delivery policies, for example, become critical signals, not just solid UX. Brands may compete on how well their backend systems serve AI agents, not how their homepage looks, feels, and functions.

Supply chains become central to conversion

Speed, reliability, visibility of inventory, shipping: all must be visible to agents. If an AI agent can see shorter delivery or lower shipping cost, that may triumph over brand. The backend may take precedent over the frontend.

Deep research and reasoning

Increasingly complex agents are performing deep research and reasoning, helping to shape the shopping experience. They can check hundreds of sites and figure out which one is best for that individual customer. Their recommendations are guided by factors like credibility, relevance, unique value, and consumer trust. This makes it more important than ever for retailers to have a strong, convincing value proposition online.

The retailer’s roadmap for agentic commerce 

Large-scale shifts in consumer habits don’t happen overnight. But while adoption of AI-powered agents is still in its early stages, the signals are clear: consumers are experimenting with agentic tools, and the infrastructure to support them is rapidly emerging.

That means brands and retailers need to prepare. Here’s what the roadmap looks like:

1. Audit product data
   AI agents rely on accurate, structured product information. Clean up catalogs, standardize metadata, ensure real-time inventory is accessible. The better your backend communicates with AI, the more likely your products will be recommended and bought.

2. Make logistics machine-friendly
   Agents evaluate delivery times, stock availability, and return policies when making decisions. So streamline fulfillment processes, provide real-time tracking, and make your operations visible to AI. A fast, transparent supply chain can become a competitive advantage.

3. Implement API-first systems
   Expose your key systems, like catalogs, inventory, pricing, and checkout, through APIs. Event-driven architectures allow AI agents to respond instantly to changes in stock, pricing, or promotions, keeping your offerings relevant.

4. Enable trust, identity, and governance
   Consumers must feel confident letting AI act on their behalf. Build systems that verify agent identities, enforce permissions, and provide audit trails for purchases. Clear rules and safeguards around refunds, errors, and compliance will be critical.

5. Enforce security

Retailers have been the victim of increasing cyber-attacks in recent years. One study found ransomware attacks on retailers increased 58% in Q2 2025 compared to Q1. Put simply, security cannot be overstated. When this comes to AI agents, they must be grounded in trusted, verified data sources to prevent manipulation or bad recommendations.

In addition, retailers should establish strict guardrails that define what agents can and cannot do, alongside continuous monitoring to catch anomalies in real time. Audit logging is essential here, not only for compliance, but to trace decisions, resolve disputes, and most importantly, build trust with consumers.

6. Experiment with agentic interactions
Start small. Pilot AI-assisted discovery and automated reorders for specific products. Then, measure customer satisfaction and conversion impact before scaling. 

7. Evolve with the ecosystem
Agentic commerce is quickly So track AI platform developments, competitor initiatives, and regulatory updates. Be ready to adjust your strategy as new capabilities and consumer behaviors emerge.

Start today 

Agentic commerce is arriving. For retailers, success will depend on preparing today: optimizing data, modernizing infrastructure, and building trust with AI-driven interactions.

Those who embrace the shift early can turn AI agents into a powerful advantage.